Re: [Emc-users] Homing and f-error
Ray Henry wrote: It could also be me reading it wrong. I had an issue trying to home using a shared limit pin. The one axis I was homing was fine but all the others would fault. No way around it after the ignore limit code was made single axis. If the limit switches are shared, then the first axis must be moved off the limit switch before the 2nd axis can search for home, then again for the 3rd axis. Just setting the home position to a large enough non-zero value should accomplish that. Long ago we found that motion stuttered during homing. Some sort of pause in the motion while EMC looked at the home pin. If this is still the case, it might cause the increased following error that was observed. No pause anymore. There is a bit of a glitch, at least on some encoder counter schemes, when the encoder count is suddenly set to zero by the index pulse. It is pretty small, as the driver masks most of the discontinuity. Missed you at the CNC Workshop! Jon - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Homing and f-error
Hi Chris I guess I missed it also. Saw the later post where there was a suggested work around since this is a Mesa card setup. It would be possible to add a normally closed real relay contact between the limit prox and the drive's directional inhibit input that, when pulled by the homing pin in HAL would disconnect that input. I'll put my flame suit on after that suggestion. Ray On Wed, 13 Jun 2007 10:40:48 -0500, Enhanced Machine Controller (EMC) emc-users@lists.sourceforge.net said: On Wed, Jun 13, 2007 at 08:48:43AM -0500, Ray Henry wrote: It could also be me reading it wrong. I had an issue trying to home using a shared limit pin. The one axis I was homing was fine but all the others would fault. No way around it after the ignore limit code was made single axis. Long ago we found that motion stuttered during homing. Some sort of pause in the motion while EMC looked at the home pin. If this is still the case, it might cause the increased following error that was observed. Ray Ray I missed the last paragraph of Richard's message at the first reading, but down there he says that his switches disable the servo amp during the homing process, which causes a larger following error. The motion (decelerating to a stop after touching the switch) is commanded smoothly, but his machine can't track it well, because the amp is off, and the axis is coasting. Chris - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Homing and f-error
On Tue, Jun 12, 2007 at 05:51:37PM +0100, Richard Arthur wrote: I have a problem with f-error being larger than normal when homing. ... Am I correct in thinking the larger than nomal f-error is caused by the servo amp being disabled by the the limit switch during the homing sequence? Yes I'm sure you're right What can you suggest to overcome this so that I can run it with minimum FERROR and MIN_FERROR settings? You have to arrange not to do that. A separate home switch seems like the best answer, but it's probably not the easiest. (I'm not sure what the easiest would be). I say the separate switch is best, because any other solution would involve sometimes disabling your safety measures on the limit switches. Chris - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Homing and f-error
Ray, I don't think it's a matter of development thinking. If he hits teh limit switch and that physically disables his drives then it's not soemtihng you can overcome by software, or maybe am I reading it all wrong? Regards, Alex Wah. We really need to get some sort of order in our development thinking here. Sure professional machines use three switches at one end of travel and two at the other but that does not obviate the need for lesser systems for us little light machine guys. Note: Indepentantly of EMC2, the limit switches disable the servo amps in the current direction of travel only. - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Homing and f-error
It could also be me reading it wrong. I had an issue trying to home using a shared limit pin. The one axis I was homing was fine but all the others would fault. No way around it after the ignore limit code was made single axis. Long ago we found that motion stuttered during homing. Some sort of pause in the motion while EMC looked at the home pin. If this is still the case, it might cause the increased following error that was observed. Ray On Wed, 2007-06-13 at 16:30 +0300, Alex Joni wrote: Ray, I don't think it's a matter of development thinking. If he hits teh limit switch and that physically disables his drives then it's not soemtihng you can overcome by software, or maybe am I reading it all wrong? Regards, Alex Wah. We really need to get some sort of order in our development thinking here. Sure professional machines use three switches at one end of travel and two at the other but that does not obviate the need for lesser systems for us little light machine guys. Note: Indepentantly of EMC2, the limit switches disable the servo amps in the current direction of travel only. - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Homing and f-error
Wah. We really need to get some sort of order in our development thinking here. Sure professional machines use three switches at one end of travel and two at the other but that does not obviate the need for lesser systems for us little light machine guys. The fact that some folk would like to stack all of the limits on a single pin and then try to home from any one of them is not an error, it is an integrator decision that should/must be allowed in the software. Some of the recent changes to HAL and Motion have made it very difficult to use EMC2 in these circumstances. IMO we are getting close to a critical lack of proper flexible task planning. Ray On Wed, 2007-06-13 at 08:07 -0500, Chris Radek wrote: On Tue, Jun 12, 2007 at 05:51:37PM +0100, Richard Arthur wrote: I have a problem with f-error being larger than normal when homing. ... Am I correct in thinking the larger than nomal f-error is caused by the servo amp being disabled by the the limit switch during the homing sequence? Yes I'm sure you're right What can you suggest to overcome this so that I can run it with minimum FERROR and MIN_FERROR settings? You have to arrange not to do that. A separate home switch seems like the best answer, but it's probably not the easiest. (I'm not sure what the easiest would be). I say the separate switch is best, because any other solution would involve sometimes disabling your safety measures on the limit switches. Chris - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Homing and f-error
On Wed, Jun 13, 2007 at 08:48:43AM -0500, Ray Henry wrote: It could also be me reading it wrong. I had an issue trying to home using a shared limit pin. The one axis I was homing was fine but all the others would fault. No way around it after the ignore limit code was made single axis. Long ago we found that motion stuttered during homing. Some sort of pause in the motion while EMC looked at the home pin. If this is still the case, it might cause the increased following error that was observed. Ray Ray I missed the last paragraph of Richard's message at the first reading, but down there he says that his switches disable the servo amp during the homing process, which causes a larger following error. The motion (decelerating to a stop after touching the switch) is commanded smoothly, but his machine can't track it well, because the amp is off, and the axis is coasting. Chris - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Homing and f-error
- Original Message - From: Chris Radek [EMAIL PROTECTED] To: Enhanced Machine Controller (EMC) emc-users@lists.sourceforge.net Sent: Wednesday, June 13, 2007 8:07 AM Subject: Re: [Emc-users] Homing and f-error You have to arrange not to do that. A separate home switch seems like the best answer, but it's probably not the easiest. (I'm not sure what the easiest would be). I say the separate switch is best, because any other solution would involve sometimes disabling your safety measures on the limit switches. Chris I had an idea I plan to try on my machine when I get to that point. Have a push button that jumps limit switches only while held in. My machine had motor and belt guards which means I have to dis-assemble the machine or hotwire to enable the amps to jog it off the limit switch. My purpose was to be able to jog off limit switches but it could also be held while homing from limit switches. Roger - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
[Emc-users] Homing and f-error
I have a problem with f-error being larger than normal when homing. I ran this program : % G00 G40 G49 G80 G90 G54 G20 G61 G00 X0.2 G01 X0.1 F6. X0.2 X0.1 X0.2 X0.1 X0.2 M02 % and took a screenshot of HAL Oscilloscope to demonstrate normal f-error: http://imagebin.org/8953 This next screenshot shows the f-error when homing: http://imagebin.org/8954 My mill set-up: EMC 2.1.6 3 axis Servos 2 Hall effect limit switches/axis (no home switch) Note: Indepentantly of EMC2, the limit switches disable the servo amps in the current direction of travel only. From m5i20 .ini: # First axis [AXIS_0] TYPE = LINEAR HOME = 0.0 MAX_VELOCITY = 4.0 MAX_ACCELERATION = 20.0 BACKLASH = 0.000 INPUT_SCALE = 2 OUTPUT_SCALE = 1 OUTPUT_OFFSET = 0 MIN_LIMIT = -32.0 MAX_LIMIT = 20.0 FERROR =0.02 MIN_FERROR = 0.02 HOME_OFFSET = -0.1 HOME_SEARCH_VEL = 0.1 HOME_LATCH_VEL =0.01 HOME_USE_INDEX =FALSE HOME_IGNORE_LIMITS =YES MAX_OUTPUT =10 # PID tuning params DEADBAND = 0.15 P = 200 I = 0 D = 0 FF0 = 0 FF1 = 1 BIAS = 0 The problem exists on all axis. f-error reduces as HOME_SEARCH_VEL and HOME_LATCH_VEL decrease. Am I correct in thinking the larger than nomal f-error is caused by the servo amp being disabled by the the limit switch during the homing sequence? What can you suggest to overcome this so that I can run it with minimum FERROR and MIN_FERROR settings? Thank you. Richard. - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
