Anilam 1100M Servo Bearings

Got the replacement servo in yesterday and it started out running away and following error shut down the drives. Today it will turn at normal speed for a bit then stop and following error shuts down the drives. The servo looks almost never used, I took a brush out and they look brand new... don't know what to try next.

JT

Remove the motor and tacho brushes, use compressed air on brush holes.
Do mind what brush goes where so you can put them exactly where they were and oriented properly.
Use a flashlight to see inside brush holes and check the state of the commutators, they can be mildly blacked out but not totally black.
- testing
-enable the drives possibly with no input, the motor should slowly spin and easily stopped by hand (depending on how big the motor is), spinning it by hand should not cause it to jump or speed up on it's own.
--if at any point during the above the motor jumps or spins very fast = tacho or motor polarity wrong, change one of them, not both and test again.
I would not advise to change anything on the servo drive (no DIP switches, no moving the pots) if you are sure it was working properly.

Jerry Bouvier (used to be Anilam tech support) got back with me today and said that either the tacho or armature wires are backwards or both. I need to cobble up a simple way to tell the 7i77 to output 1v and ignore following errors.

Hmm, if the armature wires are backward to the encoder direction I just need to reverse the analog out wires to make it go the right way...

JT

1V might be to much, a 4000RPM motor will be doing 400RPM !
set the jog velocity to something very slow and short press the jog buttons, keep upping the length, then up the jog velocity a bit, repeat.

Right now the drive faults out after a 1/4 turn so pondering getting an AC servo from Chinlee... need 3.5 NM of torque or more and 2400 rpm or more. I can move the replacement servo to the Y axis and same result. It does run fine when you send a DC voltage to it but dunno if the tacho is really a resolver or not.

JT

If it is always at 1/4, looks like positional error, not velocity.
Tacho = 2 wires
Resolver = 6 wires
Usually they have more, but those are the important ones.

Ok, only two wires so must be a tacho. I've ran it on the bench with 48vdc for a while just to make sure the brushes have been seated in. The encoder goes straight to the 7i77 only the two wires from the tacho go to the drive. Sometimes it will go a bit more and sometimes it faults out straight away. Not sure what to try next... oh I have the following error huge in LinuxCNC.

JT

Forget LinuxCNC for now, lets fix the velocity loop.
Does the drive fault or only LinuxCNC?
To check the drive/velocity loop, disconnect the analog wires from Mesa and use DVM set to MOhm first, if no movement set to KOhm, then Ohm, and if those do not work set it to Diode, that should move it but quite fast so only short touch the wires.
It should move smoothly so try reversing wires and it should move in the other direction.
If all that works, connect to 7i77, try to give it some tuning with FF1, pretty sure you can do that since i learned most of it from you and Andy. Thank you, again and again.
BTW did you try reversing the tacho wires ?

My DVM only has an Ohm setting, I was thinking of using my lab DC power supply to apply a little DC voltage to the drive and see what that does.

Yes I reversed the tacho wires and it ran away and LinuxCNC shut down with a following error.

I think the drive is faulting out because the red fault led is lit after I try to move.

With 48vdc on the armature wires I measured 9v on the tacho and -9v when I reverse the wires on the armature, I'm thinking I need to test the Y servo and see what it does.

JT

Original servo from the Y axis with 48vdc on the armature I get 8.5vdc on the tacho so the new one has too much voltage from the tacho... if I try and adjust the tach to anything other than 10k ohms the board will stay in fault.

The setup page for the drive says
adjust the SIG pot to obtain 4.5K ohms
adjust the TAC pot to obtain 10K ohms
adjust the COMP pot to obtain 2.2K ohms

There is a BAL pot if the axis drifts that needs to be adjusted to stop drift.
There is one more pot marked LIM, I'm not sure what that does.

I've attached the 3 pdf's I have for that system, a couple of pages each.

JT