[CreativeOffroad]

Hello all.

I am having a problem with my Mitek bed mill. My setup is listed below...

One of the servos was damaged in transport. The servo was a Glentek servo (GM4050-38). I was unable to locate a reasonably priced replacement, but did find an SEM MTS30U-42 to replace it. When I ordered the servo, I had them wire it the same as the Glentek servo per documents from Glentek.

When I hooked the servo up and started the machine, the servo ran uncontrolled until the controller errored out (I haven't mounted it yet).
I called tech support for the servo, and they told me to reverse the tach leads. I did, and after adjusting the balance on the servo amp, got the servo to stay still.

The problem now, is that the servo won't move. The controller shows the Z axis coordinates changing, but the servo isn't turning. It does register movement when I turn it by hand.

Any ideas? I know the BEMF (?) is different than the original servo, but thought I could adjust for it with the driver?

Thanks,
Scott

Mitek VBM50-F Bed Mill
Acramatic 2100 controller
Glentek SMA7115 Servo Amps
Glentek GM4050-38 Servos on X and Y axis'
SEM MTS30U-42

[Al_The_Man]

First power up normally and if the power is on the servo's, try turning the shaft of the motor, if it resists you, then the controller has command of the servo and means CNC is inputing a ±10vdc signal to the drive.
If you can turn the shaft, then measure the ±10vdc signal which should vary when you turn the motor shaft, this means the CNC is trying to command the drive but it is not responding.
Make sure the drive enable is on.
To trouble shoot the drive without the CNC controller, you can remove the ±10v signal and feed it with a 1.5~12v battery box.
Also to check the drive if all are the same size, temporarily swop two of the drives over.
Al.

[roseng]

also check to see if the current limiting pot on the drive is backed all the way off (counter clockwise) if so, apply a small drive signal, then slowly turn current limiting pot clockwise until motor turns.

Hope this Helps

[CreativeOffroad]

Thanks! I will try that. I cannot turn the servo shaft right now with the controller powered, but the motor will not turn by the controls.

I will try your tips, and post back this evening.

Thanks again,
Scott

[roseng]

Thanks! I will try that. I cannot turn the servo shaft right now with the controller powered, but the motor will not turn by the controls.

I will try your tips, and post back this evening.

Thanks again,
Scott

If you can't turn the shaft, then that rules out my current limiter theory.
have you tried reversing the analog speed command wires ?
if it uses a single ended command voltage, that would do it

[Al_The_Man]

have you tried reversing the analog speed command wires ?
if it uses a single ended command voltage, that would do it

That usually causes a run away, in my experience.
If the controller is holding the motor in position, it appear the analogue signal is able to control the servo in position, but it appears there is no analogue move signal coming out.
Al.

[CreativeOffroad]

When I test across the two leads coming from the Acramatic to the servo amp, I get -10.35V on the Z axis (even witht the E-Stop pressed), and just a couple mv on the X and Y. It doesn't change with machine power applied either. Does that sound right?

[Al_The_Man]

Thats strange, with that type of input the motor should be running flat out.
Did you try swopping the drives to see if it transfers to another axis?
Al.

[CreativeOffroad]

It did run away until I stopped it with the balance pot. It does move to the Y axis when I swap. Sound like a problem with the Acramatic on the Z axis? Should I move the Z to the 4 axis? Is that an option?

[Al_The_Man]

I would trouble shoot it with the Control disconnected. If you disconnect the 10v analogue signal and jumper the two pins out you should be able to zero the drive with the balance pot, You should at this point find resistance to the shaft without runaway.
Once you have it stationary, apply 1.5v to the analogue in and you should be able to stop it by removing the battery and short the terminals together, do the same with the reverse polarity, if it runs continous in one direction, then swop the tach leads and try again.
You have to get this loop right before the CNC Loop.
Al.

[CreativeOffroad]

I get the -10.35 volts from controller even with it disconnected from the servo drive.

[Al_The_Man]

I get the -10.35 volts from controller even with it disconnected from the servo drive.

That is because the motor/drive is not moving under the control of the analogue signal to move to zero (null) output.
Al.

[CreativeOffroad]

That is because the motor/drive is not moving under the control of the analogue signal to move to zero (null) output.
Al.

Is there a way to force/remedy that? Please forgive my lack of knowledge about how this works - I am learning as I go...

[CreativeOffroad]

Should it be in that state as soon as it boots up?

[Al_The_Man]

The servo should be under control as soon as the controller has booted up and the drive supply is present and the drives enabled, (e-stop out).
If all this present and the analogue signal is there the servo should move.
If not, you generally have a situation where either the motor can be turned by hand (no control) or it will run away. I use a +-10v battery box for testing that allows a 0v-to 10v variable signal to the analogue in.
Al.

[CreativeOffroad]

For what it's worth, the 10.35V goes away when I unplug from the Acramatic and plug the loopback plug in...

So I should disconnect the two wires from the controller, and connect the battery box to test?

Is there a chance the problem lies with the wiring of the motor?

[Al_The_Man]

Yes

So I should disconnect the two wires from the controller, and connect the battery box to test?

Is there a chance the problem lies with the wiring of the motor?

Sounds like it.
Al.

[CreativeOffroad]

Here is my latest results (I will make a battery box tomorrow for further testing).

I took the belt off of the X axis to prevent mechanical damage, and hooked it up to the Z axis amp. It behaved the same as the new motor. I

I then looped back the Z axis to supress the errors when it is unhooked and hooked the new motor up to the X axis connector with everything else back the way it belonged. The new motor seemed to work correctly, and be controlled fine by the X axis controls. (It does look like it will need some tweaking to match the controller perfectly).

Does this seem to indicate a bad servo amp. Like I said, I will test it more once I make my battery box tomorrow.

Thanks again for all of the help here. I am not sure what I would do without it...

[CreativeOffroad]

From tonight's testing:

Unhooked signal output from servo drive. Attached Voltmeter to it.

a. Voltage stays at 0V until about step 2 of the Acramatic software loading, then it goes to 10.35V. It doesn't change when I release the E-Stop and apply power. It stays at 10.35V(even when I push the up or down buttons on the control for the Z axis). Applying voltage to the servo amp at this point does nothing.

b. If I go into axis setup in the Acramatic, disable the Z axis (uncheck the "Axis Present" box), and restart the system, the signal voltage stay 0V the entire time (even after applying machine power). Applying voltage to the servo amp now makes the servo move.

Again, the servo itself acts fine if I plug it into the connector (at the table) for the X axis. I can't find anything in the setup that looks different for the two....

[CreativeOffroad]

Fixed the initial issue. Feel pretty silly working on that much hardware and finding it to be a software issue.

Turn out the SEM Servo has a Heidenhain encoder on it. This flagged with me (accidentally of course) when I was looking through the axis setup backup file I saved to a floppy disk, and saw a line with "fdbk_heidenhain=0" in it. I took a chance, and changed the "0" to a "1", rebooted, and the controller worked fine - spun the servo just like it should (for a while).

After a few minutes of moving the servo through the controller, it started "oscillating" (for lack of a better term). It is moving about a 1/4 turn each direction staying centered in the same place. The center point does move if I press the controls either direction.

Any ideas for this problem? It worked flawlessly for a few minutes - I'm baffled once again...