Re: [CAD_CAM_EDM_DRO] Re: High voltage servo controller

Jon,
I thought we've been over this.... ;-) The HP UHU is NOT the same as the
'old' one. We discussed it a couple of months ago remeber?

The HP UHU incorporates MOSFET drivers with deadtime, RLC filter on the
output, hardware cycle-by-cycle current limit and has been tested at 185V
with good results (so far). Personally I've run mine at 160V with a 19A
continous current rated motor - on the bench though, still need to get the
machine together...

If you follow Rogers Youtube-link you'll see MY test-setup, running at 130V
which is what I'll have on the machine. And here:
http://www.youtube.com/watch?v=eB1AaByTjYc you can see it running a 1kW
lathe spindle, IIRC he's running at 160V.

My problem right now is the limited encoder count frequency, I have
3600lines encoders on a 2000rpm motor, 4-5 times what the UHU can handle....

As more people are building them I'm sure we'll hear success or failure
stories in the near future.

/Henrik.-




-----Original message-----
Oh, man, don't believe it! I critiqued the UHU at last year's CNC Workshop
for a guy, and I think I went on for nearly a solid HOUR detailing what he'd
done wrong. The freewheeling diodes are standard rectifiers, and about 1000
times too slow to protect the FETs. The gate turn-on resistors are 270 Ohm,
which causes insane switching losses in the FETs, but it has to, to save
them from the destructive effects of unclamped inductive pulses from the
motor, since the diodes are too slow. The current sensing circuit is blind
to recirculating current and braking of the motor, since it only has one
sense resistor.
The power circuit topology is poor, with large loop areas.

If you think you can get it to run at 160 V, go ahead, but be sure to wear
ear muffs and goggles when testng it under load.

(There seems to be a new version 3 board, maybe it has incorporated some of
these improvements.)

Jon