On Sunday 07 February 2010, richard harris wrote: >The encoders on this machine are differential sealed units from BEI. When > i got the machine one was nonfunctional so i sent all three to BEI and had > them rebuilt and upgraded from an incandescent bulb to led. The encoder > is mounted to the rear of the motor, and shows no sign of any oil or > coolant contamination. Currently the encoders are wired as single ended, i > plan on ordering differential receiver from US digital on monday. All > wires are shielded and grounded at the controller side only. All cables > are run in flexible metal conduits. I checked all the grounds along with > all the mechanical couplers, nothing looked unacceptable. > >I wrote the code below to see if i could manage to duplicate the error > without scraping parts. G1 X.5 F30 >M3 S4000 >#1 = 0 >O101 while [#1 lt 150] >G1 X2.0 Y2.0 z-1.0 F40 >G1 X.5 Y0.0 Z-1.716 >#1 = [#1+1] >O102 if [#1 EQ 30] >S3000 >(MSG, 30 loops down) >O102 endif >O102 if [#1 EQ 60] >S2000 >(MSG, 60 loops down) >O102 endif >O102 if [#1 EQ 90] >S1500 >(MSG, 90 loops down) >O102 endif >O102 if [#1 EQ 120] >S1000 >(MSG, 140 loops down) >O102 endif >O101 endwhile >M5 >G1 X.05 Y0.0 Z-1.716 F30 >G1 X0.0 F10 >M2 >This code has more linear inches of motion and is longer than the program > that was exhibiting issues. I ran it three times without the coolant or > spindle running, i had two other machines running at the same time as i > did the other day. I also had halmeter showing the encoder count, not sure > if it would show anything of value but i was curious. The machine repeated > within the tolerance of the indicator.. I then ran it three times with the > coolant and spindle running and witnessed two thousands of motion lost > over the three runs. However Halmeter showed the same number of encoder > counts. I am assuming two things, 1 my problem is noise, 2 the error was > less during the test because i did not load the spindle. Do those sound > like reasonable assumptions?
Very much so. Guessing at a distance, I'd blame the spindle motors brushes & start waving a scope probe around in the vicinity of its power cable. I don't have encoders, entirely dependent on the steppers sstaying in time here. but I did do that and was pleasantly surprised that my motors power cable, which is not shielded in its run from the PWM controller I moved from the gear head into a box about 4 feet of cable away so it could be hooked directly to the PMDX-106 controlling it, and my 100mhz scope can't find either the PWM noise, or the brush noise from 4 or 5" away from that twisted cable. > My plan at the moment is get a differential > receiver on all the encoders, and run new wires on the x-axis just in > case. If these mods don't improve the situation I'm thinking of adding a > line reactor to the VFD. The idea I believe, is to add the ferrite stuff so it attacks the longitudinal noise, in other words both conductors going to the motor go through the ferrite core as if they were zip cord, and that zip cord was wound through the core several times. That slows down the unbalanced component of the noise, without seriously effecting the VFD's ability to read the motor & see what its doing. It also, because the huge majority of the currents are equal but opposite directions, the controlling currents balance each other out magnetically, so the core will not easily saturate and disappear from the circuit. > Any other straight forward things I can do to > improve the noise situations? Star grounding, when running through conductive conduit, often isn't what one thinks he has. The wires themselves should also be shielded according to function, and the only place the shields and the grounded conduits are actually tied together should be at the 'star' point itself, at the controller. The shielding should not be connected to the conduit any other place. If you are depending on the conduit for ground, problems such as this are a possibility. Good luck. >--- On Fri, 2/5/10, richard harris <[email protected]> wrote: > >From: richard harris <[email protected]> >Subject: [Emc-users] Directional motion loss >To: "Enhanced Machine Controller (EMC)" <[email protected]> >Date: Friday, February 5, 2010, 6:17 PM > >Hey, > >I am running emc 3.2.0 off the BDI with Jon Elson's pwm servo boards. The > machine is a knee mill, of similar size and mass to a hurco. > >This machine gets little use as its bigger brother does all the heavy > lifting so the problem i describe below might have existed for some time > and not been noticed based upon part configuration and tolerance. I notice > during the last few days, while machining a thin wall box that one > sidewall was thicker than the other. At first i assumed that i had a chip > in the vise when i loaded the part, however the next part had a similar > condition. I re-zeroed the machine and noted that it showed i was .010" > off. Ran the next part and it was fine, the second part showed some > issues, and the third clearly had a problem. Part is taking 17 minutes, > and the properties pull down shows 400" of motion. I checked the machine > for a loose belt, loose ball screw to thrust washer, and loose ball nut, > all of these looked fine. I continued to run the code, cutting air and > double checking the machine position to the zero point on the vise. I am > loosing .002-.005" to the in the minus x direction. I have checked for > backlash within the working envelope with a dial indicator and have not > found anything that would alarm me. Any ideas on what would causing a loss > of position in one direction only it is repeatably losing in the minus > direction only. Mechanical, electrical, anything at this point. As always > thanks. >Richard > > > > >--------------------------------------------------------------------------- >--- The Planet: dedicated and managed hosting, cloud storage, colocation > Stay online with enterprise data centers and the best network in the > business Choose flexible plans and management services without long-term > contracts Personal 24x7 support from experience hosting pros just a phone > call away. http://p.sf.net/sfu/theplanet-com >_______________________________________________ >Emc-users mailing list >[email protected] >https://lists.sourceforge.net/lists/listinfo/emc-users > > > > >--------------------------------------------------------------------------- >--- The Planet: dedicated and managed hosting, cloud storage, colocation > Stay online with enterprise data centers and the best network in the > business Choose flexible plans and management services without long-term > contracts Personal 24x7 support from experience hosting pros just a phone > call away. http://p.sf.net/sfu/theplanet-com >_______________________________________________ >Emc-users mailing list >[email protected] >https://lists.sourceforge.net/lists/listinfo/emc-users > -- 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) I put up my thumb... and it blotted out the planet Earth. -- Neil Armstrong ------------------------------------------------------------------------------ The Planet: dedicated and managed hosting, cloud storage, colocation Stay online with enterprise data centers and the best network in the business Choose flexible plans and management services without long-term contracts Personal 24x7 support from experience hosting pros just a phone call away. http://p.sf.net/sfu/theplanet-com _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
