dirt simple turbo pump controller?

[AllenWallace]

What would happen if one drove a turbo pump like a stepping motor? With no feedback from any motor position sensors?
You would have to ramp it up slowly at first, just like a stepping motor, and you couldn't get really high rpm.
I'm looking to get from , say, 50 millitorr down to 1 millitorr. Nothing like the sub-millitorr pressures that turbo pumps can deliver.
The electronics would be very simple.
Allen

[Ross Moffett]

It's not a stepping motor, so you can't drive it like one. If you have a single-phase motor you can find a schematic for a driver here in this forum. If it's a 3-phase one you can make a driver using a 3-phase driver chip, similar to this one featured on hackaday. Since you don't know the difference between the motor types, my guess is you'll be better off buying the manufacturers controller.

My turbopump does not have a motor RPM or position sensor, just what looks to be a thermal or vibration switch. The performance of the pump is monitored by the current draw from the motor, if it's monitored at all.

[Nick Peskosky]

Allen - An additional problem you could run into with "stepping" a motor that isn't designed for it (such as most single and three phase TMPs) is that if you mis-time commutation of a phase winding, you could end up freezing the motor between magnet poles or worse inducing a severe back EMF on one of the out of cycle windings. Both of these scenarios negatively impact motor efficiency and place undue wear on the driver FETs/IGBTs.

Ross - There is a good chance that your motor is driven through back EMF control, which is especially likely if it is both a somewhat older pump and wired for 3-phase. My Alcatel 5150 works on this principle and at one point I considered building a controller but abandoned the effort because of the necessary controller complexity and cost versus rebuilding an eBay scrap heap unit .The controller times its advancing of the energizing of the windings by sampling the induced EMF and zero crossing point of the non-energized winding. If you want to read more about this theory check out the attached link:

http://www.digikey.com/us/en/techzone/m ... k-emf.html

[John Futter]

Nick Allen
I have designed a turbo controller that works with all early Leybold units
it is three phase
there is no feedback required. Rotor position is inferred
more about it here
http://www.coultersmithing.com/forums/v ... ?f=12&t=83

[AllenWallace]

This is all great advice! My unit is a Phieffer TPU-050 with four drive coils. Normally a pair of Hall effect sensors determine which coil to energize next, however my sensors are defective.

[Peter Schmelcher]

Hi Allen,

I have the same pump missing all the bottom bearing assembly plus a defective Hall sensor. The markings on the sensor are A UGN 3177 515. The UGN3177 is no longer in production. The recommended Allegro replacement part is A1210. I don't have any track record rebuilding these pumps but it looks doable via the 4 bottom end bolts and the shaft retaining nut.

If anyone ever disassembles or rebuilds the bottom end I would appreciate the markings on the ball bearing and some assembly pictures.

BTW a Leybold NT controller has a resonance detection circuit to protect the ball bearings. If the rpm does not pass through 50K in under 1 minute it shuts down. The drive frequency is increased in steps with a coast in between to measure the rpm.

-Peter