Re: [CAD_CAM_EDM_DRO] Re: E-stop relay welds contacts for load dump resistor

Dan,

I looked at the recommended schematic for the DC servo power supply on
the Geckodrive web site. I had not realized that these were servo
drives as opposed to stepper drives. Servo drives are not the same as
stepper controllers.

As another poster mentioned, the circuit supplied is suboptimal, because
it passes all the initial charging current for the filter cap through
the momentary pushbutton before the holding relay takes over. That is
immaterial to the problem at hand, but can be addressed with a
rearrangement of the circuit.

Usually, the injunction is against a switch between the drive and the
motor. That would certainly be destructive to the drive. I can see
that a switch between the drive and supply could also be an issue if you
are driving a DC permanent magnet motor in a servo configuration. If
the motor was turning at a high rate of speed, and you open the supply -
to - driver connection, the motor would act as a generator and force
energy back into the drive. With its supply end open, there would be no
place for the energy to go except into the semiconductors of the
driver. That would let all the magic smoke out. Pffft!

I would maintain that using a dump circuit as suggested is not the best
approach, as it leaves the motor freewheeling and uncontrolled every
time the E-stop is engaged. It's like turning off the ignition key of a
runaway car on the highway at high speed. Yes, it stops the engine, but
you are still going 65 and now you have lost your power steering and
power brakes. I do submit that it would be helpful in case of a driver
runaway condition, where the driver is running the motor independently
and in contradiction to the input signal. If the driver is well
designed, then this should be a rare occurrence.

However, in the event of a computer hang or descent into mumble mode, I
would personally prefer having the driver fully powered, and exerting
its influence to bring the motor, lead screw, and table to a stop,
rather than having them coast uncontrolled to a stop under their own
inertia. In my experience, failures of the control computer or software
are more frequent than internal servo driver failures. YMMV.

You could say it's a matter of electrical momentum vs. mechanical
momentum. If you just break the line power to the transformer, the
stored energy in the filter cap will drive the motors until the cap is
depleted. On a three axis machine, you can figure that two of the axes
will be healthy and obedient, and one axis will be in runaway
condition. All will remain powered until the cap is drained.

OTOH, if you dump the power cap, now all three drivers will be
helpless, and all three motors will be freewheeling. If you were in the
midst of a high speed movement, all of the axes in motion will coast
uncontrolled to a stop at a rate determined by their mechanical
rotational momentum. Personally, I would prefer the former situation to
the latter.

A simple experiment would be to initiate a full speed movement starting
from one extreme of your longest axis toward the other, and then
initiate an E-stop condition with no dump resistor circuit. In other
words, you would be just disconnecting the line power input, and letting
the filter cap drain on its own, but with all step signals to the drive
inhibited. The distance the table moves will indicate the electrical
momentum of the system. Then try the same maneuver with the dump
circuit in place. The distance the system coasts will indicate the
mechanical momentum of the system. With this information, you can
decide if the dump circuit is valuable or not. You will have to decide
if you are more worried about a driver module runaway or a main control
system failure.

Another poster suggested using a beefy SCR as a dump element. These are
available for a few bucks on eBay, and would be ideal as a substitute
for a relay contact in a dump circuit. Just be sure to use a resistor
high enough in value to keep the peak current well below the rating of
the SCR.

Dave

On 1/31/2011 10:54 PM, kzf817 wrote:
> Dave: I was thinking about your advice, below and then double checked
> Geckodrives documentation for the G320X servo drive. I did most of my
> electronics enclosure assembly and wiring last winter so I had to look
> over some of my schematics and documentation. Mariss specifically
> states in his tech sheet for the G320X:
>
> "IMPORTANT: Never put a switch on the DC side of the power supply!
> This will damage, if not destroy, your drive!
>
> I am sure that Mariss is speaking from experience, so this is why my
> E-stop circuit is designed the way that Geckodrive specifies. Thanks!
>
> Dan.


[Non-text portions of this message have been removed]