Re: [Emc-users] Installation of emc2.2
Craig Muller wrote: Hi All I Previously had emc installed via the BDI I have recently tried to upgrade to emc2,2 . I downloaded the ISO and cut it to a disc seemingly without a hiccup. I can boot my machine using the live cd however when I try to install the software the installation hangs at step 3 (Keyboard layout) does any one have any I deas how to solve this? make sure you have enough video memory on your video card. I think you need 4 MB. If you have on the motherboard video, you may need to add a plug-in video card. You need at least 2-3 GB of hard drive and 256 MB of main memory. If you disk is partitioned from the BDI, that may be confusing this install. You might have to do a custom install and delete all partitions, or do some tinkering if you want to save files on the system. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Bandit Steppers
Kirk Wallace wrote: The IA and IB sink the LED's on two optocoupler inputs each. The high side of the LED's is driven by the Q and /Q quadrature signals. I am guessing this forms a NAND function or an Inhibit? I wonder if the inhibits are needed to prevent shoot-through? If that is the case I will need to be very careful not to allow shoot-through. Here are some pictures of one of the drives. http://www.wallacecompany.com/machine_shop/Shizuoka/1-1a.jpg http://www.wallacecompany.com/machine_shop/Shizuoka/2-1a.jpg I am wondering if this is just a simple H-bridge with current limit? If anyone has more information on this drive, I would appreciate hearing it. It seems to have a LOT of parts for an H-bridge, and you mentioned before common wires on the motors. So, it sounds like a very typical current-limiting unipolar drive. It would need 4 transistors for the unipolar phases, and then more stuff for the current regulator(s). I see 4 copies of some circuitry at the bottom of your first picture, so I'm guessing the bottom 4 transistors are those phase drivers. There are 7 big diodes, too. I can't quite explain the whole topology from the pics, but a dual H-bridge would not use the motor center taps, and would need 8 identical power transistors. You have 6 identical TO-218 units, and 2 TO-220s and one more (different) TO-218. Yes, it seems reasonable they need to turn off one transistor before turning on the opposing one. It wouldn't actually cause shoot through, but an equally rough situation due to the center-tapped transformer action of the motor winding. If those transistors are Darlingtons (quite likely) then they can take a LONG time to turn off, in the several microseconds range. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Bandit Steppers
Kirk Wallace wrote: I scoped the driver board inputs again and corrected my signal diagram on the schematic. The gray traces are the result of the coil input and it's inhibit. I confirmed this with scoping the far side of the input optocoupler. I played with the axis speed and noticed the inhibit gets smaller and goes away around half speed. The trace leads me to believe the designers were trying to soften the low speed steps, with a brief half step, so that the motor would not overshoot and degrade the surface finish. If this is the case and not for some sort of electrical reason, I can switch over to EMC2 right away, and not worry about blowing out a board. Of course Stepgen doesn't have this inhibit signal available, so I may need to find another way, maybe using selective and adjustable delays in HAL. It would be real easy to rig an edge detector to detect when the phases changed, and enforce a blanking on that phase for a set time. A one shot could set the blanking time and allow you to adjust to the same as the Bandit controller produced. If step overshoot is a common stepper problem, is there a common way to fix it? Do you mean jumping steps? The worst cause of this is mid-band resonance, where the (full) step frequency matches the rotor's pendulum-like natural frequency due to rotor mass and the magnetic pull of the motor's poles. The Gecko drive detects an electrica phase shift in the motor's waveform and counteracts it. Otherwise, you can have the controller skip over the resonant frequency, install viscous dampers, or try to come up with other electronic damping solutions (RC dampers, etc.) None of these schemes is perfect, I think the latter RC dampers are almost certainly a disaster due to component size. The Gecko also greatly solves the problem with microstepping, as it reduces the excess energy supplied at the full-step frequency so as not to excite the resonance. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Bandit Steppers
Kirk Wallace wrote: I mean that when the motor is commanded to take a step the shaft will try its best to reach the new step, overshoot and then settle down. If this over shoot is large enough, it might cause the cutter to cut too deep for an instant. At moderate speeds, I suspect this isn't a problem because the motor might never come to rest between steps. I got this thought when I saw that with increasing speed the inhibit signal got smaller as a percentage of the already decreasing step pulse length and went away at mid speed. If the inhibit were for electrical protection the inhibit signal would get larger, relative to the pulse width. Hmm, it is more sophisticated than I had thought! It apparently gradually goes from half-step to full-step as speed increases. Very ingenious! Well, as for stair-step motion, there is no simple cure. You can gear down the stepper with a belt and pulleys, but then you lose top speed, rather severely with steppers. This is why a lot of higher-end systems go to servo motors. You can get encoders with an arbitrary number of counts/rev. and the speed limit is much higher. Micro-stepping and Geckos sound like a good thing, except they're not exactly cheap and are in the budget. It would almost be cheaper to go with a UnivPWM system, which would be even better. If I can get my old stepper drivers sorted out, the conversion to EMC2 will be basically free. I will only need a PC and breakout card that I already have. If these drivers don't work, a servo system may not be possible because there is no room for a belt reduction on X, unless I tolerate the loss of some travel on one side. Microstepping doesn't solve the problem, either. The microsteps build up force until the machine moves (some amount) in a jump. Might be less than, could be even more than, a full step. You'll never know. With a servo system that feeds back to the CPU, you can always check the performance, and at least read the jumpiness of movements as seen at the encoder. There are all sorts of ways to re-rig motors to clear moving parts. If you use a big enough motor, you CAN direct-drive the screws. Just select a high resolution encoder there. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] 5 axis cinci
Stuart Stevenson wrote: Gentlemen, Here is a video of my cinci. Notice the dual arm tool changer. It is very generic and can do many things. :) :) that's me. http://www.youtube.com/watch?v=mxxdq6y8z8M Wow, that is pretty cool! I'm usually not intimidated by big stuff, but there's a place near the end where the ram holding the head moves outward toward you, and that made me lean back in my chair! So, that is using my PPMC board set? Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Very low PRM system.
[EMAIL PROTECTED] wrote: Hi I want to build system to tool grinder and that means that my system should behave stable on very low RPM. Minimum PRM 0.0001 per minute. I want to use direct drive with ball screw 5 rev per inch or 1 rev for 5 mm in case of metric system. My motors have 8192 pulse per ….. it takes 4x8192 per revolution. It is AG industrial from servo dynamics. Any fundamentals ideas? Should I use higher resolution encoder? If you really need .0001 Rev/Minute, that is 32768 counts * .0001 = 3.3 encoder counts/MINUTE, or 18 seconds between each encoder count. You can't get smooth motion like that. Of course, .0001 RPM x 5 TPI on the screw is a movement rate of .2 IPM. Do you truly need it this slow? To get smooth motion, you really want an encoder count rate of maybe 15 counts/second, or even better, 60. So, for 15 cts/sec at .0001 RPM, you need 15 * 60 * 1 = 9 000 000 counts/rev, or 2 250 000 pulses/rev. This will be a pretty expensive encoder. (Multiply by 4 for 60 counts/sec.) Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] PID-tuning
Anders Wallin wrote: Hi all, I got my x-axis servo motor mounted today and ran some tests. There is a video and some hal-scope screen shots here: http://www.anderswallin.net/2008/04/x-axis-test/ This is a system with no motor current, motor voltage, or velocity feedback. An m5i20 generates PWM output for pico-systems DC-brush amps and 4000 pulse/rev encoders provide feedback. This might make the system a bit different to tune from other systems which run another loop on current/voltage/velocity. I found FF1 worked quite predictably and reduced the cruise-phase error to very little. Only P and FF1 was needed for that. (2nd pic on page) Then I added a lot of I-gain which helped a bit for the acc/dec phases. With FF2 I became confused on a higher level (3rd pic on page). It seems to either fix the acc. phase or the dec. phase but not both. I couldn't find your serial number in my data base, so I don't know what version of it you have. The earliest models had a little bug in them during deceleration. The drive uses a sign-magnitude scheme where it shorts the motor whenever it is not applying power from the DC supply. If the current exceeds the current limit, it goes back to shorting. This is fine during acceleration, but makes the problem worse during deceleration.These are the rev A through C versions. So, I had to change the circuit to go to open-circuit on current limit, since the current sense logic can't tell acceleration from deceleration. Rev D and E have this fix. The last pic shows moves at different feed-rates. All of them show around 16 to 25 count spikes during the acc/dec. phases. Presumably at the moment when acceleration changes and jerk is infinite. How much have other people been able to reduce these spikes? Well, that tells me either the FF1 is correcting too little or too much, or you are demanding more accel/decel current than your current limit is set for. It is VERY easy to over-correct with FF2, so I always check for having too much. Since you are direct-driving with a small motor, it is also easy to ask for too much torque from the motor. See http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?PWM_Servo_Amplifiers for some of my tuning results. The last picture actually represents a fair number of encoder counts, but at the resolution of 128,000 counts/inch, it is still a far smaller movement than the machine's accuracy. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] PID-tuning
Anders Wallin wrote: I couldn't find your serial number in my data base, so I don't know what version of it you have. The earliest models had a you sent these to me in late December 2006. Can you tell which rev. I have based on the date? Yes, Rev D started in Feb, 2006, so your amps are also Rev. D, with the proper action on braking. So, that isn't the problem. All these amps DO have an asymmetry when on the current limit, due to variations in the .005 Ohm current sense resistors and the traces on the board. I select other components to try to compensate for that on each batch of boards. If you are at the current limit (the motors tend to sing in that mode) you either need to trun up the limit or back off the acceleration. Current limiting of the amplifiers breaks the closed-loop operation, and should be avoided. it is a safety feature for the amps and motors. I should add I am not sure you ARE hitting the current limit, and one way to check is to slightly increase the acceleration. If the error becomes vastly larger, then you are on the current limit. I will have to check the current limit. I know the PWM signal is not maxing out, but as the inductance of the motor keeps changing the DAC output is not a good indication of current. The DAC output looks trapezoidal, so I just assumed there are no problems during accel/decel, but you are right - the current limit might be setting in... The PWM would be expected to go to ~95% when the amp is in current limit, as it is no longer responding to the command. The DAC output is much more proportional to output voltage than current, and so is nearly an analogue of velocity. So, it doesn't sound like current limit. Other's comments about integral windup sound very relevant. I use I very cautiously, as it can have a number of side effects like this. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] PID-tuning
Anders Wallin wrote: somebody wrote : 3. What sample frequency are you using? with torque mode and straight PWM amps you will need a high sample rate than used with velocity mode servos. A a higher sample rate will allow higher P gains with these amps, improving performance. This is the standard 1 ms servo-thread. I could try decreasing that a bit. Will EMC2 report about real-time delays if I reduce the servo period too much? Not necessarily, but it does increase CPU load, of course. If you have an adequate computer, you can go to 5 KHz, for instance. This moves the 1/freq sampling jitter up to 2.5 KHz, well outside the band of motor response. At the default 1 KHz rate, the 1/freq produces a lot of noise at 500 Hz, which can lead to exciting some instabilities in the mechanical system. I'm not sure this is really any panacea, but it does seem to help with whistling servos. It certainly is easy to try, though. Save your parameters first, they will all change at a different frequency. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] VFD Interference
Erik Christiansen wrote: On Sun, Apr 20, 2008 at 10:42:31PM -0700, Kirk Wallace wrote: I got the spindle VFD working on my Shizuoka and the speed varies quite a bit. This seems to be due to interference coming from the spindle motor leads. The cable has no shielding but there are six wires. One for each phase, and three more that seem to connect together at the motor end, but not to anything else. That sounds an awful lot like the star point of your 3-phase motor. [1] While it's usually within a few volts (or ten) of ground, _if_ supply and load are reasonably balanced, it's not_for_finger_poking_or_connecting_to_ground, I submit. (If one phase goes open, there'll then be smoke, if not before. ;-) The star point would be near ground with balanced 3-phase Wye power. There are also corner-grounded and center-tap grounded Delta systems where it is nowhere NEAR ground. With a VFD, it will NEVER be near ground, and will guarantee a wild fireworks show if grounded. The VFD has a floating 400 V DC supply that is bouncing between the peaks and valleys of the input power waveform, and the motor's star point will be centered between those + and - voltages. I put a power line filter on the INPUT to my VFD, and have no shielding or other treatment on the wires between VFD and motor. I had some hash on my computer screen before I instaled that filter. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] VFD Interference
At the request of a potential customer, I did some performance tests with EMC2. I created a program with 1 blocks like N123456 G01 F30 X1. Y0. with the coordinates working around a 2 diameter circle. Each chord is roughly 0.0006 long. I ran it with the feedrate at 30 and 60 IPM, no difference, so it wasn't acceleration-related. I got 4 minutes and 17 seconds both times. That works out to 38.91 blocks/second or 2335 blocks/minute. This is on a 600 MHz Pentium III running my universal PWM controller at a servo update rate of 1 KHz. Presumably a hot 3.0 GHz CPU would do this at least 5 times faster. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
[Emc-users] block processing rate
Kenneth Lerman wrote: To the best of my knowledge, there has been no attempt to optimize the performance of the interpreter. If there is strong feeling that this might be a problem, I suspect that it could be improved significantly. My general experience with products that have never been optimized is that a factor of two is generally easy. A factor of five or ten isn't unusual. I don't know if this is a problem. If a factor of five can be obtained with just a modern CPU, that would be over 100 blocks a second. I think that would compare quite favorably with commercial controls. I wish the guy that called me had left his number, so I could call him back with the results. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
[Emc-users] block processing speed
John Kasunich wrote: Jon Elson wrote: At the request of a potential customer, I did some performance tests with EMC2. I created a program with 1 blocks like N123456 G01 F30 X1. Y0. with the coordinates working around a 2 diameter circle. Each chord is roughly 0.0006 long. I ran it with the feedrate at 30 and 60 IPM, no difference, so it wasn't acceleration-related. I got 4 minutes and 17 seconds both times. That works out to 38.91 blocks/second or 2335 blocks/minute. This is on a 600 MHz Pentium III running my universal PWM controller at a servo update rate of 1 KHz. Presumably a hot 3.0 GHz CPU would do this at least 5 times faster. Jon Your test is not measuring interpter speed, or anything else that is likely to be improved by a faster PC. I am almost certain you are seeing acceleration limited behavior, but without knowing the accel limits of your config I can't do the math to be sure. I don't think so. I am sure I had it in G64, and the 10,000 vectors describe one orbit around a 2 diameter circle, so the individual vectors are very nearly colinear. snip What about when blending is turned on? EMC2 has one-segment lookahead. It looks at the next move, and blends the two together if possible. The blending is fundamentally quite simple (although dealing with all the possible cases is not.) EMC2 calculates the accel, cruise (if any) and decel numbers for each move as if it was in exact stop mode. But then it starts the accel phase of the next move as soon as the decel phase of the current move begins, and adds them together. Move 1 is slowing down, move 2 is speeding up, and if the two are at a mild angle to each other they ramps cancel out leaving a very smooth movement. If the two moves represent a sharp turn, the overlap between their accel and decel periods tends to round off the corner. EMC2 has special code to detect that this would happen, and if the rounding is worse than the specified tolerance EMC2 slows it down some more. Since EMC2 calculates each segment as if it were doing exact stop, blending doesn't increase the maximum speed. It does increase the average speed, because the tool isn't constantly slowing down and speeding up. And it makes the movement _much_ smoother. In the case above, where the accel limit is 50, blending would result in an average speed of 14.7 inches per minute. Jon's program made a 2.00 diameter circle, with a circumference of 6.28, and it took 4 minutes 17 seconds. That works out to an average speed of 1.46 ipm. I worked the equations backwards and got an accel value of almost exactly 1.0 inches per second squared. Jon: Is the accel limit on your test config 1.0 inches per second squared? No, both default and max accel in the TRAJ section is 5.0 user units (Inches) and max_accel in the axis sections is also 5.0 I have posted the program to http://pico-systems.com/codes/contour.ngc just so we can all have a standard to work with. But, I think you have defined the problem quite clearly, and I did not know it worked this way. The blending has a horizon of ONLY one line! Looking at the first 17 blocks : N10 G01 F30 X0. Y1. N40 G01 F30 X0.0006 Y1. N70 G01 F30 X0.0013 Y1. N100 G01 F30 X0.0019 Y1. N130 G01 F30 X0.0025 Y1. N160 G01 F30 X0.0031 Y1. N190 G01 F30 X0.0038 Y1. N220 G01 F30 X0.0044 Y1. N250 G01 F30 X0.0050 Y1. N280 G01 F30 X0.0057 Y1. N310 G01 F30 X0.0063 Y1. N340 G01 F30 X0.0069 Y1. N370 G01 F30 X0.0075 Y1. N400 G01 F30 X0.0082 Y1. N430 G01 F30 X0.0088 Y1. N460 G01 F30 X0.0094 Y1. N490 G01 F30 X0.0101 Y0. we see that the first sixteen vectors are perfectly co-linear, due to roundoff! There is no need to decelerate at all here. And the next 12 are again colinear, and so on for some time. I have no idea how hard it is to do better with this lookahead, I expect it is not easy to handle all possible combinations of moves, with compensations being turned on/off, etc. But, for people doing contouring of surfaces, they often want to skim rapidly across surfaces at 60 IPM with several thousand vectors/inch, therefore, at least 1000 vectors/second is needed. This doesn't require large acceleration, as the vectors comprise a smooth curve. If there is a discontinuity, then the machine has to slow down, of course. Any suggestions are welcome. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] block processing rate
sam sokolik wrote: I made this program http://pastebin.ca/993663 (same thing I think as Jon made) I ran it on the live cd (hardy is all I have handy at the moment) on a dual core 2.2ghz. (not the greatest latency - around 20us. Anyway.. With the default stepper_inch.ini - acceleration set to 20In/s/s the program takes 1min40sec. (side not - the velocity on the axis screen reads 0) I upped the traj and axis acceleration (x and y) to 100In/s/s and the program took 1min40sec.. :) as much as I could tell with a second hand on the clock. So that calculates out to about 3.77ipm. Now - If I add G64P.0005 - the program and leave the acceleration to 20In/s/s it cruises along at 30ipm.. :) cool. Takes about 13 seconds. Yes, that does it! WOW! I left my accel at 5.0 In/sec^2, and got 17 seconds. I can actually hear it accelerating at the beginning of interpolating the circle and slowing down at the end. So, there is a HUGE difference between G64 and G64 P0.0005 ! So, that is 588 blocks a second! Thanks, Sam! Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] block processing speed
John Kasunich wrote: Jon Elson wrote: I have no idea how hard it is to do better with this lookahead, It's hard. :-( That's the rub - if there is a discontinuity the machine has to slow down. But it doesn't know there is a discontinuity until it gets there. (Or in EMC's case, until it gets within one segment of there.) Well, then, what is the difference between G64 and G64 P0.0005 ? In this particular program, it makes a 4 minute difference, or a factor of 15:1 ! it still didn't get up to the programmed feed rate, but it got a lot closer. I think I must have left the file with the 60 IPM feedrate in it, and it did 6.28 inches in 17 seconds, or 22 IPM. I can understand horrible performance in G61 mode, that would be expected. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] block processing speed
John Kasunich wrote: Jon Elson wrote: John Kasunich wrote: Jon Elson wrote: I have no idea how hard it is to do better with this lookahead, It's hard. :-( That's the rub - if there is a discontinuity the machine has to slow down. But it doesn't know there is a discontinuity until it gets there. (Or in EMC's case, until it gets within one segment of there.) Well, then, what is the difference between G64 and G64 P0.0005 ? In this particular program, it makes a 4 minute difference, or a factor of 15:1 ! it still didn't get up to the programmed feed rate, but it got a lot closer. I think I must have left the file with the 60 IPM feedrate in it, and it did 6.28 inches in 17 seconds, or 22 IPM. I can understand horrible performance in G61 mode, that would be expected. Maybe the lines in this program are close enough to collinear that when you give G64 a tolerance to work with Jeff's code is combining bunches of them into much longer segments that allow much higher velocity. Well, that is the idea. With it taking 10,000 chords to make a full circle, they should be VERY close to co-linear. I think these chords are all within .0001 of being colinear. I think G64 without a tolerance can only blend moves that are exactly tangent - that would mostly apply to paths made up of lines joined by arcs. Well, at least near the corners there were a lot of segments that were truly co-linear due to roundoff, but it didn't seem to speed up at those spots in G64. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] block processing speed
Anders Wallin wrote:
I think G64 without a tolerance can only blend moves that are exactly
tangent - that would mostly apply to paths made up of lines joined by arcs.
I thought G61 ('exact path' mode) did this already?
G61.1 is 'exact stop' mode which will stop at the beginning and end of
each move.
The Pxxx for G64 was added quite recently. Before that, with low
accelerations it was quite possible to get EMC2 to misbehave and cut
corners a lot by specifying only G64 and no tolerance. Jon's test shows
that now G64 without a tolerance can produce a path which is
unnecessarily close to the programmed path and thus slow.
IMO G64 without a tolerance could/should be made illegal, and only a
G64Pxxx allowed by the interpreter.
Well, I'm not sure that is necessary. I would like to see a
good description of what all these options are. I used to know
what G61 vs. G64 did, but now it gets more complicated. It
looks like we have 4 options now, G61, G61.1, G64 and G64 Pxx.
The documentation on this in the user_manual.pdf is really
lacking. I'd be glad to try to complete it, but I need to know
how it actually works, first.
Jon
-
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Very low PRM system.
Roland Jollivet wrote: Hi I (we?) have no idea of the application, but if the system is always going to operate at low speeds, what about a fat flywheel on the drive motor to smooth things out? You'd need an incredible flywheel to smooth out a motor movong at 1.8 degrees per minute! I don't think even a railway wheel would do it. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] limit switch error
John Thornton wrote: On 25 Apr 2008 at 11:16, Jon Elson wrote: With my controllers, for instance, there are 3-line stanzas for each limit switch, such as : newsig Xminlim bit linksp Xminlim = ppmc.0.din.01.in linksp Xminlim = axis.0.neg-lim-sw-in Jon Jon the new format is a single line using net. I don't have my emc machine handy but it's something like this net my-signal axis.0.neg-sim-sw-in = paraport.0.pin-13-in O! I'm treading water trying to keep up! Glub! Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] CNC Fest
[EMAIL PROTECTED] wrote: Hi I want to ask about next CNC festival. Will be any class, as a class where is instructor that taking question and addressing questions? 2 years ago Ray Henry was instructor. Last year Jon Elson teach about tune up. I will arrange to do two seminars. One on hardware setup for my board products, one on servo tuning. I plan to be there the whole week, I think. Probably come in on Monday afternoon. I am sure good class organization will make more incentive to more people to come to CNC work shop. I am wrong? No, I think the class scheduling was not well organized last year, especially posting of what classes were when and where. We ought to put it on the web. Maybe we could add something to our linuxcnc wiki page so it could be updated easily. Maybe we could hang a whiteboard somewhere and post today's and tomorrow's classes each day. I have an old whiteboard here I could bring. I have topic for class, program MPG-4 pendant and make it fully functional. I think many will be interesting to see how it can be done. If any preoperational work can be done, please let me know. Yeah, I'd like that, too. I've been thinking of a way to add a pendant to my PPMC systems for a long time, but just haven't done anything about it. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] CNC Masters Supra
Peet de Vos wrote: Hi there, Does anybody use or know whether EMC will work (without headaches or major customization) with the following *CNC Masters* Kit: _CNC Supra Vertical Knee Mill Retrofit Kit___ http://www.cncmasters.com/CNC%2520Supra%2520Retrofit%2520Kit.htm They are not real clear what motor drivers they are using. The software they supply is clearly Mach3. If they are using Gecko drivers, it should be quite easy to trash the Mach and use EMC. But, whay pay so much for something, when you don't want it in the first place? See if you can just buy the mechanical part, or that plus the drive box. Finally, at that price, you could have a servo system! You might want to think about that. Some servo systems allow you to use the machine manually, and switch between CNC and manual with DRO without losing the axis alignment between the two modes. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] EMC Fest 2008
Dale Ertley wrote: Hello, Will there be any good training for newbies at the EMC Fest 2008? I am new to EMC. I am also in the process of putting a 4+ axis (xyzw+) full size mill on EMC2. I'm not sure exactly what formal classes there will be. I hope we can have something more organized than last year. Ray Henry will be there, so that is likely to be a guarantee! But, even without formal classes, we will have a true EMC critical mass there as usual. Just buttonhole one of the developers and ask a question. I'm planning on doing two classes that are oriented a bit toward my products. But, maybe I need to do a general class as well, mostly focusing on the retrofit process and selection of components. If you don't have EMC running yet, bring your computer and we will get it installed and at least partially configured for your own setup. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] helical code
Dave Engvall wrote: Hi, I'm milling a tapered hole using helical interp. The program works fine but can't keep up. With a HSS ball mill I've been millling at 4 ips. With carbide I could raise that to 10 ips. At either speed it maintains speed 'til about 1/2 way thru the program then starts to pause and go, pause and go., etc. I tried G64 P0.001 and that didn't help. Hmm, the G64 P0.0005 made a HUGE difference in the performance on a flat XY-plane circle specified in tiny G01 moves. It sped up by 15 : 1. My test machine is kind of torn apart right now, but if you'd send me the whole program I could try it here. Of course, for tapered holes, a tapered end mill is another solution. The slowing down at about halfway through is a real puzzle. It seems odd that the user-mode half of EMC couldn't keep up with the real-time part. The program is only about 500 blocks long? I'm wondering So, you start at the top, with a larger diameter, and then get smaller as you go down? I'm wondering if the curvature is just on the bitter edge of triggering some limit for merging the line segments, and as the diameter shrinks, the trajectory planner switches modes from smooth merging of the segments to not merging them without deceleration. I'd guess that P0.001 parameter might have an effect on this, you might have to make the P larger or make your line segments shorter. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Master-Slave stepper set up
Mark Wendt (Contractor) wrote: I've in the process of building a two axis machine - using the X and Z axis, and I have a three axis stepper/driver combo already set up and basically ready to go. I was wondering if there is some way to configure EMC so that I could slave the Y axis to the X axis. and use two steppers, one on each side of the gantry. There are much better ways to do this. Depending on the motion hardware you have, this slaving can be done in the HAL layer, and once it is set up right, you never need to know there are two motors moving in unison again. The tricky bit is to figure out how to get the two sides of the gantry synchronized so the machine is square and not distorted. The large gantry mills have bearings in the gantry towers so that when the gantry sides are out of sync a little it doesn't stress or break the machine's structure. Homing the axis brings both sides to preset alignment with individual sensors. I'm sure there are ways to do this easily in HAL with a couple AND gates that shuts off steps to the motor that reaches home first, then waits until the second motor gets there before completing the home sequence. This gantry scheme is called a tandem axis. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] EMC Fest 2008
Dave Engvall wrote: You get to take your chances. Several of the attendees stayed at the Econo Inn. I stayed there two years ago. They had just changed the beds and the new ones were too firm for me. I will stay at the Holiday Inn Express which is more expensive and farther away; but then my wife will be there also. The Econo Inn is about as close as any motel I'm aware of. Price two years ago was $50/nite. All the above is true. But, you can also bring your camper/RV and stay AT the Workshop! There is another low-budget motel there that I stayed in a couple years ago. No way would my family stay at such a dive! Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] 155 ms real-time delay every 10 minutes?
Kent A. Reed wrote: Anders: I hesitate to throw what may be just another red herring into the discussion, but there are a number of 600-second defaults floating around Linux, and it is possible you're getting tripped up by one of them rather than by some clock problem. For example, by default, the stock Ubuntu 7.10 on my home PC clears the ip4 routing table every 600 seconds (defined in /proc/sys/net/ip4/secret_interval). I would assume the time it takes to do this is dependent on the size of the routing table (this created a problem with a Redhat system we built at work to monitor experiments). Anything happening in IP routing should be totally unable to affect the RT system. Maybe it does something else at the same time that triggers something in the ethernet driver, like a big DMA burst. I also used to trip over the various ways screen blanking intervals are defined in X11, though that seems an unlikely cause of your problem. Some of the computers with on-board video handle certain things in BIOS ROM routines, and these can cause upsets to the RT environment on some machines. Sometimes you can turn these things off, or set them to be done by Linux-level software rather than hidden features, by turning down the X-windows acceleration support in the configuration menu. I can easily see how a screensaver could trigger some bit-blt activity that might have side effects. I think this one is well worth pursuing. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Master-Slave stepper set up
Mark Wendt (Contractor) wrote: Thanks for the info. I was doing a little reading on the wiki and in the user doc over the weekend, and missed seeing anything on that. Is there any documentation that I may have glossed over? It likely is NOT documented at the present. I think someone is working on a standard configs file set for this machine config. We have batted this around a bit, I have a fairly clear idea how I would do it if I were building such a machine, but I haven't actually done it. I think all the HAL components you would need already exist. Most likely, just one AND gate for each motor, that passes the step pulses until that motor's home switch trips. Then one and gate to send EMC the home switch signal ONLY when both home switches are tripped. That signal also sets a flip-flop to pass all further step pulses by bypassing the and gates, using a 2-1 multiplexer. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Master-Slave stepper set up
Kirk Wallace wrote: I wonder. If you don't know the state of the two end carriages, and they might be in a binding condition, it might be necessary to sync the carriages before any substantial move. Could a laser be used to get the carriages synced enough to for a move to a proper sync? Well, in a step-controlled system, they will always move together, so the binding/misalignment would not get any worse than when the step drivers were turned on. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] 155 ms real-time delay every 10 minutes?
Kenneth Lerman wrote:
The suggestion was made that arp or some other ethernet related protocol
might run every ten seconds. Disabling ethernet in the bios might be
*causing* the problem if all of the protocols are not disabled in software.
I sure don't know what would happen if something is trying to use the
ethernet and there was no hardware there.
(That's probably not the problem, but it can't hurt to suggest looking at
it.) :-{
I have a Dell desktop box here that is in a location where I
share ethernet jacks. I know that you just about can't log
in and get X running without the ethernet being connected.
After 10+ minutes, it finally gets running and appears OK.
(At one time I knew how to disable this, but now I've forgotten.
Certainly shutting down one of the network daemons. But, before
it gives up, the system is insanely slow, it literally takes
10+minutes for the login stuff for X to complete.
I don't think this affects the RT system, though.
Jon
-
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Buying used CNC machine - what to look for??
Witek GB wrote: I am thinking of taking the plunge and buying a VMC maybe with an ATC. I plan on buying it for the iron and upgrading it to EMC. I have manual machine tools already so I know what to look for when I see a manual machine, but with CNC I really do not know much. What should I look for? Are there any resources on the Internet that show you what to look for when buying a used cnc? How can I determine if the ball screws are good or bad? Servos?... I will use this machine for hobby work and maybe some light production runs. Any help is appreciated Servo motors are pretty tough. Really old encoders had light bulbs in them. They can be retrofitted with infrared LEDs. Many older machines had resolvers instead of encoders. (I'm working on a lower-cost retrofit for those, but it may still be cheaper to install US Digital encoders where they will fit.) Ball screws are fairly hard to evaluate in the field. You can put a dial indicator on the table and see how far you can turn the screw by hand to check backlash. The general condition of the machine can be a useful guide. If every nook and cranny is stuffed with chips, and layers of brown sludge coats everything, you know the machine was run hard in heavy production. That will take its toll. As for the ATC, the simpler it is, the easier it will be to get EMC to control it. If it is a massively complicated hydraulic monster with several changer arms, lots of position sensors, etc. then it will be more complicated. Ray Henry did manage to get the Mazak working at Roland's shop. That is a pretty complex ATC. Remember that the tooling is a major expense on these machines, a rack of NMTB 40 holders and collets is worth a LOT more than most older machines. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Buying used CNC machine - what to look for??
Dave Engvall wrote: I didn't (don't) know any better so I consider the Mazak ATC as pretty standard in complexity. Some of those on the horizontals must be simpler since they change tools in 2 sec or so. The Mazak Micro 5 has a part that swings the holder 90 degrees from the plane of the tool wheel to the plane of the spindle, then the changer arm swaps them, then that part swings the tool back up to the wheel. Many smaller changers have the tool carousel in the same plane as the spindle so there is just the one arm that does a down-spin-up motion. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Buying used CNC machine - what to look for??
Witek GB wrote: Dave, I live in Lenexa Kansas a suburb of Kansas City. I hopped to go to the EMC CNC workshop in IL however around that time I have a prior engagement with my wife. If I would go she would kill me and the machining hobby would be out the window. I might go next year. Is there anyone else in KC using EMC? Stuart Stevenson has it runing on several large machines in Wichita, which is not too far. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Buying used CNC machine - what to look for??
John Thornton wrote: Just goes to show you never know what is tucked away in the hills of south east Missouri... If I had to guess I would have guessed a job shop to support the lumber industry. I'm glad it is bigger than that. There's a LOT of aerospace work in MO. The old McDonnell-Douglas plant, now owned by Boeing, in St. Louis, also Sabreliner has a big overhaul facility here. McDonnell sold off its major fabrication plant to GKN, which does work for just about every aerospace outfit. Boeing has had a major facility in Wichita for nearly forever, as well as Cessna, and there are subcontractors for several hundred miles around. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] CNC Fest classes
Ray Henry wrote: I won't presume to speak for other presenters. I move a around a lot for an old guy. Don't think that a web cam is going to do folk much good although I'm not at all opposed capturing my events. A good lapel mike and a camera focised on your whiteboard or easel would get all the important stuff. Another issue will be my use of the projector -- a lot. I presume that we could hire someone with a bit of video production experience to switch between the display and the lab and catch the essence of things. If we use a stand alone video camera we could probably get one of the developers to write us a capture and store program. It may be possible to capture the projector info digitally, and mix later. Wireless mikes would be good also. I don't have a clue how Roland's link would stand up to streaming. We might have to upload after the fact. Yeah, I'm sure that would overload his link. Live streaming video from the workshop would be WAY cool, though! My wife's favorite church 1000 miles from home does this, but they hire a service to do the internet broadcast. Still, they have to have a fast upload feed to that service. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] EMC Fest 2008
Ed wrote: Which brings up the question of power availability. I would like to bring a project that would need 240V 3Phase at about 10Amps. Does anyone know if that is available? Hard wiring or common plugs are not a problem. Ed. Roland most certainly has good 3-phase power at his location. He has both 480 and 240 V panels. I don't know how easy it is to get additional power rigged, but there are a number of twist-lock outlets scattered about. You have to be real careful to get the right voltage before you plug in. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] EMC Fest 2008
John Kasunich wrote: The Mazak is three-phase, and plugs into a twist-lock receptacle, but I don't recall if it is 240V or 480V. Depending on how continuously you will need power we could unplug the Mazak and plug in your machine. I seem to recall that the plug is fed from a disconnect, which is handy for safety. My recollection is the Mazak is 480. I recall there is a 480 panel and a 240 panel. I think Roland has a dry transformer inside the building to get the 240. The 240 may not be available from the bus duct, only 480. I'm certain that there is 480V available - can you scrounge a suitable 480 to 240 transformer? I think you need about 5KVA. A 5 KVA single-phase transformer isn't all that big, but in 3 phase they start to get pretty heavy to lug around. I have a 2 KVA single-phase step-down transformer. Would that be needed? I can bring it along. Jon - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Thread Milling
Kirk Wallace wrote: Is tapping obsolete? http://www.emuge.com/carbide_thread/ Others have given good answers, but there is an alternative: SINGLE-row thread mills. These have the advantage that they will mill a range of thread pitches with one cutter, and they are cheaper than multi-row thread mills. They are also slower as you have to mill the entire length of the thread helix, while a multi-row thread mill can do a length of thread in one orbit of the hole. For a small shop, and in cases where you need to mill non-standard thread diameters or pitches, these are great. I have a 1/4 single-row thread mill from Micro-100, and have used it on some odd jobs. One was a 1 20 TPI thread that needed to be milled at an angle to the rest of the part. I had no idea how to fixture it on the lathe so I mounted it on a sine bar in a milling vise and drilled and thread milled it with excellent results. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Subject: Re: Micro stepper driver MSD542 / KL-5042
Ian W. Wright wrote: 'Scuse me for stepping in but I'm a bit confused by a couple of things being said in this thread. 1. Exactly why does a low accel round corners more than a fast accel.? The program calculates all moves based on the specified machine acceleration in the ini file, and the velocities specified in the G-code. For each move, a point at the end of the move is chosen where deceleration would need to begin to stop at the exact endpoint specified. On machines with higher acceleration limits, that point will be closer to the endpoint than on machines with less accel. available. This point near the end of the move is where the acceleration to the next move in the G-code is blended in. So, at a right-angle corner, with G64 in force, the X axis will actually start moving before the Y axis comes to a complete stop. That will, of course, round your corners. Does this suggest that a machine with slightly underpowered drive motors is inherently less precise than one with big drive motors? Yes. 2. Is there some advantage in using G64 with a tolerance over G61 if you want tight adherance to a toolpath? Yes, it allows the program to run MUCH faster,while maintaining the required accuracy. I ran some tests a few weeks ago between G64 (no parameter) and G64 P0.0005 (in inches) and the difference was four minutes vs. 17 seconds! This was a 10,000 segment circle 2 in diameter. G61 would certainly be as bad as the G64 with no tolerance, if not even worse. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Thread Milling
Mark Wendt (Contractor) wrote: I think I have a grasp of how this thing works now. My only question left is, if the teeth are in rings, and there are multiple rings (and assuming the thread mill is spinning at that same time as the x and y axes are moving), how does the tool keep from munging the threads that are already cut if there is no screw ramp to them? From the looks of the thread mill in the picture, the thread mill's rings all seem to be parallel with each other, and have no screw ramp. For inside threads, the thread mill has to be a good deal smaller than the hole diameter, so that the curvature of the thread mill gets the cutting teeth out of the thread faster than the helix angle of the thread it is cutting. If this works out, then the tips of the thread mill never touch the sides of the thread except at the one axial line on the hole wall. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Thread Milling
Glenn R. Edwards wrote: Other issues with thread milling: 1)The thread mill tool will not follow the pre-drilled hole as will a tap. 2)Getting the correct pitch diameter from a thread mill is a trial and error routine. I don't think so. I bought a single-row 1/4 thread mill from Micro-100. I wrote a program to create the G-code, and after trying one pass in air, I drilled a hole and let it run. It produced a thread as close as I could tell identical to a tap. I didn't have a real certified go/no-go gauge, but a threaded part felt the same in the hole. I was very impressed. I had bought the thing because I had a job coming up that needed an NSEF thread, and I was worried about getting a tap for that. Then a sale flier on special taps crossed my desk, and I was able to get the right tap at a very reasonable price. So, I've only done one job where I needed to thread an off-size hole at a wierd angle. But, it produces a thread with very predictable pitch diameter. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Thread Milling
Glenn R. Edwards wrote: Jon, I hope it doesn't appear as though I am jostling you for the final word, however you pointed out a major difference between us hobbyists and professionals. A professional must deliver a part in spec 100% of the time ( six sigma- don't get me started...) no matter if the part was cut in the cold morning or during a hot afternoon, if the part was mounted at the center of the table or if mounted near the edge, if the machine was being run by your best operator or the guy you hired yesterday. A tap can wear, but so can a thread mill cutter and probably faster. However, thread milling introduces two additional tolerances that tapping does not. Each of those tolerances can (and eventually will) produce an unacceptable thread. A loose fitting thread can have a LOT less pull-out strength than a properly formed thread. And, you might not be aware of a loose thread unless you have a GO NO-GO gage. In my day job, I want tapped holes from the machine shop (In fact, I am getting quite accustomed to roll-tapping and the superior strength thread it produces). During the weekend, I will enjoy making my own threaded holes with a thread mill. Yup, I have to say I do not do aerospace work, or much where a mechanical failure would endanger life or limb. And, I am not a job shop where all my parts wind up passing under the scrutiny of my customer's incoming parts inspector. On the other hand, I buy a lot of stuff that is SUPPOSED to be tested, and I have a little pill bottle full of screws with no threads or no recess in the head, and such like. My guess is that thread milling with a worn cutter will result in an under-size internal thread or an oversize external one, in either case the parts can't be assembled, rather than a too-loose thread. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Micro stepper driver MSD542 / KL-5042 cont.
Kenneth Lerman wrote: I really think we need some pictures and/or details. Instead of using letters for a test case, could you try using short lines? First plot a set of parallel lines about 5 cm long and 2 cm apart, parallel to the X axis and centered on the Y axis. Then do the same thing parallel to the Y axis and centered on the X axis. Another really useful test is to make a circle out of small linear segments. If there is some error happening on axis reversal, it will show up at the points where the movement reverses. Many step/direction drives have restrictions on when the direction signal can change relative to the last (or next) step pulse. If the required polarity of the step pulse gets reversed, it can make the cnc computer violate that requirement, and lose a step every reversal. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Micro stepper driver MSD542 / KL-5042 cont.
John Kasunich wrote: It sounds like you are losing position during rapid moves only. The move between letters is probably a G0 rapid, which goes as fast as EMC thinks the machine can go, while the moves within the letters are G1 lines or G2/G3 arcs, both at some specified feedrate. A quick test is to use a text editor to change all ocurrences G0 or G00 to G01, and re-run the test. If the results are different, this strongly confirms John's thinking above. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Way OT: Contact Lube
Kirk Wallace wrote: I have a Hobart TIG with really stiff AC/DC+/DC- and AMP range selectors. They are eight inch rotary switches with 1/8th inch thick blades and wipers. Does anyone have a favorite lube or should I run them dry? I tried an electronics cleaner/lube and it turned into a stiff wax. (EMC tie in: I can't work on the Shizuoka until I get it fixed.) Very likely it is the SHAFT and bushing that are stiff, not the contacts themselves. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Harmonic gear and EMC2. ????
John Kasunich wrote: Kenneth Lerman wrote: A simple solution is to use a dual drive. Use a course mechanism with long travel for course positioning and a fine mechanism with limited travel for fine positioning. The fine mechanism has a large mechanical advantage and can use a smaller motor and driver. Of course, some sort of clutch mechanism could be used to allow switching between the course and fine speeds. You might be able to avoid the clutch as well. One possibility would be to stack a very high resolution slow stage on top of a fast coarse one. With the right speed reducer, I think these extra complications are not needed. The problem is that asking a motor to give smooth motion at a speed of a few degrees per MINUTE is just too slow. But, a servo motor can also run at pretty high speed. So, a modest speed reduction between motor/encoder and leadscrew should satisfy both the slow and fast requirements. The speed reducer must be stiff and backlash-free, which requires a higher class of unit. Either a worm drive or planetary would work, as long as it was designed from the ground up for zero backlash. A problem with a worm drive is it has sliding friction, and will wear. There is a new style of zero-backlash planetary drives that use slightly offset pins on the planetary carrier, causing it to spring-load the gears. I'm guessing this is a patented technique, and so there's only one supplier for a few years. You could also do things like having the fast motor turn the screw, and the slow motor turn the nut with a worm gear to increase the resolution and decrease the speed. In this latter case, the stiffness of the fast motor will be an issue, even though it isn't turning. You haven't answered the question of what type of mechanism you are planning to use to provide smooth linear and rotary motions at this slow speed. I think Ken has hit the nail on the head here. At very low speeds, it is very hard to get smooth motion. You are more likely get stick-slip behavior, where the motor turns a little but the table doesn't move. The screw and other parts deflect until they build up enough force to overcome the static friction and start it moving. As soon as it moves, the dynamic friction is much lower than the static friction, and it moves farther than you wanted it to, then stops. And the whole cycle repeats... The individual movements are tiny, but at the extremely slow speeds you are talking about, stick-slip is more likely to be the limiting factor than your encoder resolution. Yes, you need all rolling elements, leadscrew, slide, etc. to combat this. Gearing down does not help - the issue is the flexibility and friction of even the most rigid screw/nut/bearing combination. At these scales, you almost have to think of the metal parts as if they were hard rubber. Sliding ways are usually the worst choice for stick-slip. Rolling element Linear bearings are better, but they are not very happy in a grinding environment with abrasive dust. Depending on the loads, you might want to consider air bearings. They do not suffer from stick-slip, and to some degree are self-cleaning in a dusty environment. One of the regulars on IRC, who uses the name toastydeath, works (or worked) for a company that makes air bearings, and may have some suggestions. Air bearings are insanely easy to make, although getting just the right geometry to prevent oscillations is slightly tougher. Really, you can make surprisingly good air bearings with a drill press, a block of aluminum and a surface plate. You drill a hole in the block, then make a slight relief in the center of the block, and lap the rest of the block flat with fine sandpaper and the surface plate. Apply air pressure to the hole, and voila- an air bearing slider! Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Lathe screw cutting problem
[EMAIL PROTECTED] wrote: Hello. Has anyone managed to set a screw cutting lathe up successfully with EMC2? Mine is a simple setup. The encoder has 1ppr and also a disk with 48 slots. The problem is that the lathe carriage does not synchronise with the spindle nicely - it overshoots and undershoots and by the time it settles down the thread is half cut. This happens at any speed above about 100rpm. Other movements are beautifully smooth. I have set up lathe-type threading on my mini-mill, with a real quadrature encoder on the spindle. It has 1728 lines, for 6912 quadrature counts/revolution. So, at 100 RPM, it provides 11,520 counts/second, or 11.52 counts/servo period, assuming a 1 KHz servo update cycle. I didn't think I needed that kind of resolution or count rate, it was just a one-of-a-kind encoder I had laying around. I did also see some of this hunting behavior when I first started setting this up several years ago. I just dug through the EMC-users and developer's lists, but I can't find the message. I had problems in my encoder counter card that occasionally failed to perform the find index pulse function correctly, and that may have been the reason for the axis hunting. Also, if your Z axis is asked to sync to the spindle in too short a distance, it will reach its accel limit and suffer some hunting. A fine-pitch thread will be much easier to sync to than a coarse one. What is your accel limit set to in the ini file - both in the [TRAJ] section and the [AXIS_2] section (for the Z axis)? Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Lathe screw cutting problem
Kirk Wallace wrote: On Sat, 2008-05-24 at 13:49 -0400, [EMAIL PROTECTED] wrote: Hello. Has anyone managed to set a screw cutting lathe up successfully with EMC2? Mine is a simple setup. The encoder has 1ppr and also a disk with 48 slots. The problem is that the lathe carriage does not synchronise with the spindle nicely - it overshoots and undershoots and by the time it settles down the thread is half cut. This happens at any speed above about 100rpm. Other movements are beautifully smooth. Right. The demo threading G-code sample does this spindle sync in air, and then dives the cutter in in X to start the cut, and then does a diagonal retract at the end of the cut, before dropping out of sync. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] 20V80 on eBay
Kirk Wallace wrote: This one could use a little EMC: http://cgi.ebay.com/Cincinnati-20V80-5-axis-mill_W0QQitemZ110255408981 Must be picked up within 14 days I doubt you could arrange the riggers and freight in that time. H, too big even for my garage! Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] servo with brake issue
Kirk Wallace wrote: Thank you for posting your results. I kind of take the mains for granted. Oh,no! That can be a FATAL mistake! I have been BLITZED a number of times by defective grounds, and systems where multiple pieces of equipment were plugged into different outlets. I had a ground in a house I used to own where a romex staple had pierced the safety ground and tied it to hot. The staple was driven with the power off, and it cut the safety ground back to the breaker panel, so the hot and ground were connected together with no short, and just left that way. It took me WAY too long to find that one! Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Rigid tapping
Anders Wallin wrote: Hi all, Just wanted to let everyone know that we have now mounted an encoder on the spindle motor of our mill and with a little HAL-tweaking it was easy to get rigid tapping to work. There are two videos of M3 tapping and M6 tapping on youtube: http://www.anderswallin.net/2008/06/rigid-tapping/ I'm not sure how many EMCers are doing rigid tapping out there but at least over here the spirits are high after the first successful tests! Great! I am going to set up my minimill to do some thread drilling in the next day or so. I use these $8 combined drill-taps. I have been doing it on the Bridgeport with a Procunier tapping head, but that is such a long assembly it gets a bit wobbly. Rapidly reversing the big Bridgeport motor is going to be hard on things, and getting an encoder on the spindle looks difficult. I have this little minimill, and have to come up with the right fixturing for the workpiece, but its motor should be able to reverse a lot more easily, and I already have an encoder on it. Jon - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Putting Encoder on AC servomotor???
[EMAIL PROTECTED] wrote: Hi I am looking for more info about why I can not take out encoder from AC servomotor and put another in? If the motor has commutation sensors (Hall signals) built into the motor, and separate from the encoder, then there is no problem. Many newer motors, like the SEM and Servo Dynamics, have integrated these functions, and so the commutation encoder needs to be accuratly aligned with the motor position. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Request for information at the CNC Workshop
Matt Shaver wrote: I put this on the wiki page as well, but I thought I'd make this request on-list as well. I'm looking for the following things: 1. Copies of ANSI, CENELEC, FM, IEC, ISO, OSHA, NFPA electrical and/or machine tool safety standards, European [Economic] Community Directives (EC or EEC), European Norms (EN), or any other standards, draft or published, of any age or origin relating to electrical controls and machine tools. The goal is to learn as much as possible about designing standards compliant machines and electrical controls. I don't think there is truly ONE way to do things. You pretty much have to use safety-rated controls from Pilz, Crouzet, Schaffner, Faulhaber, etc., and then set them up so they can function correctly to stop motion, even after a component failure. 2. Equipment or instruments to measure RFI, or practical information on the subject of accurately AND inexpensively measuring RFI emissions (both conducted and radiated) from electronic equipment with the goal of detecting (and ultimately reducing) RFI generated by machine tool controls, including VFDs and axis motor drives. I have been through the process of testing a product to FCC part B standards some years ago. It is practically impossible to do this yourself, due to the vast sea of electronic emitters spewing signals. Some testing labs have anechoic chambers, others use open-air test facilities in really out of the way locations where there are only a few RF sources, which they calibrate out. The equipment is QUITE expensive, mostly special-purpose spectrum analyzers. You can get off-lease analyzers from the equipment rental outfits, but they are still pretty expensive. Then, you get into calibrated wideband antennas, and on and on. Finally, you have to test every different configuration, so you generally have to test each specific installation, if you are doing one-off machine retrofits. Most machine tool controls have had totally uncontrolled emissions, because they were mostly exempt as heavy industrial equipment. Note the lack of output filters on most PWM servo amps. I need to make my CNC designs acceptable in real industrial environments, not just in the US, but throughout the world. Also I've become rather committed to making my designs more resistant to the effects of external electrical interference and power line disturbances, while simultaneously minimizing the level of EMI and RFI my controls leak into the environment. Of particular interest to me are: UL508A (Industrial Control Panels) UL50 (Enclosures) IEC60947 (Low Voltage Switch Gear) EN418 (Emergency Stop Equipment) EC Machinery, Low Voltage and EMC Directives The first step is to fully examine the standards that cover machine tools and controls so that I can produce compliant designs. If you have any of this type of information, and are coming to the CNC Workshop, I'd love to see it! Devices to inject disturbances into the line and equipment chassis are not as bad as the spectum analyzer type gear, and you can even make much of this yourself, like with auto ignition coils and pulser circuits. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Putting Encoder on AC servomotor???
Stephen Wille Padnos wrote: Jon Elson wrote: Jeff Epler wrote: In emc2 there are practical limits on the precision of position values due to the use of float as the fundamental type for analog values such as encoder positions. For instance, the values 1.0 and 1.0 + 0.5**23 (approximately 1.001) are both exactly representable, but no values in between those two can be represented. Similarly, 2.0 and 2.0 + 0.5**22 (approximately 2.002) are both exactly representable, but no values in between those two can be represented. http://en.wikipedia.org/wiki/Floating_Point#Implementation_in_actual_computers:_IEEE_floating-point A very good point! It seems there would be a 25.4 X benefit then by using mm as the user unit instead of inches, in the case of these very high resolution encoders. (I think you mean + 0.5 * 10 ^ -23 in the above) Nope, it's 0.5 ^ 23. 0.5 is the first digit past the binary point, and 0.5^23 is the last bit past the binary point. It seems to me that : the value 1.0 + 0.5**23 (approximately 1.001) would be 5, followed by 22 zeroes, followed by 1.0, assuming that you mean **23 as x 10 to the 23rd power. This is not a number slightly larger than 1.0, unless I'm totally misunderstanding what you are trying to say above. This may really be moot in Aram's project as the one million count/rev encoder would provide 5 million counts/inch with a 5 TPI screw, and that is only about 22 bits of resolution, assuming only a couple inches of travel. Even if you tried to do this with a very large machine, say 100 of travel, that's only 100 X 5 X 1 million = 500 million, or about 29 bits of resolution. That's just a few bits past the mantissa of a 32-bit float, which is 24 bits. Umm, yeah, I had been thinking this was all carried as DOUBLE floats! Definitely could be a problem with a high res encoder and a longer machine travel. I think the 24-bit counter is still OK for an encoder that gives 1 million counts, or even 4 million counts/rev. Probably would be OK, even with +/- 8 million counts/rev, since 24 bits can represent +/- 8,388,607 integers. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Request for information at the CNC Workshop
Matt Shaver wrote: On Thu, 2008-06-05 at 00:18 -0500, Jon Elson wrote: I don't think there is truly ONE way to do things. You pretty much have to use safety-rated controls from Pilz, Crouzet, Schaffner, Faulhaber, etc., and then set them up so they can function correctly to stop motion, even after a component failure. Yes. In fact, that's what I've learned so far! The most recent control I've built uses a commercial safety relay in the estop circuit, as well as an EN418 qualified (anti tease, positive acting) estop button. There's also some redundancy of power control elements and monitoring circuits that prevent resetting estop if a contact has welded closed. What I'm interested in is what I've missed! I learned all those techniques from manufacturers literature that refers to these standards, but I've never actually seen the standards documents themselves, and I'd like to... The problem is they all cost several hundred $ each, and every document references something like 100 other documents, with no indication how important that one is. ... pretty expensive. Then, you get into calibrated wideband antennas, and on and on. Finally, you have to test every different configuration, so you generally have to test each specific installation, if you are doing one-off machine retrofits. All true of course. I guess I put too much emphasis on the accuracy part of my request. I don't doubt that elaborate test equipment and facilities would be required to certify performance to a particular standard or spec. The most basic test for RFI is to tune around the AM radio band on a receiver with the antenna close to the equipment under consideration, listening for changes in the level of static. I'm just hoping to improve on this; to take it to the next step. Maybe build some sort of hobby standard wide band RFI detector useful for comparison purposes, rather than a calibrated measurement instrument. The problem is even a hand-held 7 transistor radio is way too sensitive. Try it with your home computer, it will pick up all sorts of fairly loud birdies, like the disk drive spindle motor drive circuit, the main power supply, etc. All the CPU stuff will be several orders of magnitude above the AM band, so you won't hear any of that, except sub-harmonics from memory and I/O data patterns. You'll be fairly surprised at how BAD the average PC is, no WAY they'd pass a real FCC class A test, but they come in, anyway. Even the cheapest radio has an AGC circuit or AM wouldn't work at all, but this makes it impossible to compare one buzz against another. They will all appear as loud, as the AGC's JOB is to make them all the same. It might not be real hard to hack a small radio from the 1970's to disable the AGC, it is probably all in one chip today. For example, if I put a ferrite bead on a cable, did it help reduce the RFI, or was there no change at all? Right now, I'm operating in the dark, installing filters, chokes and beads in an anticipatory, prophylactic way ;). What I need is a relative indicator to detect electrical interference, locate the source, and compare the effect of different mitigation methods. You might have some success tuning the radio to listen to local stations, and compare the noise WHILE tuned to the station, that will lock the AGC to the station's power level, so you can maike relative comparisons. But, of course, that only works on wide-band emissions. Most machine tool controls have had totally uncontrolled emissions, because they were mostly exempt as heavy industrial equipment. Note the lack of output filters on most PWM servo amps. Your own being an exception to this! Well, yes. And, I had some prior hints that people were having real trouble with various drives that messed up encoder and step/direction signals, so I thought it would be wise. Having 100 ns risetime pulses of 50 -100 V flying about systems that need to correctly sense 5 V signals which also have fairly fast risetimes, so you can't massively RC filter them seemed a bad idea. Devices to inject disturbances into the line and equipment chassis are not as bad as the spectum analyzer type gear, and you can even make much of this yourself, like with auto ignition coils and pulser circuits. I did make a vibrating relay type noise generator which helped recreate an intermittent problem with a spindle speed control board. Once they were able to reliably produce the problem, the manufacturer of the board was able to add hardware filtering and do some software modifications that eliminated the problem. VFDs, of course, are pretty powerful noise generators, too. Anyway, for noise suppression, the best thing is shielded cables tightly bonded to the cabinet, a cabinet free of large holes and unsecured slots, and a good power line filter module. Adding filtering to PWM outputs is also a big help. Slots formed by covers
Re: [Emc-users] PWM Amp Basics
Kirk Wallace wrote: A while back I played with a KBIC speed controller which demonstrated fairly good torque at high and low speeds. I was curious about how a PWM amp would do, so I set up a Pico amp to a parport PWM signal at about 25kHz and was surprised to find that low end torque was much less. I thought the PWM frequency might have had something to do with the torque issue, so I made a similar setup with a 50kHz PWM signal from a Pico PWM controller and got similar results. Then I got to thinking, the Pico amp is a PWM to voltage converter. The motor torque will be the torque for the voltage that the motor happens to be driven with at the time. I thought the KBIC had feed-back during the zero crossing period ,but I didn't realize how important it is. So, in order for the PWM amp to have good torque and speed control, feedback needs to be set up. Does this seem correct? Yes, you have it exactly. My PWM servo amps are truly PWM = voltage output, with no compensation for the motor resistance or back-EMF. You could, of course, use some kind of feedback. If you have a spindle encoder, you could make HAL read the encoder velocity and add a correction to the PWM output. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] PWM Amp Basics
Kirk Wallace wrote: What are the issues with choosing a PWM frequency? I would think that efficiency would be one, where you would balance switching losses and iron or copper losses? Evidence of this might be seen as higher motor or amp temperatures? Lower frequencies might produce noise. I believe the Pico PWM amp has a 25kHz to 100 kHz range? What would happen with lower or higher frequencies? Well, in my PWM servo amps, specifically, the output filters are tuned for 50 KHz, so that the PWM square wave is reduced to a small sine wave ripple on a DC voltage. I wouldn't increase the PWM frequency, as it increases losses with no benefit. lowering the frequency will What are the symptoms of not setting the bootstrap parameter, like for my parport setup? On my DC brush PWM servo amps, the IR FET driver chips need these pulses to come out of their shutdown mode. The brushless servo amp does it by internal logic, so you wouldn't need this with those amps. The bootstrap option in the PPMC driver generates a 5% PWM in one direction for one servo cycle, then in the opposite direction on the next servo cycle, then it goes to normal operation. It does this every time the amps are brought out of estop. Does the current limit resistor affect the amp performance or just set the current at which the amp alarms out? No, the current limit limits the current to protect the motor. Effectively, if you cause current to exceed the limit set by the resistor, then the on-time of the high side transistor is cut off. This limiting resets every PWM cycle. There is a second, fixed limit which causes the amp to go into fault state. These are two totally different current sense mechanisms. The fault sensing also will detect motor grounds, where the fault current would bypass the current sense resistor used for the limiting. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] PWM Amp Basics
Peter C. Wallace wrote: A reasonable hack might be to use the previous delta for velocity if index (and a count clear event) happened Good thought, this condition will only exist for one servo cycle, so it shouldn't cause any problems for a spindle speed control algorithm. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] PWM Amp Basics
Kirk Wallace wrote: Feedback would start with ppmc.X.encoder.X.delta? Is encoder delta reset to zero and at what event? This would then be scaled/converted to RPM and fed to pid.X.feedback and possibly a speed display. Encoder.xx.delta is never reset. It is derived by subtracting the last position from the current position. When the spindle index feature is used, the spindle encoder index pulse clears the enocer counter, and the driver mades the axis position zero. So, there would be a large jump in position at this instant, causing delta to be a large number on one servo cycle. You know when this happens, as the HAL signal ppmc.0.encoder.xx.index-enable will be true before the event, and will go to zero on the same servo cycle as you read the wild delta number. It looks like there is still some magic that needs to conjured before this will work. This should be quite simple to work out. I'm not even sure delta goes wild at these times, I've looked at the signals with halscope, but it has been a while. Reading the driver code, it seems it should go wild when the spindle index function is used, unless something else fixes it. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Pico on Pluto
Kirk Wallace wrote: I received a Pluto today and started playing with it. Thanks to some help from the IRC (Thank you Skunkworks) I got the driver to load using LPT1 (378). Although, it loads without the apparent fix (epp_wide=0) now too. The driver and EMC loads with my PCI parport card, but it it doesn't respond to AXIS commands, so there is more to troubleshoot. I set up the Pluto to drive my Pico PWM amp without feedback, and it seems to work fine but the large inductors got smelly, Do Not Touch, hot. I suspect the Pluto PWM frequency (19.5kHz ?) might be the problem but I don't see any way to change it. My guess is the Pluto is using synchronous antiphase modulation, where the bridge is switching the transistors on 50% of the time at idle. This provides full DC supply voltage across the inductor at the PWM frequency, and heating of the inductors is guaranteed, over a wide frequency range. Have you looked at the control waveforms with a scope? If it is synchronous antiphase, then you would lock the PWM optocoupler on by just hooking a resistor to the 5 V power, and drive the amp completely with only one signal to the direction input. Above 50 V or so, however, the output filters will overheat for sure. These amps are designed for sign-magnitude PWM. At idle, there should be no current in the PWM optocoupler, and the high-side transistors will be off. The low-side transistors will be on, shorting the motor. As the servo jitters, it will send very short pulses of current to the PWM optocoupler, which will turn one of the high-side transistors on for a short pulse. Briefly looking at the Altera datasheet it looks like the output pins are good for 25mA? I'd appreciate any comments on using Pluto's with PCI cards and how to keep the Pico amp inductors cool(er). If the Pluto tests out to actually be putting out the right sign/magnitude signals, then you must have a subharmonic oscillation that is reversing the inductor current on every nth cycle. With the low frequency of the Pluto's PWM, these oscillations should be audible. Proper tuning should get rid of this, and you should be able to see the PWM output command using Halscope. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Pico on Pluto
Kirk Wallace wrote: I haven't looked at the control signals with a scope. One inductor was much hotter than the other. I assumed because it was because I was driving the motor mostly one direction. I was planning on using 110 Volts for the motor supply. Hmm, synchronous antiphase would have both inductors equally hot. A lot of PWM jitter in one direction could heat the one giving the positive output more. The heating is worst at 50% on time. I am not sure what you mean, but I get the impression that you are saying that for synchronous anti-phase, the PWM signal and the direction signal are both modulated to control motor speed? No, PWM is just turned on 100%, and the only signal that is changing is direction. This would make it work like the output stages of a Gecko 320, for instance. I did put a voltmeter on the direction signal and got 0 and 3.3 Volts DC with a direction change. The PWM signal went between 0 to 3.3 Volts DC dependent on the speed setting. There was never any audible noise from the motor other than bearings and brushes, which actually was kind of strange. After looking at the Pluto blurb on their web site (a bit sketchy) it seems like sign/magnitude is the default, they call it pwmdir, I think. I am feeding the spindle command straight into the Pluto PWM input, so, I believe, there is no PID or anything else that could be tuned. I am a little surprised the Pluto PWM frequency isn't higher. Hmm, this is a strange problem. First, I don't think you will get very good speed control without a control loop, unless the motor resistance is very low, but second, I don't understand the heating. At 19.5 KHz, the output current will be discontinuous, and that will cause some slight increase in iron losses in the inductor. Maybe if you were testing mostly at half speed, that would put the PWM duty cycle at 50%, the worst case. What is your DC supply voltage? Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] PWM Amp Basics
Kirk Wallace wrote: Sometimes the the spindle context keeps me from seeing things as they should be. More thinking aloud. An encoder still acts the same whether it's on a linear axis or spindle. So if the spindle moves, the distance is counted, and sometimes, it's a very long move. The delta math is done each time the thread that contains the encoder function loops (servo-thread?), so the delta divided by the thread period gives the average velocity for the period just previously completed(?). A discontinuity occurs if a move in one direction is long enough to wrap the counter register. At least on the Pico Systems Universal PWM and Stepper controllers, the 24-bit hardware encoder count is rolled over to a 32-bit count. So, the HAL pin won't see an overflow except every 4 billion counts. But, there is a discontinuity when the spindle index feature zeroes the count when the axis is synced to the index. this is usually only done on an axis home operation, but would be done at the beginning of every threading pass for a spindle axis. So, with the index-enable feature, it is like making your counter register just big enough to count one encoder revolution before it wraps? But it seems that when you see an index event, you would know to account for it. No. The spindle count is set to zero at the beginning of the operation, and then counts up as many turns as the spindle rotates. encoder.xx.pos is scaled to count to 1.0 on the first turn, and continue like that. It does NOT reset to zero at each turn after completing the index function. For G33.1 (rigid tapping) it would start winding back down after the spindle reverses direction. I kind of assumed that the index-enable feature was created to fix the discontinuity problem for threading. I haven't a clue how though. No, it simple sets encoder.xx.pos to zero when it needs to align the spindle to the index pulse at the beginning of each threading pass, so the linear axis can sync repeatably to the thread. The only discontinuity is that you WANT to break the continuity of the linear axis to the spindle, or you'd have to rewind the spindle for each threading pass. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Pico on Pluto
Jeff Epler wrote: When you set pluto-servo.pwm.#.pwmdir TRUE, the output should be what you call sign-magnitude: value after scale+offsetPWM Dir 0.0 0%FALSE 0.5 50%FALSE 1.0100%* FALSE -0.5 50%TRUE -1.0100%* TRUE * may be reduced by max-dc paramter In emc2 versions 2.2.0 through 2.2.3 there was a bug in the pluto drover's pwmdir mode that rendered it completely broken. It should be fixed in 2.2.4 and newer. When in the default mode (pwmdir FALSE), the output is: value after scale+offsetUp Down 0.0 0% 0% 0.5 50% 0% 1.0100%* 0% -0.5 0% 50% -1.0 0%100%* * may be reduced by max-dc paramter There is no way to modify the PWM frequency of the pluto without recompiling the firmware. WOW, Dr. Eppler, I believe you have a DIAGNOSIS! With a negative command into the driver, it looks like synchronous antiphase! Kirk, can you check if the heating problem is only showing up in one direction? it also ought to have a non-symmetrical response. 50% command should give 50% velocity for a + input value, and the minus valuses should go from -100% speed to +100% speed, with a command input of -0.5 giving zero speed. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] wiki request
Andre' Blanchard wrote: What is the delay (in degrees of rotation) between reading the spindle encoder and positioning the Z axis? That delay would be in opposite directions going and coming out. Unless there is mechanical backlash between the spindle itself and the spindle encoder, this delay should be minute, somewhere in the couple millisecond range. At a couple hundred spindle RPM, it should be negligable. Backlash in the Z axis, combining all factors such as axis following error, leadscrew backlash, spindle axial backlash and spring in the link between ballscrew and quill are most likely the culprit. This might be a lot easier to measure than the actual error during a threading cycle. It would be a VERY interesting experiment to perform, though, and could possibly turn up some kind of quirk in EMC2, such as a software delay introduced in the Z axis motion as it follows the spindle. I seem to recall there is some smoothing applied to the Z motion. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Pico on Pluto
Kirk Wallace wrote: On Sat, 2008-06-07 at 13:06 -0500, Jon Elson wrote: ...snip would put the PWM duty cycle at 50%, the worst case. What is your DC supply voltage? Jon 53 VDC unregulated with 500 uF of filter caps. So far the motor has had no load. If you need more information let me know. If you want, I could also ship this setup to you. On the other hand, I was just being curious and currently have no application that will use this configuration. Well, I've also found a goof in the servo amps made since about Nov 2007. I changed power transistors, and should have re-tuned the deat time setting. Some units have feedback between the power transistors and the opto-isolators, and oscillate when the controller is sending small pulse widths. Some amps trip the fault sensing, others just run hot, whistle, or work fine. I just discovered this this weekend. It is a fairly simple fix, but requires taking the whole amp apart to put two caps on the back of the board. Your last order of servo amps may be included in this batch. If the power transistors say FB260 on them, they need the fix. If they say 31N20D then they don't. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Pico on Pluto
That should have read re-tuned the DEAD time Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Pico on Pluto
Jeff Epler wrote: Jon, What *should* the driving signals look like at the input of your amplifiers? I assume from your response that I'm making some kind of rookie mistake here, compounded by the fact that I still don't understand what the mistake is. No, maybe not. My amps were designed for sign/magnitude. The direction signal is pretty obvious, and the PWM signal should have current in the optocoupler's LED when you want the power transistor on. So at idle, there should be only very short current pulses in the PWM opto, amd the direction will be flipping back and forth as the control loop dithers. What you want to avoid is dithering where the PWM is more than a couple percent and the direction is flipping back and forth each PWM cycle. The PWM can go to zero%, but should not exceed about 97% on-time, to keep the bootstrap capacitors charged. These amps CAN be driven in synchronous-antiphase, where you turn the PWM on all the time and control only the direction signal, but above about 50 V on the DC supply, the filter inductors will really overheat badly. So, I don't recommend that mode of operation. Here, the idle situation would have a 50% duty cycle on the DIR input. There's also the bootstrap problem, where the amps need a pulse in each direction every time they come out of estop to reset the shutdown FFs in the FET driver chips. (My AC servo amps handle this in the CPLD, but the 4000-series CMOS in the DC brush version of the amp is the epitome of dumbness.) Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Pico on Pluto
Kirk Wallace wrote: I can certainly try, but I am using the latest 2.2.5 which should be fixed. I have pwmdir set true. I measured only logic levels on the dir pin and analog on the PWM pin. The no load forward and reverse RPM through the range (0 to 5500) was within 50 RPM. OK, not a diagnosis, then. Well, it may be the feedback problem, it manifests differently on different units. Do you get faults on the amps, too? Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Pico on Pluto
Kirk Wallace wrote: These are a little embarrassing, but here are some pictures of my Pico/Pluto setup: http://www.wallacecompany.com/machine_shop/Pluto/ The cores of the inductors seem to be heating and not the copper, so it takes a while for the heat to be detected. It seems that one or the other inductor heats up dependent on direction of the motor rotation. I seem to get no heating at zero speed and have only done brief testing up to 30% speed. The upper trace is the PWM signal at the amp connector. The lower trace is the direction signal at the amp connector. Wow, I thought I was the only one who did that desktop computer thing! Well, I'm a bit mystified by this core heating thing. It is not unusual for them to get warm, and if you are running the servo motor at constant speed for several minutes, then I wouldn't be terribly surprised at this. The hot core would be on the side that is sending a positive voltage to the motor. On the other side the ground transistor would be on all the time, and that inductor would run cooler. You say you are experiencing this with only 50 V on the DC supply? I can do some experiments here to see whether I can duplicate the result. I was actually running one of these amps on the spindle of my minimill tonight, but the program was reversing the spindle a bunch as I've been experimenting with rigid tapping. I didn't notice any unusual heating there. And, it could be this feedback thing, too. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] wiki request
Chris Radek wrote: That was exactly my thinking too. But as others have pointed out, there could be other things at work. Looseness/backlash in the quill spline on my bridgeport is significant for instance. The first step to compensating for it is measuring it, so I was concentrating on finding a way to do that first. I figured out how I might fix this on the ancient Bridgeport M head I used to have, but I exchanged it for a much better J head. Anyway, I got the idea that I could cut strips of brass shim stock of a thickness of half the total spline gap, and slip them between the spline teeth, and then bend the ends over on the pulley. It might take a bit of fiddling to find the right shim thickness that would fill the gap but not bind the travel of the spindle. - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Pico on Pluto
Kirk Wallace wrote: On Sun, 2008-06-08 at 01:21 -0500, Jon Elson wrote: ... snip Your last order of servo amps may be included in this batch. If the power transistors say FB260 on them, they need the fix. If they say 31N20D then they don't. Jon My amps have FB260's. Should I put some caps on them? If you are truly a masochist, there are two places on the bottom of the PC board that have no component there. You should place 100 pF 0805 capacitors in those two locations. You have to take the 4 screws out of the transistors and then the 4 screws that hold the board in place. You then have to reassemble it, with the thermal conductive insulators under the transistors, and the plastic step washers and plain washer on top. And, if you don't want to perform this surgery, you can send them to me any time you are not needing them. All of this will make the amp a bit more stable, but it probably won't have much effect on the heating problem. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
[Emc-users] rigid tapping
Well, I have finally gotten the rigid tapping working, but I'm not totally satisfied with the results. I'm getting a little mauling or damage to the top part of the threads. I am using combined drill-taps. What I am doing now is drilling the hole through first, then pulling the tap back, and starting the G33.1 tapping operation with the tip of the drill-tap even with the top of the material. This makes the G33.1 return the tap to be completely outside the material while the spindle is in reverse. So, I think the damage to the threads is because I'm giving the drill-tap too much axial freedom in a home-made spring-loaded tap driver. I had one anomaly where the spindle just mashed down onto the tap and smashed it to pieces. At least, that's what I THINK happened, it happened very quickly. It reminded me of the problems I had last year when the spindle sync/home to index pulse function had several problems in it. I ran a bunch of air tests without a repeat, so I wnt back to production and had no further problems. I did manage to tap 208 holes with this setup. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] My Mill on EMC2
On Jun 8, 2008, at 6:53 PM, Greg Michalski wrote: Just thought I'd share my benchtop mill powered by EMC. A question though - when I have the file loaded and touch off to the workpiece - it takes FOREVER to get back to ready to let me touch off the next axis. What is this the result of? The .ngc file is www.distinctperspectives.com/X3/LimitBases.ngc I know it is a lot of milling (it took a while to run) but it took longer to update after touching off than it did to parse the file at loadup. I know I'm not running a top end system (it's a P3 650MHz w/384mb memory and onboard video that I picked off the curb one winter afternoon) I wonder if just backing off on the BASE_PERIOD setting would help. The real time section with software step generation may be taking a little too much CPU time. One other thing, do you have it on a network? Something I've noticed on one system where I share the network jack is that any user interface stuff will run insanely slow for 10-15 minutes after you boot. I forget which network process is burning all the CPU time trying to establish some connection. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] rigid tapping
Anders Wallin wrote: So, I think the damage to the threads is because I'm giving the drill-tap too much axial freedom in a home-made spring-loaded tap driver. We clamped the tap in a standard ER25 collet, no axial freedom (it's called _rigid_ tapping after all...). We had the g33.1 move start with the tip of the tap around 5mm above the hole so the spindle and Z would have time to accelerate to cruise speed before hitting the hole. Our threads appear OK, perhaps somewhat loose, but that may be because the drill makes a slightly oversize hole. I'm comparing this to using my Procunier CNC tapping head on the Bridgeport. It has a very small amount of axial freedom which selects which clutch to engage. I guess I will have to try it without the springs and see if that looks better. I thought these springs were stiff enough to constrain a 4-40 tap in aluminum. My concern was that this sloppy, home-made machine retrofit would have too much mechanical backlash in the Z travel, and would be putting a lot of axial loads on the tap, especially after it reverses. Thanks, Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] My Mill on EMC2
Jon Elson wrote: One other thing, do you have it on a network? Something I've noticed on one system where I share the network jack is that any user interface stuff will run insanely slow for 10-15 minutes after you boot. I forget which network process is burning all the CPU time trying to establish some connection. I was really unclear in this message. What I was trying to say was that if the network is unplugged, the machine logs in and runs very slow (several minutes from mouse click to terminal window opening, for instance.) This usually takes under 5 seconds with the network plugged in. About 10-15 minutes after booting with the network unplugged, whatever was hogging the CPU gives up and it runs normally. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Pico on Pluto
Jon Elson wrote: Kirk Wallace wrote: On Sun, 2008-06-08 at 01:21 -0500, Jon Elson wrote: ... snip Your last order of servo amps may be included in this batch. If the power transistors say FB260 on them, they need the fix. If they say 31N20D then they don't. Jon My amps have FB260's. Should I put some caps on them? I have another test. Try running the system exactly as it is when the inductors overheat, but with the motor disconnected. If it still overheats, you can remove C14 and C15 (right near the power connector). These are the capacitor part of the output filter. Of course, this will allow the servo amp to radiate EMI, but it will remove part (maybe a major part) of the circulating alternating current from the inductors. Let me know what you find, I'm still trying to decide if this is a major problem in all applications, or only a concern on spindle drive applications. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Pico on Pluto
Kirk Wallace wrote: On Mon, 2008-06-09 at 20:24 -0500, Jon Elson wrote: ... snip I have another test. Try running the system exactly as it is when the inductors overheat, but with the motor disconnected. If it still overheats, you can remove C14 and C15 (right near the power connector). These are the capacitor part of the output filter. Of course, this will allow the servo amp to radiate EMI, but it will remove part (maybe a major part) of the circulating alternating current from the inductors. Let me know what you find, I'm still trying to decide if this is a major problem in all applications, or only a concern on spindle drive applications. Jon The inductor still heats up with the motor disconnected. I haven't tried removing C14 and C15 yet. I'll let you know what happens. OK, then. This will work, and will allow the inductors to run cool (or at least I have strong reasons to suspect so.) it is a comdination of the low PWM frequency and the continuous operation at higher duty cycles. Hopefully, there won't be interference problems with encoders or whatever. If so, a small 250 V low-loss capacitor would need to be put in C14 and 15. But, you should be able to try it without, first. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Pico on Pluto
Kirk Wallace wrote: Removing C14 and C15 stopped the inductor heating without affecting anything else, as far as I could tell. I wonder if smaller caps or a different core type might help. Excellent! Well, these are 47 uH 15 A inductors, and pretty amazing at that size. I have found nothing else anywhere near the size or price of these units that matches that rating. I think the caps are way too big. It is not necessary to completely suppress the PWM waveform at the motor, just knock down all the harmonics. A much smaller capacitor would do fine, maybe .001 uF or so. I'd use a 250 V film capacitor for low losses. A ceramic might well get very hot due to the AC current. But, if everything is running fine, there is no need to try adding small caps back. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Gear Cutting
Kirk Wallace wrote: Has anyone tried cutting gears with something similar to this arrangement? http://www.wallacecompany.com/machine_shop/Gear_Cutter-1b.png I was thinking a slot saw (gray disk) could be used, centered on the gear shaft(violet). The gear tooth form could be followed with Y while rotating the gear (bronze color) and incrementing X on each gear rotation until the width of the gear is complete. Or successive passes in X and incrementing Y and A could make a complete tooth so that one gear rotation would complete the gear. Slot saws aren't very stiff and don't side cut, so some other cutter would be needed. Part of my thinking is that I would like to avoid special cutters like those needed for normal gear cutting. In theory, this can be done. A thin slitting saw would deflect too much to get an accurate tooth profile. You can buy gear tooth cutters and run them like this, and it will go much faster, which is still fairly slow. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Gear Cutting
Steve Blackmore wrote: On Thu, 12 Jun 2008 00:38:02 -0500, you wrote: In theory, this can be done. A thin slitting saw would deflect too much to get an accurate tooth profile. You can buy gear tooth cutters and run them like this, and it will go much faster, which is still fairly slow. http://www.jeffree.co.uk/Pages/cnc-wheel-cutting-engine.htm I've seen that in operation and used it at a show here in the UK, it's not slow, gear cutter rpm was about 2500 rpm and you can stuff the cutter through the blank full depth - I guess feed was about 100 ipm. If you do it on a horizontal mill with arbor supported at both ends, you can do it like that. 2500 RPM certainly wasn't an HSS cutter in a steel gear blank. On thin brass blanks, I reckon it takes no longer than 1 second a tooth. You can't do that as the gear blank willl fold up. So, you need some discs on the side to support it. For moderately thin gears the support might be made slightly under the tooth root diameter, for really thing clockwork gears, the support discs probably need to be full tip diameter. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Gear Cutting
Ian W. Wright wrote: One of the main reasons I want to try to generate gears and, particularly, pinions is the great problem I have in trying to make working pinion cutters small enough for the watches I work on. It just seems to me that if you have a CNC machine of any type, you should be able to cut a master tool for the form required. Then, that tool could cut the gear teeth directly, and reduce a 5 hour job to 15 minutes! Even if you only have a mill, you can mount a disc on the spindle and a lathe-type tool in the vise, and make a gear cutter by laboriously following the tooth profile. Then, you could cut radial slots to form the cutting teeth, harden it, and you'd have a pretty professional gear cutter for any tooth profile you need. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Gear Cutting
Greg Michalski wrote: snip Another thought comes to mind, could wire EDM be used to cut very small gears and pinions? You bet! For hair-thin wheels and intricate escapement profiles, they are the way to go. Of course, wire EDM is a whole other domain, and you can't buy a $500 Chinese wire EDM machine like the desktop mills. I wish we could get Robert Langlois to come down and show his stuff at the CNC Workshop. I met him years ago at NAMES, and he was working on some incredible stuff even back then. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] E-stop Surprise
Kirk Wallace wrote: My mill isn't fully functional yet and I neglected to realize the ramifications of this. A bracket I needed to make required a tool change, so I decided to just do it manually. I checked the program by lowering the knee and single stepping through the program, which at the time seemed okay. I raised the knee, started the program and drilled the holes. The table homed for the tool change displayed the tool change message and proceeded to go ahead with the program without stopping. Unfortunately, the new tool that didn't get changed is shorter so the tool collided with the table. I tried to click the e-stop button in AXIS but it had no effect. Then I realized that the tool change message had the window focus and I needed to clear it before the e-stop would work. I realize I made a few mistakes here, by not having a proper config file and a hardware e-stop button, but I wonder if there should be a way to have the AXIS e-stop always on top and available. You really HAVE to have a hardware E-stop! The business with the pointer focus on the GUI is pretty dangerous, and I run into it all the time. Usually just hitting Enter will get your focus back to the main GUI window, but that is still dangerous. Given how the windowing system works, I don't think there's a way to do what you want. Put in a harware E-stop button. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Endmills
Andrew Ayre wrote: Sorry if this is a bit OT, but I'm stuck. I have EMC2 running and my CNC machine assembled and now it's time to try cutting. I want to focus on wax and wood and start with some pine. I'm using a Dremel for the spindle. I've been reading the CNC forums and various websites but I cannot figure out what kind of endmills I need. Can anyone give me some pointers? Is there a beginner's guide on the web somewhere? All I want to do at this stage is: - cut objects in pine - expect to break a few endmills so looking for cheap - want to avoid junk endmills - not sure when you would use a ballnose mill - difference between endmills for roughing and finishing I expect to be cutting parts that are perhaps 20 - 30mm on each side with the smallest feature perhaps 2mm. My scrap pine for testing is 3/4 thick so I guess I have to do some pocketing before cutting. Well, the Dremel's collet and speed steer you to a smaller size endmill. I used to get 1/8 2 and 4-flute endmills in solid carbide for about US $3.00. They probably are more, now. I would buy them by the dozen at one of the eBay sellers like Reliable End Mill, I think they are now Relaibel Tool. It may be in these materials, that standard Dremel router bits would work fine, these should be available at hobby shops, hardware stores, etc. In these sizes, you are not likely to find roughing end mills. Ballnose mills are good for profiling 3-D surfaces. You probably won't find junk in solid carbide. There were some true junk in HSS end mills with 3/8 shanks, in square blue plastic tubes with the size rubber stamped on a paper label. These were the worst crap China ever made, and I'm insulting the sight-impaired by saying they looked like they were hand sharpened by an 80-year old blind man with shaky hands! Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] You guys might find some aspects of this EMC Survey interesting.......
Renewables Not Reactors wrote: See here for details... http://www.cnczone.com/forums/showthread.php?t=59635 Sorry, can't fill this one out. it asks why I DON'T use EMC! Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
[Emc-users] EMC Fest
So, Who is at the fest and set up? I'm frantically putting stuff together and trying to get ready to head up there tomorrow. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] EMC Fest
Jeff Epler wrote: so far we've got Jeff Epler John Kasunich Steve Padnos Kenneth Lerman Steve Stallings Matt Shaver Ray Henry Great! Where's the webcam? It looks like I will be assembling a servo system there, as I have run out of time here to put it together. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] EMC Fest
Jeff Epler wrote: On Sun, Jun 15, 2008 at 11:02:05PM -0500, Jon Elson wrote: Great! Where's the webcam? http://linuxcnc.org/compile_farm/festcam.shtml HUH! Bunch of pikers - where the heck are you? All I see are empty chairs! Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Shutdown during milling
Andrew Ayre wrote: I have noticed that perhaps one time in five Linux will partially shutdown during milling. It's a bit hard to describe but the Gnome desktop is replaced with a text screen of what looks like the tail end of the bootup sequence: Operation [OK] Another operation [OK] Running startup scripts in /etc/rc.local [OK] There are no messages and EMC2 continues to generate pulses - which cause the X and Y axis to move the spindle in a straight line. A quick press of the power button turns off the power. I booted and checked dmesg, user.log, syslog and messages. Nothing was recorded to indicate a problem. The syslog only had the 30 minute marks in it. This may be a crash of the X-windows server. You didn't, by any chance, hit Ctrl-Alt-Backspace? That is the command to kill the X server. If X is actually crashing, the message file is in /var/log/XFree86.0.log and may show some useful info. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Ball Nose
Andrew Ayre wrote: Thanks John and Kirk, I will use an end mill for roughing. Andy John Thornton wrote: Use a end mill for roughing unless there is very little to remove. Ball mills don't cut as well. Likewise use a drill to remove most of a hole the final size it with an end mill... John On 22 Jun 2008 at 11:17, Andrew Ayre wrote: Suppose I wish to cut a dome at and the edge of the dome I have to cut 0.5 below the surface of the wood. Do I use a end mill for roughing then a ball nose for finishing, or just use a ball nose for the entire job? There are also bull nose or radius end mills that are halfway between square end mills and full ball mills. These have the advantage that there is no point to chip. Since the radius is less than that of the entire end mill, you have to calculate things differently for these. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Website Found
Matt Shaver wrote: The content seems pretty old: Those who bought Anilam's original 386sx-16 could easily upgrade to the current standard, 486dx-33. Hurco will soon offer a PC motherboard as the main processing engine ... Looking through the fora on the main menu, I see a few that have posts as recent as Feb 2008, but nothing more recent. Gee, that's pretty stale in this business, never mind the 386/486 stuff. Outside of eBay, it would be hard to find even a 486 today. I even trashed all the 486's out of MY basement a number of years ago! Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] FYI, HNC on eBay
Kirk Wallace wrote: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItemitem=330246010908 Will they MAKE you take the GE 550 control? If so, you'll have to dispose of that clunker. I don't believe there's anything there that would be usable. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] FYI, HNC on eBay
Kirk Wallace wrote: On Thu, 2008-06-26 at 11:09 -0500, Jon Elson wrote: Kirk Wallace wrote: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItemitem=330246010908 Will they MAKE you take the GE 550 control? If so, you'll have to dispose of that clunker. I don't believe there's anything there that would be usable. Jon I reused the servo power supply. There are other higher(ish) current supplies that might be good for something else. Whenever I need a piece of wire, screw, switch, etc. I visit the 550. It's like having my own little RadioShack. My WHOLE basement is like a Radio Shack, only FAR better! I guess if you don't already have such an archive of goodies, it might be of use. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Trash?
Ian W. Wright wrote: Jon Elson wrote... I even trashed all the 486's out of MY basement a number of years ago! If you are throwing out old PC boards like this it is worth stripping out the old processors and maybe some of the other big chips first. The ceramic processor packages make excellent polishing hones for finishing tool tips or gravers on. just push the pins into a lump of wood or cork first I have at least 100 Lbs of PC boards with FAR heavier gold than PCs. This stuff dates as far back as the 1950's, when gold was used routinely as the etch resist. I attack the pile as I have time, chisel the pins off big wire-wrap backplanes and such. I got about 1 Kg of WW pins off one box of trashed backplanes. I also have maybe 20 Lbs of sawed-off card edge connectors. I am hooked up with a group of gold refiners, and one guy came up with an incredibly cool process that removes the gold as metal flakes from these pins, using all hardware store chemicals! I haven't tried it yet, though. PC's, however, really don't have a lot of gold in them. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] FYI, HNC on eBay
[EMAIL PROTECTED] wrote: If you had to take it would still be worth it because nobody makes spindles as good as the one in that HNC -Maybe the control would fall off the truck on the way home if you were lucky Well, the spindle on my Sheldon R15-6 (D1-6, 2.25 spindle through hole, double-row zero-clearance ball bearings about 5 diameter) is mighty fine, and will go to 1250 RPM without overspeeding the motor. Some models used a two-speed motor to get up to 2500 RPM. I'm not going to do that with a chuck on there. One of these days I will get around to CNC'ing it. I want to make it so it can still be used as a manual lathe. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] FYI, HNC on eBay
Tony Bussan wrote: Now, nobody bid. It's mine, all mine! Tony Heh heh! Lucky you! Well, there is a community of EMC retrofit owners developing on these. Jon - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] EMC the right choice for us?
Kai Schaeffer wrote: This sound not so bad. Actually we need this speed only for the rapid movements. For the cutting speed we have to max. speed of 5m/min. That would mean 0.02mm per cycle. I guess we could work with this. Or what do you think? Depending on thew hardware interface, you could probably up the servo cycle to 10 KHz with little trouble, given a current, fast CPU. There's also an issue with the speed at which the interpreter can process and output G-code lines, but others know more about this... Good point. What happens if the system is running out of lines? Does the machine stop properly and continues when more data is available? Why would it run out of lines? It should always have a buffer of interpreted G-code to read ahead. I did some experiments with the relatively new G64 P command to set the allowable tolerance during contouring. I was doing 588 blocks of G-code a second, and that seemed to be limited by the feedrate I had set and acceleration limits for the machine, not the CPU. This was on a 600 MHz Pentium III, so much faster hardware is available. - Measurement of the surface for a Z-correction probing? At the beginning of each program we measure the Z-profile of the surface of the sheet. This profile is used to correct the position of the Z-axis to get a precise cutting depth. EMC currently doesn't have a feature like that. I suspect it could be done, but it wouldn't be trivial. - We would prefer to use a bus like CANopen, Sercos or EtherCAT to control the motors you will have to write and test the drivers for the PCI-cards that provide these communication formats. OK, and how hard is this? Well, I wonder if it is really necessary. What control signals do the current drives use? Can they use PWM or +/- 10 V velocity commands? - SPC functionality what is this? Stored Program Control (in German called SPS). OK, we have Classic Ladder that allows you to graphically design ladder diagrams for the control interface. We also have HAL, which is more text-based, but you can write your own functions algorithmically, then connect with signals. You can mix HAL and C.L. parts, too. Jon - Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! Studies have shown that voting for your favorite open source project, along with a healthy diet, reduces your potential for chronic lameness and boredom. Vote Now at http://www.sourceforge.net/community/cca08 ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] EMC the right choice for us?
Kai Schaeffer wrote: Andre' Blanchard schrieb: What does it wouldn't be trivial mean in man-months? Actually, I think those were my words. EMC doesn't currently have a feature to scan a surface and then modify the meaning of a G-code file. There was some discussion once about a general transform system to map arc-type deviations in axes and non-orthogonality. I don't think any of that was ever implemented as a standard part of EMC. Actually, a kinematics routine that interpolated deviations from a matrix shouldn't be real hard to do. EMC2 already has some scanning cycles that could acquire the grid of Z points. Then, you'd need to have a mechanism to send that matrix to the kinematics for compensation. This might be fairly easy to do, especially for someone who knows the kinematics code (not me). Jon - Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! Studies have shown that voting for your favorite open source project, along with a healthy diet, reduces your potential for chronic lameness and boredom. Vote Now at http://www.sourceforge.net/community/cca08 ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Editing the GUI
Renewables Not Reactors wrote: What software is recommended for editing the gui for emc? Any text editor will work. Some like vi, or gedit, I prefer emacs (an optional software package). Jon - Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! Studies have shown that voting for your favorite open source project, along with a healthy diet, reduces your potential for chronic lameness and boredom. Vote Now at http://www.sourceforge.net/community/cca08 ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Rigid Tapping
Kirk Wallace wrote: On Wed, 2008-07-02 at 19:44 -0500, Chris Radek wrote: One thing that jumps out at me is that it seems your accelerations on X/Z are extremely low. It looks like it takes an inch or more to get up to your rapid speed. That's correct. I seem to recall that I had to balance between acceleration and tripping the following error alarm. I should probably go back and optimize my tuning sometime (after I start making money). If accel really needs to be so low you should try starting the tapping move way outside the part (an inch) or tapping at very low rpm and see if that helps. It's possible it hasn't settled into sync with the spindle quite yet, when it's cutting the first couple threads. OK, I think I see a problem. Giving more approach in air will help as the tap enters the workpiece, but then you have the problem the Z axis has to reverse in sync with the spindle. If the spindle reverses too fast, the Z can't keep up, and the approach in air can't help that at all. The only way to help this is to increase Z acceleration or slow down spindle acceleration. Kirk and I talked about the importance of cranking up the SERVO and TRAJ rates and I think he is running both at at least 1kHz now. Any slower just isn't enough to get a nice synchronized reversal. Chris Oops, by the time I got out to the shop, I forgot about the rate corrections. Still, I think g33.1 is working well as is with a .1 to .2 lead-in. I think the problem is inherent with the tap or fluid I am using. The more I think about it, the more I think running a second g33.1 to chase the thread is the way to go. To be honest, I really can't tell the difference between the double g33.1 parts and the hand chased parts. I suppose I should tap a part on my manual lathe for comparison. If you haven't upped the traj_period rate, I think it will do more than ANYTHING else to fix the problem. I spent hours fooling with every other parameter with little benefit, then at the CNC Workshop Jeff told me about this and it made an immense improvement. Jon - Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! Studies have shown that voting for your favorite open source project, along with a healthy diet, reduces your potential for chronic lameness and boredom. Vote Now at http://www.sourceforge.net/community/cca08 ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] EMC Gallery added to CNCZone
Renewables Not Reactors wrote: I have added the EMC video gallery to CNCZone. Expect a lots of newbies to appear here :) http://www.cnczone.com/forums/showthread.php?t=60619 Great! I believe either the 3- or 5-axis CNC mill at Hogeschool Antwerpen is controlled by my Universal PWM Controller and PWM servo amps. A lot of COOL videos! Thanks, Jon - Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! Studies have shown that voting for your favorite open source project, along with a healthy diet, reduces your potential for chronic lameness and boredom. Vote Now at http://www.sourceforge.net/community/cca08 ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
