Re: [CAD_CAM_EDM_DRO] Question about Servo Drives

John Dammeyer wrote:

> Hi,
>
> I've started building one of the UHU-Servo drives with the circuit board
> from Embeddedtronics.com and the processor from Uli in Germany. When
> all was said and done, the cost of parts along with shipping has
> exceeded the price of a Gecko 320 on sale and pretty well matches the
> Gecko 340.
>
> The transistors I've chosen are good to 200V and 50A. I've also
> selected connectors that are removable so I don't have to turn clamping
> screws while playing around with it. With the right set of resistors
> the UHU-Servo can drive a much larger motor.
>
> Looking at the Gecko block diagram it appears to be a totally hardware
> solution while the UHU-Servo has an ATMEL processor and setup is done
> via the serial port.
>
> So is the biggest advantage to the UHU-Servo over a Gecko that it can
> drive a bigger motor or does the software capable tuning also make it
> more versatile?

Ummmm, maybe. Before you say it CAN drive a bigger motor, you
might check that it actually can do that without problems. I
have looked at what happened with one of these drives when
somebody tried it. I gave him some advice on the shortcomings
of the original UHU design, and that Embeddedtronics was going
in the right direction to try to solve some of the problems, but
that I didn't think they were "there", yet.

One of the problems I ran into undershoot. When one of the
half-bridges is sourcing current to the motor, and is commanded
to bring that terminal to ground, it is trying to interrupt the
flow of current through the motor's inductance. This causes the
inductance to try to pull the terminal below ground to keep that
current flowing. A diode needs to be placed such that this
current has a source. The body diodes in the FETS take
MICROseconds to turn on, but the high-side FET turns off in
nanoseconds. So, an equally fast (or faster) diode is needed.
The IR driver chips are rated for -5 V on the VS pin, and I have
measured that they can just tolerate about -7 V before failure.
I easily developed -12 V and worse in this condition.

Also, this is yet another synchronous-antiphase drive, which is
hard on the motor, especially if it is low inductance. Adding
an L-C output filter would help reduce current ripple, but it
makes the above problem worse, of course.

I don't want to explain too much further or I'll be revealing my
own trade secrets.

Jon