#emc-devel | Logs for 2008-07-26

[01:25:32] <steves_logging> steves_logging is now known as steve_stallings
[02:59:40] <cradek> rayh: I could use some advice about my Z axis. It still oscillates. when I null the vel input it still does it, so it's not the position loop oscillating - it's in the vel loop/amp
[02:59:56] <cradek> the tach output looks nice and clean.
[03:00:23] <cradek> I can damp the oscillation with an allen wrench on the end of the screw and it easily stops. but when I move, it starts up again.
[03:01:58] <cradek> turning the gain control up and down doesn't make it better or worse.
[03:03:06] <cradek> I replaced two nasty looking electrolytics on the amp and that didn't change anything.
[03:14:18] <steve_stallings> steve_stallings is now known as steves_logging
[05:25:14] <CIA-35> EMC: 03cmorley 07TRUNK * 10emc2/src/hal/classicladder/module_hal.c: How did I miss hal_exit(compId) in exit code?
[13:43:03] <rayh> Morning cradek.
[14:01:03] <cradek> good morning
[14:03:26] <BigJohnT> morning rayh and cradek
[14:06:26] <cradek> huh my simple enco order is coming in four separate shipments. it's good I found the 'free shipping' code.
[14:08:46] <CIA-35> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/Master_Developer.lyx: Added more NML messages 40 down 112 to go...
[14:08:54] <CIA-35> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/code/NML_Messages.lyx: Added more NML messages 40 down 112 to go...
[14:08:59] <BigJohnT> what are you getting
[14:10:19] <cradek> oh just various supplies. It started with me needing a gallon of way oil, and ended up costing $150, you know how it is
[14:10:46] <BigJohnT> yep
[14:10:48] <cradek> one roughing end mill is coming in its own box, so is one box of shop towels
[14:11:35] <BigJohnT> I've been getting my end mills from Lakeshore carbide and like them
[14:11:40] <rayh> Hi John
[14:11:53] <BigJohnT> Hi Ray
[14:12:09] <rayh> Keep up the good work on those NML messages. I need em.
[14:12:34] <BigJohnT> wish there was an easy way to get them from the spreadsheet to lyx
[14:12:50] <rayh> cradek, I've known those little pots on that drive board to not make good contact.
[14:13:01] <cradek> I still have not needed a solid carbide over 1/4". they are expensive.
[14:13:18] <BigJohnT> does the format look ok on the NML messages
[14:13:30] <rayh> I've had to turn them back and forth a few times to get them to settle down. Seems the coolant makes a film on the wiper.
[14:13:45] <BigJohnT> I use solid carbide for all even aluminum
[14:13:47] <rayh> Great format.
[14:13:51] <BigJohnT> thanks
[14:14:36] <cradek> rayh: ok I'll meter it -- but I have turned them a lot.
[14:14:50] <rayh> Then that may not be it.
[14:15:02] <cradek> the offset one definitely works
[14:15:29] <cradek> and the one at the bottom of the amp [unmarked] also does something (makes the oscillation change but not get better)
[14:15:32] <rayh> Good. That is a step in the right direction.
[14:16:20] <rayh> It's good that tach feedback is clean.
[14:16:33] <BigJohnT> I use this one with aluminum dry and it seems to last forever. http://www.lakeshorecarbide.com/index.asp?PageAction=VIEWPROD&ProdID=59
[14:16:35] <cradek> I unhooked the vel command signal and nulled the offset so it didn't move, then moved it with a 1.5v battery both directions
[14:16:50] <cradek> moving with the battery had the same grumbling oscillation
[14:17:39] <rayh> I looked but couldn't find any of the manuals for that drive.
[14:17:58] <cradek> darn, thanks for trying
[14:20:02] <rayh> The tach v command comparator is right close to those pots.
[14:20:54] <rayh> A single transistor if I remember.
[14:21:24] <cradek> there are carbon comp resistors nearby too that I should check
[14:22:02] <cradek> also I have not done an obvious thing - swap the amps - just because it's a lot of screws and stuff. I should do that.
[14:25:11] <cradek> BigJohnT: do you use that with cold air? or just nothing?
[14:26:13] <cradek> holy cow, they recommend 1800 sfm
[14:27:25] <cradek> no way can I do that on my mill except with a VERY big end mill
[14:28:03] <cradek> haha I would need 1.75"
[14:28:36] <BigJohnT> cradek: just compressed air
[14:29:48] <BigJohnT> I rigged up a small regulator to some loc-line with a nozzle and it works real well
[14:31:20] <BigJohnT> the cool thing about that series of endmill is is handles deep pocket corners without chatter and you can run it as fast as you dare...
[14:31:59] <BigJohnT> and with a corner radius it lasts 10 times longer than other carbides
[14:32:09] <skunkworks> cradek: checked the brushes?
[14:32:19] <skunkworks> * skunkworks bets the servo is full of carbon
[14:32:51] <cradek> no I have not done anything at all to the motor
[14:33:35] <skunkworks> do they look like this? http://www.electronicsam.com/images/KandT/servostart/smallservo.JPG
[14:33:54] <skunkworks> I think they where hardinge..
[14:34:18] <cradek> yes I think so
[14:34:20] <skunkworks> they where packed full of carbon. I was impressed ;)
[14:34:37] <cradek> and thanks for the photo, which tells me where the brushes are hidden :-)
[14:35:15] <cradek> any hints for opening and cleaning it?
[14:44:17] <rayh> Those motors were made by ElectraCraft in Minnesota. Bought out by
[14:44:24] <rayh> Reliance
[14:45:11] <rayh> Don't just pull them apart.
[14:45:18] <rayh> You'll weaken the magnets.
[14:45:37] <skunkworks> I think we just pulled the endcap off
[14:45:51] <rayh> We used a hunk of hollow steel tubing about the same size as the armature.
[14:45:54] <skunkworks> although ithas been a while.
[14:46:42] <rayh> If the carbon gets real bad it will short out the brush holder and then blow some transistors and resistors under the heat sink on the drive.
[14:47:27] <cradek> yikes
[14:47:42] <rayh> Oh they are pretty easy to replace.
[14:48:23] <cradek> I had some trouble with the caps - with the magic disintegrating 70s circuit board traces
[14:48:53] <rayh> Looks like I did replace 4 transistors on the board I've got in my hands here.
[14:49:08] <cradek> this one has a bunch of rework on the back. it may have blown up once before.
[14:49:29] <rayh> SK3260 is the replacement I used.
[14:49:50] <cradek> logger_dev: hear that? remember for me please.
[14:49:50] <cradek> I'm logging. I don't understand 'hear that? remember for me please.', cradek. Try /msg logger_dev help
[14:50:34] <cradek> rayh: want to stop over for coffee and have a look? :-)
[14:51:20] <rayh> You bet. Be right there in 14 hours.
[14:51:44] <cradek> I'll fire up my transporter
[14:51:59] <cradek> (it's loud so I don't let it run all the time)
[14:52:45] <rayh> Okay. "Now would be a good time, Scotty"
[14:54:08] <rayh> cradek, Does the growl persist at all speeds?
[14:55:12] <cradek> if I stop the oscillation with the wrench I can sometimes jog at very slow speeds and it won't start oscillating. but once it starts it seems to persist at all speeds, even stopped
[14:55:45] <rayh> Do the repairs to the board traces seem like they are recent or are they covered with crud?
[14:56:00] <cradek> they are pretty old
[14:56:27] <rayh> So it seems like it worked a while after those repairs.
[14:56:35] <cradek> I think so
[14:56:58] <cradek> he said he was using it 3 months ago. I have a hard time believing that but he seemed honest.
[14:57:10] <cradek> I suppose he may have been using it with the oscillation.
[14:57:42] <rayh> Want me to send you one?
[14:58:36] <cradek> let me make sure it is the amp first. I will troubleshoot some more including swapping the two amps.
[14:58:51] <cradek> if it is the amp and I can't seem to fix it, I'd appreciate that a whole lot.
[14:59:23] <rayh> Sure. Don't expect good behavior without retune after the swap
[14:59:45] <rayh> No two the same sort of thing.
[14:59:48] <cradek> ok
[15:00:02] <cradek> do you know what the third pot near the bottom is?
[15:00:07] <cradek> mine aren't marked
[15:00:37] <cradek> the one I'm calling gain is marked "max speed" and the offset one is easy to identify by its behavior. the third at the bottom is a mystery
[15:00:42] <rayh> I used to know but...
[15:01:08] <rayh> Let me look brb
[15:03:23] <rayh> not a clue
[15:03:59] <cradek> ok thanks
[15:04:14] <cradek> I will mark it first if I get the urge to turn it on the working one
[15:04:27] <rayh> Might be a motor current feedback adjustment.
[15:05:06] <rayh> Although a lot of that scaling was done with the resisitors and caps on those solder posts
[15:05:45] <rayh> On this board, that bottom pot was glued to a specific setting.
[15:05:57] <cradek> one of those is 2.2Mohm carbon comp - I should desolder one end and check it.
[15:06:51] <cradek> oh hmm, mine isn't glued, and I've turned it all different ways.
[15:08:34] <rayh> Okay.
[15:09:31] <rayh> You might pull the heat sink and look at the current sense resistors.
[15:10:01] <cradek> ok good idea. I could scope them and maybe see a motor brush problem too
[15:10:28] <rayh> If it was blown up and they replaced some with less accurate it might cause something like what you are seeing.
[15:10:57] <rayh> They are 0.1 ohm precision
[15:11:10] <cradek> ok
[15:16:05] <cradek> thanks again ray, I'll let you know what I come up with. I better get going - lots to do this weekend!
[15:16:19] <rayh> Good luck.
[21:18:07] <rayh> cradek P3, the pot at the bottom is current loop balance. I found a set of the papers for that. Also found a description of the tool holders that Hardinge used. Send me an email with your addy and I'll get them out to you.
[21:28:08] <cradek> cool! thanks.
[21:28:46] <cradek> I might be on to a problem. the tach isn't as clean as I thought. when I ground the vel and tach inputs, and move the motor by changing the input offset, it moves smoothly
[21:29:10] <cradek> so I think it's a tach problem. I'm trying to figure out whether I should/can open up the tach
[21:30:39] <skunkworks_> * skunkworks_ hates tachs
[21:31:50] <cradek> shhhhhh it will hear you!
[21:32:46] <skunkworks_> heh - the old k&t had tachs for the velosity loop. They caused a few problems.
[21:37:03] <rayh> Ah. I've rebuilt them. When your reassemble them you'll need to chuck them up in your mill and adjust the back brush holder so you get the same voltage both ways.
[21:39:53] <cradek> ok so I should take these two screws out? :-)
[21:40:41] <skunkworks_> a lot of them also have holes to place pins so the brushes are retracted for easy assembly.
[21:40:50] <skunkworks_> (if you can get to them from the outside)
[21:42:14] <cradek> cool, the brushes face forward
[21:42:28] <cradek> can I safely pull the armature out?
[21:44:21] <rayh> Yes I believe so
[21:45:10] <rayh> I've got several of these as well. They may be for a bit later version of the feedback but you could swap brushes or the front end.
[21:45:41] <skunkworks_> so - the commutator is like a disk?
[21:45:42] <rayh> I'd make a scratch mark between the end and the sides.
[21:45:50] <rayh> Exactly
[21:46:01] <skunkworks_> don't know if I have seen that.
[22:56:24] <CIA-35> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/code/NML_Messages.lyx: WOW they are all done! NML messages galore
[23:00:10] <BigJohnT> when do you guys plan on putting 2.2.6 out?
[23:02:16] <CIA-35> EMC: 03bigjohnt 07TRUNK * 10emc2/debian/extras-Ubuntu-5.10/usr/share/applications/emc2-integratormanual.desktop: changed comment for integrator manual
[23:02:38] <CIA-35> EMC: 03bigjohnt 07TRUNK * 10emc2/debian/extras-Ubuntu-6.06/usr/share/applications/emc2-integratormanual.desktop: changed comment for integrator manual
[23:02:39] <CIA-35> EMC: 03bigjohnt 07TRUNK * 10emc2/debian/extras-Ubuntu-7.10/usr/share/applications/emc2-integratormanual.desktop: changed comment for integrator manual
[23:02:40] <CIA-35> EMC: 03bigjohnt 07TRUNK * 10emc2/debian/extras-Ubuntu-8.04/usr/share/applications/emc2-integratormanual.desktop: changed comment for integrator manual
[23:02:43] <CIA-35> EMC: 03bigjohnt 07TRUNK * 10emc2/debian/extras-sim-Ubuntu-5.10/usr/share/applications/emc2-integratormanual.desktop: changed comment for integrator manual
[23:02:46] <CIA-35> EMC: 03bigjohnt 07TRUNK * 10emc2/debian/extras-sim-Ubuntu-6.06/usr/share/applications/emc2-integratormanual.desktop: changed comment for integrator manual
[23:08:44] <BigJohnT> time to chill