Re: [CAD_CAM_EDM_DRO] Re: Looking for a servo amplifier.....

dougrasmussen@... wrote:

> Matt,
>
> Yes, I have the tuning info, etc. The problem is the BAL pot does
> not seem to be active, the drive won't balance. When the control is
> put in Servo adjust mode to tune the axis lag, it's impossible to get
> the value adjusted properly. The axis lag is much higher in one
> direction than the other, and impossible to adjust that difference
> out. This IM-balance causes in-position errors and occasionally a
> servo error during rapid, a while back I reset the max rapid down to
> 250 IPM to help eliminate errors.

1. Balance needs to be set with the signal input wires from the cnc
disconnected. This adjustment is meant to compensate for offset errors
in the amp itself, not the output of the Servo I/F board in the Delta
20. Disconnect the 15 pin Dsub connector from the Servo I/F board (has
three, 2 conductor grey cables coming out of it). You may need to go to
the setup screen and select the Test Mode that disables all axis
movement (my memory is a little sketchy here) to keep from going into
e-stop due to following error while you adjust things.

2. Since you have big size 42 motors with 7V/1000RPM tachs, I would
expect that a good starting point for adjustment would be:

BAL, CMP and SIG - Midrange (all the way CCW, then 10 turns CW).

TACH - 7k ohms from COMMON (TP1) to the TACH test point.

CL and RMS- Maximum (all the way CW, then 1turn or so CCW to make sure
the wiper of the pot isn't off the end of the pot).

With the signal input disconnected (as in 1 above) try to balance the
amp. If you can't get the motor shaft to stop turning, or at least slow
down to a bare creep, the amp has problems. If you can get the amp to
balance out, reconnect the signal inputs (and get out of the "stop
everything" test mode). Jog the axis back and forth slowly, the motion
should be smooth. Now turn the CMP pot one or two turns CW and jog again
to make sure the motion is still smooth. Keep going CW a turn or two at
a time until you get rough running or oscillation, but in any case I
wouldn't go more than 8 turns CW from the starting point if roughness or
oscillation never develops. With that done, try to increase the gain by
turning the SIG pot one or two turns CW at a time and then jog a little
to test. If the axis starts running rough, back off a turn or two CCW on
the CMP pot to smooth it out. Since you want to run at high speed. the
idea is to get the gain (SIG) up as high as you can (but don't run the
wiper off the end of the pot) while still retaining smooth motion. Don't
go CCW past the center point on the CMP pot because the center position
is the minimum setting. When all this is done, disconnect the signal
input again and rebalance as best you can. The Y axis is the hardest to
adjust because it has the most weight (on a knee mill) to throw around.

> I tuned per my book, results were not good. Tech suggested using the
> test points off a good axis (X) card and tuning bad axis (Y) to the
> same reading. That improved axis performance a bit.
>
> This is the Y axis, which is notorious for allowing coolant to leak
> onto the motor. That was my problem 5 years ago, the motor got
> coolant into the encoder area, new motor cured problem. Performance
> symtoms were similair to now, but the Servo Dynamics tech tells me
> the balance problem is strictly related to the amplifier card and not
> the motor.
>
> I do intend to switch motors between X & Y to make sure the problem
> doesn't follow the motor. Meanwhile due to work load I need to keep
> this machine running since it has the longest travel of my mills.
>
> Any thoughts?

Try switching the motors for your own satisfaction. You could also
switch the amps to see if the problem follows the amp, but you'll really
need to readjust their pots for it to be a fair test. Servo Dynamics is
a good company so you could get then to fix your amp if need be. Worst
case is a new amp at about $400, and since you have the work for that
machine...

Matt
(410) 521-3715