Thanks for the help, guys. 20,000 counts per revolution is in fact pretty high for an encoder. The "latest thing" however is the new sin/cos encoders which output counts in the millions. http://www.engineeringtalk.com/news/enc/enc109.html I originally thought the video card might be the problem source so I replaced it last week (and lowered the period to .0001) but the problem still remains. My next step is to make an attempt at recompiling without the debounce code. I have *some programming experience but a duffer's guide on this would be great! Really, any instructions or recommended reading you could suggest would be appreciated. Regards, Craig *-----Original Message----- *From: [email protected] [mailto:bdi4emc-help- *[email protected]] On Behalf Of bdi4emc-help- *[email protected] *Sent: Friday, November 24, 2006 3:00 AM *To: [email protected] *Subject: Bdi4emc-help Digest, Vol 10, Issue 4 * *Send Bdi4emc-help mailing list submissions to * [email protected] * *To subscribe or unsubscribe via the World Wide Web, visit * https://lists.ourproject.org/cgi-bin/mailman/listinfo/bdi4emc-help *or, via email, send a message with subject or body 'help' to * [email protected] * *You can reach the person managing the list at * [email protected] * *When replying, please edit your Subject line so it is more specific *than "Re: Contents of Bdi4emc-help digest..." * * *Today's Topics: * * 1. Re: X-Axis Display Problem (David Szakovits) * 2. Re: X-Axis Display Problem (Paul) * * *---------------------------------------------------------------------- * *Message: 1 *Date: Thu, 23 Nov 2006 11:43:13 -0500 *From: David Szakovits <[email protected]> *Subject: Re: [Bdi4emc-help] X-Axis Display Problem *To: "Help and advice for BDI-4 users." * <[email protected]> *Message-ID: <[email protected]> *Content-Type: text/plain; charset=ISO-8859-1; format=flowed * *i am not sure that i am not seeing the same sort of thing. i definitely *get some freezing. i have been warned about the on board video that i *have and i am going to buy an AGP card to see if that might fix the *symptom but being a emc-novice, i do not know what to expect or if it *will fix the problem or if at all related. *djs * *Craig Gadd wrote: * *>Hello Paul, *> *>I'm having a problem with my BDI setup and was hoping that you might have *>some advice on a problem I have. I'm running BDI 4.51 with a Motenc Lite *>Card installed. The card is connected through a 50 pin ribbon cable to a *>Vital Systems breakout board which in turn has been wired to a single *>encoder which has 20,000 quad/cts per revolution). *> *>When I launch the generic.run, or emc.run things seem to behave normally *>with the units showing for the axis which has the encoder on it. The *units *>will update if I turn the motor slowly, and everything seems to work. *When *>I spin the motor quickly however, the unit display of the axis seems to *>"freeze" and will not recover. The rest of the interface continues to *>function properly though. Any ideas on what might cause this and how I *>might fix it? I've tried Tkemc,Xemc, and mini but they all seem to behave *>the same. *> *>Thanks, *> *>Craig *> *> *> *>_______________________________________________ *>Bdi4emc-help mailing list *>[email protected] *>https://lists.ourproject.org/cgi-bin/mailman/listinfo/bdi4emc-help *> *> *> * * * *------------------------------ * *Message: 2 *Date: Thu, 23 Nov 2006 17:51:41 +0000 *From: Paul <[email protected]> *Subject: Re: [Bdi4emc-help] X-Axis Display Problem *To: [email protected] *Message-ID: <[email protected]> *Content-Type: text/plain; charset="iso-8859-1" * * *Hi Craig * *On Thursday 23 November 2006 06:52, Craig Gadd wrote: *> The card is connected through a 50 pin ribbon cable to a *> Vital Systems breakout board which in turn has been wired to a single *> encoder which has 20,000 quad/cts per revolution). * *That is a pretty high resolution encoder ! * *> When I launch the generic.run, or emc.run things seem to behave normally *> with the units showing for the axis which has the encoder on it. The *units *> will update if I turn the motor slowly, and everything seems to work. *When *> I spin the motor quickly however, the unit display of the axis seems to *> "freeze" and will not recover. The rest of the interface continues to *> function properly though. Any ideas on what might cause this and how I *> might fix it? * *Many years ago, Fred Proctor (as I recall) was having problems with erratic *encoder counts and added a routine to "debounce" the input - This is hard *coded with an arbitrary limit of 10 times the max velocity. When the rate *of *change per servo cycle ecceeds this limit, any further input is ignored and *a *following error *should be triggered. * *With a 20,000 count/rev encoder and a moderate max velocity, I would *imagine *it would be quite easy to trigger the debounce routine - When you come to *connect the servo amps up and tune the system, as long as you don't go over *the top on the acceleration settings, you shouldn't experience any *problems.. *If you do, we can disable the debounce routine easily enough by *adding -DSIMULATED_MOTORS to the RTFLAGS of src/Makefile.inc.in and *recompiling vitalmod.ko * *If you need a step by step duffer's guide to setting up a build environment *and recompiling the kernel modules, just ask. * * *BTW With any servo system there is nothing to be gained by setting the *BASE_PERIOD parameter anywhere near what you might use with a stepper *system. *0.00005 or 0.0001 should be fine. * * *Regards, Paul. * * * *------------------------------ * *_______________________________________________ *Bdi4emc-help mailing list *[email protected] *https://lists.ourproject.org/cgi-bin/mailman/listinfo/bdi4emc-help * * *End of Bdi4emc-help Digest, Vol 10, Issue 4 ********************************************
On Sunday 26 November 2006 11:37, [email protected] wrote:
Thanks for the help, guys.
20,000 counts per revolution is in fact pretty high for an encoder. The "latest thing" however is the new sin/cos encoders which output counts in the millions.
Just one comment here. Nice, but where do we hide, and get power for, the Cray SX1 it will take to do the a/d conversions and transcendentals of this things output in anything even remotely resembling real time? Particularly at its full speed of 12,000 rpm. Methinks this is a solution looking for a problem.
I originally thought the video card might be the problem source so I replaced it last week (and lowered the period to .0001) but the problem still remains.
My next step is to make an attempt at recompiling without the debounce code. I have *some programming experience but a duffer's guide on this would be great! Really, any instructions or recommended reading you could suggest would be appreciated.
Regards,
Craig
*-----Original Message----- *From: [email protected] [mailto:bdi4emc-help- *[email protected]] On Behalf Of bdi4emc-help- *[email protected] *Sent: Friday, November 24, 2006 3:00 AM *To: [email protected] *Subject: Bdi4emc-help Digest, Vol 10, Issue 4 * *Send Bdi4emc-help mailing list submissions to * [email protected] * *To subscribe or unsubscribe via the World Wide Web, visit * https://lists.ourproject.org/cgi-bin/mailman/listinfo/bdi4emc-help *or, via email, send a message with subject or body 'help' to * [email protected] * *You can reach the person managing the list at * [email protected] * *When replying, please edit your Subject line so it is more specific *than "Re: Contents of Bdi4emc-help digest..." * * *Today's Topics: * * 1. Re: X-Axis Display Problem (David Szakovits) * 2. Re: X-Axis Display Problem (Paul) * * *---------------------------------------------------------------------- * *Message: 1 *Date: Thu, 23 Nov 2006 11:43:13 -0500 *From: David Szakovits <[email protected]> *Subject: Re: [Bdi4emc-help] X-Axis Display Problem *To: "Help and advice for BDI-4 users." * <[email protected]> *Message-ID: <[email protected]> *Content-Type: text/plain; charset=ISO-8859-1; format=flowed * *i am not sure that i am not seeing the same sort of thing. i definitely *get some freezing. i have been warned about the on board video that i *have and i am going to buy an AGP card to see if that might fix the *symptom but being a emc-novice, i do not know what to expect or if it *will fix the problem or if at all related. *djs * *Craig Gadd wrote: * *>Hello Paul, *> *>I'm having a problem with my BDI setup and was hoping that you might have *>some advice on a problem I have. I'm running BDI 4.51 with a Motenc Lite *>Card installed. The card is connected through a 50 pin ribbon cable to a *>Vital Systems breakout board which in turn has been wired to a single *>encoder which has 20,000 quad/cts per revolution). *> *>When I launch the generic.run, or emc.run things seem to behave normally *>with the units showing for the axis which has the encoder on it. The *units *>will update if I turn the motor slowly, and everything seems to work. *When *>I spin the motor quickly however, the unit display of the axis seems to *>"freeze" and will not recover. The rest of the interface continues to *>function properly though. Any ideas on what might cause this and how I *>might fix it? I've tried Tkemc,Xemc, and mini but they all seem to behave *>the same. *> *>Thanks, *> *>Craig *> *> *> *>_______________________________________________ *>Bdi4emc-help mailing list *>[email protected] *>https://lists.ourproject.org/cgi-bin/mailman/listinfo/bdi4emc-help *> *> *> * * * *------------------------------ * *Message: 2 *Date: Thu, 23 Nov 2006 17:51:41 +0000 *From: Paul <[email protected]> *Subject: Re: [Bdi4emc-help] X-Axis Display Problem *To: [email protected] *Message-ID: <[email protected]> *Content-Type: text/plain; charset="iso-8859-1" * * *Hi Craig * *On Thursday 23 November 2006 06:52, Craig Gadd wrote: *> The card is connected through a 50 pin ribbon cable to a *> Vital Systems breakout board which in turn has been wired to a single *> encoder which has 20,000 quad/cts per revolution). * *That is a pretty high resolution encoder ! * *> When I launch the generic.run, or emc.run things seem to behave normally *> with the units showing for the axis which has the encoder on it. The *units *> will update if I turn the motor slowly, and everything seems to work. *When *> I spin the motor quickly however, the unit display of the axis seems to *> "freeze" and will not recover. The rest of the interface continues to *> function properly though. Any ideas on what might cause this and how I *> might fix it? * *Many years ago, Fred Proctor (as I recall) was having problems with erratic *encoder counts and added a routine to "debounce" the input - This is hard *coded with an arbitrary limit of 10 times the max velocity. When the rate *of *change per servo cycle ecceeds this limit, any further input is ignored and *a *following error *should be triggered. * *With a 20,000 count/rev encoder and a moderate max velocity, I would *imagine *it would be quite easy to trigger the debounce routine - When you come to *connect the servo amps up and tune the system, as long as you don't go over *the top on the acceleration settings, you shouldn't experience any *problems.. *If you do, we can disable the debounce routine easily enough by *adding -DSIMULATED_MOTORS to the RTFLAGS of src/Makefile.inc.in and *recompiling vitalmod.ko * *If you need a step by step duffer's guide to setting up a build environment *and recompiling the kernel modules, just ask. * * *BTW With any servo system there is nothing to be gained by setting the *BASE_PERIOD parameter anywhere near what you might use with a stepper *system. *0.00005 or 0.0001 should be fine. * * *Regards, Paul. * * * *------------------------------ * *_______________________________________________ *Bdi4emc-help mailing list *[email protected] *https://lists.ourproject.org/cgi-bin/mailman/listinfo/bdi4emc-help * * *End of Bdi4emc-help Digest, Vol 10, Issue 4 ********************************************
_______________________________________________ Bdi4emc-help mailing list [email protected] https://lists.ourproject.org/cgi-bin/mailman/listinfo/bdi4emc-help
-- Cheers, Gene "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) Yahoo.com and AOL/TW attorneys please note, additions to the above message by Gene Heskett are: Copyright 2006 by Maurice Eugene Heskett, all rights reserved.
Hi Craig On Sunday 26 November 2006 16:37, [email protected] wrote:
20,000 counts per revolution is in fact pretty high for an encoder. The "latest thing" however is the new sin/cos encoders which output counts in the millions.
The sin/cos encoder referred to in the article only has 1024 "periods" per rev. The sine waves are then interpolated to produce the 4M discrete counts with the aid of a dedicated DSP - Presumably, correction tables are programmed in to reduce cyclic errors and phase variations.
I originally thought the video card might be the problem source so I replaced it last week (and lowered the period to .0001) but the problem still remains.
Unlikely unless it shares system RAM and/or uses proprietory closed source drivers
My next step is to make an attempt at recompiling without the debounce code. I have *some programming experience but a duffer's guide on this would be great! Really, any instructions or recommended reading you could suggest would be appreciated.
First stage is to install the tools and libraries using the command: apt-get build-dep emc Then "apt-get source emc" will download and extract the sources to the current directory. Run configure in src/ and add -DSIMULATED_MOTORS to the RTFLAGS line in src/Makefile.inc - A `make modules_install` as root will compile the low level module where the "position debounce" takes place - We don't need to worry about recompiling & installing any of the user space parts.. Regards, Paul.
participants (3)
-
elnerdo＠cox.net -
Gene Heskett -
Paul