Re: [CAD_CAM_EDM_DRO] Request for assistance with Gecko and LARGE servo motor...

Kim Lux wrote:

>Hi people.
>
>I've got a LARGE (ie 3HP @ 120VDC) servo motor that powers a 200+ pound
>table via a direct drive (4TPI) ballscrew. THe motor is equipped with a
>250 count quadrature encoder and I'm driving it with a Gecko G320.
>We've got an 80V power supply supplying the system.
>
>We run 2 similar but smaller motors (ie 2 HP) on other axes, powering as
>large or larger loads (without any problems).
>
>The problem with the LARGE servo motor is that we blew up a G320 trying
>to get it to run properly. The drive would operate the motor at slow
>speeds, but not entirely correctly. The motor ran as though the damping
>was incorrectly adjusted, ie kind of rough. I spent all kinds of time
>trying to get the damping and gain set correctly, but it still ran
>rough. And then the drive blew.
>
>Any ideas as to why this could be happening ?
>
>
>
>

The large motor is most likely to have lower inductance than smaller motors.
Do you know what the inductance of the motor is? Did the motor or drive run
hot while idling? That might be a clue. As I mentioned before, and
Mariss has
confirmed it, the Gecko 320/340 drives put a 25 KHz square wave at the
DC supply voltage across the motor. If the inductance is several mH, no
problem. If it gets down to the 250 uH range, the AC ripple current
starts to get really large, approaching maybe half the current rating of the
drive alone (depending on DC voltage)! This rapidly rising current is
likely to drive the current limit circuits crazy, and may have adverse
effects on the stability of the control loops, as well. Do you have the
equipment to measure the inductance? You can add a series inductor
to raise the inductance to a more manageable value. Do you have an
AC current probe or high frequency current transformer? That would be
another way to measure the AC ripple current. If it is more than about
1 - 1.5 A, (ie approaching 10% of the rated DC current) then it seems
like it is too much.

Did the failure happen when the motor was moving at speed? There is a
problem due to the energy pumped back from the motor to the DC supply.
If this raises the DC supply voltage too much, the Gecko will pop for sure.
Again, the inductance comes into play. Normally, the DC terminal voltage
of the motor couldn't ever exceed that required to move it at that speed.
But, when the motor's series inductance is combined with the pulsing action
of the drive during deceleration, a much higher voltage is possible. I'd
decelerate the motor gradually while watching the DC supply for a rise
in voltage. If it gets close to the danger level, you need to add a voltage
limiting circuit. Mariss has posted one in our files section, I think.

Finally, you have to check the motor for possible grounding. It is not
unusual for grounding to be intermittent in older motors, due to
accumulation of brush dust around the brush holders. The Gecko drives
are totally unprotected against grounds of the motor terminals.

Jon