On Friday 19 August 2016 01:00:57 Greg Bentzinger wrote: > Hello LCNC users hive mind; > Chuckle.
> I am finally getting to the actual wiring stage of my LCNC retrofit of > my 1987 HURCO KM3. The point of this message is to say what I have > planned and if anyone notices that I’m heading for a cliff, a wall, a > Magic smoke event you can yell at me now, or if anyone has been down > this road, tips would be appreciated. > > The Hurco KM3 is a Bridgeport Vari-drive style clone knee mill with > the Iron being made by Kondia in Spain (Manual machines were also sold > in the USA under the Clausing name). The servos are big ElectroCraft 4 > brush DC units set for 1:1 direct drive with encoders and tach > feedback. These are driven by a CMC / Randtronics Servomate unit. Each > axis amp is fed 60VAC and has a healthy bridge rectifier. These I > believe are set for velocity mode using +/- 10V signals. I have not used anything in my simple systems except the 5i25. It can generate a fairly high frequency PWM that can be fed to a VFD's input at its 5 volt analog input. The noise filtering in the VFD will average the PWM into an analog signal the VFD can use. One pitfall to be aware of if there are opto-isolators in the signal path is that their ontimes and offtimes do not match and they cannot be used where the PWM carrier is above a few kilohertz. So control linearity goes to hell at both ends of the range. I have both parports in use on my mill, and the breakout board used had some sort of cheap opto's in series with every line in and out. I am controlling a Pico Systens 20 am, 160 volt rated servo amp, driving the brushed 1hp motor in the spindle of my G0704. The OEM controller, an SCR/TRIAC based unit, could just barely make 2200 revs fwd and by design, about 1100 in reverse. With this setup fed from a 123 volt dc supply good for about 2Kw continuous, I can get around 2850 revs, either direction. But to get anywhere near a linear control, I had to excise the opto from that output, jumpering it direct. That worked so well that I took the lincurve module back out of my hal setup, it was no longer needed, with a 268 edge optical encoder on the spindle, the PID handles any remaining error. And with that servo amp being a full 4 quadrant amp, I've had to insert some ramp limiters before the PID, because that servo is capable of reverseing that motor from 2500 revs fwd, to 2500 revs in reverse in 250 milliseconds, quite brutal on the gears, so I have it stretched out to a short second. The PWM-servo goes into current limit, set by its resistor that programs that to about 15 amps, and it makes a distinctive chirp when it does a higher speed reversal while doing rigid tapping which I've written a wrapper for so I can tap a hole a fraction of a turn of the tap per "peck". I have pecked a 4mm.7 hole, 25 mm's deep for gib adjustment drive screws that way. > I believe a > single servo power signal enables the entire 3 axis amp panel. With > the MAX II control rapids were 250IPM, the MAX32 would jog at 400IPM > and do 300IPM in AUTO. > > I have the Mesa 5i25 / 7i77 combo and after just re-reading the manual > I’m a bit confused as the word “field” is used for many different > things and I’m feeling like I am stuck out somewhere in left field. It > appears field can mean GPIO, or it can be a 8-32VDC power source, or > ??? I am un-clear. > So was I, in particular getting the GPIO straight in my mind as it takes 3 lines of hal code to set one up and use it. > The existing limit switch system uses Prox sensors for X & Y and a > roller microswitch for the Quill on Z. I use roller micro's for all 3 homes, and software limits set in the ini file. That has worked very well here. > The knee is manual only. These > prox switches have a working voltage of 5VDC, the documentation fails > to say what the steady state is when not tripped. The needed polarity of that signal can be changed by using the GPIO's -not pin setup option in the hal file, so that is a set & forget. > Much of the hardware > actually uses 120VAC as control voltage, however some devices use > 24VDC. Since I have removed the blazing fast 8086/8087 card racks and > the massive supporting power supply I have ordered both 5VDC and 24VDC > power supply units. These new power supplies will provide the 5VDC for > the 7i77 logic and encoders, and the 24V will provide “Field” power > and provide power for relay sources. Here is where things get messy – > the servo amps take input from the limit switches so that when you hit > a limit, motion towards the limit is disabled but motion away from the > limit is allowed. Not only that, there are inputs from the control to > kill motion if you approach a soft limit. I may have to bypass these > hardware functions, it is yet to be seen. In LCNC, limit switches disable all motion, and if you hit a limit switch, there is a mouse sequence in the axis interface, may be in others too, that will allow a manual, by the mouse & keyboard, movement off the limit switch. Clumbsy, non-programable in gcode, but it works. My toy mill's teeny ball screws each have a switch that prevents the screws from being unscrewed from their respective nuts. > > The spindle speed will continue to be manually set via vari-drive > crank. I intend to use a Huanyang 2.2Kw VFD to drive the spindle motor > and report spindle at speed. I won’t have a spindle encoder initially, > however I would love to add one eventually if possible. I don’t know > if I should use the RS-485 module for control or I should just > hardwire in for +/- 10V. I am going to hard wire direction and 5 volts of pwm to drive my inverter, a 1.5 hp model running a 1hp 3 phase motor. On the table, running free, its run at 20hz for half an hour at a time, and at 200hz, a 10/1 ratio, and if I want to listen to the backgears, thats another stepdown to watching grass grow speeds. How well it will take that under load remains to be determined. > Once upon a time, long ago and far away at > the Los Angeles convention center while attending the WESTEC Machine > tool show I saw a retrofit kit which replaced the vari-drive pulleys > with 8mm pitch toothed belts and swapped out the OEM motor for a wide > range 5hp inverter rated motor. This setup still had the Hi/low back > gear but for all speeds within a range it was fully programmable. I wouldn't be happy with anything less, but I will retain the 4 step by moving the belt speed change too that is already part of this Sheldon 11x36. I'll have an encoder on the spindle when it makes first swarf. > Coolant pump is 3ph also but I am going to try just using a run > capacitor with 230VAC 1ph. Spindle brake and Mist coolant will also > need to be wired in. I also have a control pendant with handwheel, > axis switches etc. As much as I hate the idea, I will run this machine > via PC keyboard and the jog pendant initially till I can plan out the > wiring and layout for a proper control console. Also this machine is > going to have to make some money to buy the extra bells and gizmos. Good plan. But it seems you have to spend the money first before its able to generate the profit... > I have a 8” rotary table I want to convert for an “A” 4th axis. I will > need the mill functional for that project. I am jealous, my A axis is a 4", complete with lots of backlash and a 262oz motor I installed. :) > PC main board will be a HP DC7700 CMT mATX with a Pentium D 945 (Dual > core running at 3.4 GHz) 4GB ram. Not sure, but I will likely use the > onboard Intel 965 chipset video. I have several Radeon 5450 cards I > could use if offloading the video will help things. The Radeon driver, in the cards that I've had, were a disaster, tying up the IRQ's by locking them out for hundreds of milliseconds. So of the 3 machines I have under cnc control now, 2 are atom powered, using its onboard video, and one is a much more powerfull Dell Dimension 745. Its an altogether decent box that also has intels onboard video. Available cheap as they are off-lease stuff at your local computer shop now, let them keep the hard drive & winders license and save 50 bucks. Put a 5i25 in it and install LCNC from the live dvd iso. I begged an got one with 4Gb of ram, $100 and change. 2Gb will do as thats all thats in the atom boxes. > > Any advice is appreciated. I've done small stepper systems but this is > my first foray into big iron. Thanks All. Mine too. Welcome to the party. :) > Greg, out yonder in Colorado > > ---------------------------------------------------------------------- >-------- _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) Genes Web page <http://geneslinuxbox.net:6309/gene> ------------------------------------------------------------------------------ _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
