Science/Engineering/Technical forums

Hi
I have a small Leybold TMP50 turbo pump, but have blown the controller in a flash x-ray tube driven by a marx-generator experiment. The pump can be opened and allow a look to the motor. The motor has only 3 wires ( delta circuit ? ) and looks like a normal short circuit rotor asynchron motor. I found even used controllers are very expensive so I have to build my own one. The first idea was to use a audio amplifier and generate the drive for the third connection with a capacitor. But a the run-up of the pump the drive frequency change from a round 10 Hz to 800 Hz and I think also the capacitor should be change according to the frequency change. Changing a capacitor from estimateds 4 µf down to arround 0.1 µF is not a easy task.
Perhaps one could use a controller for the meantime very popular brushless modell motors. The controllers a really cheap, and the power requirements of the pump ( 48 V, 3A) should be possible with upper range controllers. In the moment I don't know what kind of driving signals this controllers need but it shouldn't be to difficult to find out.

Thomas

Thomas
the circuit values I posted earlier in this thread are for the TMP50

Hi Johnf,
thanks a lot for the advice. So the motor ist driven by squarewaves rather than sinus.I will try to built this circuit. Maybe it will be difficult to get the 3-phase bridge driver ir2130 but I think it can be replaced by 3 single phase drivers. Perhaps if I had this information one year ealier I would have been able to repair the original controller. I tried to replace all bridge transistors and the switching transistor in the power supply but without success (I hate switch mode supplies because if they are wrong there no volatges to measure)
Thomas

I believe John is using the current sense amplifier in that driver chip as part of the input to the microprocessor (not shown) to control how fast the frequency is ramped up so as not to get ahead of the rotor, which likely has permanent magnets in it rather than being a pure induction motor. In the Pfeiffer ones I use, they are, and the turbo provides power in power failure situations to run the controller while it is spinning back down -- for plenty long enough to do other things to get a more orderly shutdown sequence should mains power just fail.

John is using square waves generated by the ring counter to avoid having to control 3 phase pwm, which results in the need to step the power supply voltage partway through spin up. His design makes the uP programming a lot easier for those who don't "do real time" in their sleep -- probably a good idea for most programmers. You might be able to do more in the uP, but you'd have to be fairly careful with most of it to keep from interrupting the drives or getting them out of phase. You wouldn't save that many parts that way, it's a tossup what would be the best. Haven't looked, but it wouldn't surprise me if some PIC has a 3 phase PWM hardware generator, though.

We can get most things here, and don't mind helping our pals overseas if they can't find them -- let us know, (maybe put up a wish list on the trading post).

Here is a thread on fusor.net that shows a two phase controller for turbo pumps -- looks like Jerry was right about those.

http://www.fusor.net/board/view.php?bn= ... 1283393372

Alex has kindly allowed us to use a picture of his 2 phase turbo driver here, so it's all in one place. Looks like Jerry was right -- if he wasn't this wouldn't work.

John, any chance in sharing the rest of the circuit for the turbo driver?

Jerry
Sorry no
the S/W is owned by work --not my doing-- all development was done with what you see using a sig gen.

clue 4 minutes to full speed--start at 300Hz finish at 7200Hz --the circuit is very forgiving--except for the DC bus clipper--we don't use a heatsink but more severe action will require one

I've given enough clues

Doug did mention that I'm monitoring the current --that is all it is --I do not act on it --so that "sticky" pumps can be used.

all for now

Jerry, I can fix you up there -- John's given us enough data to get there. The thing is, I don't have the motivation, which would be in the form of having a pump that needs it. If I had that, I'd do it. This is my longstanding rule -- if I'm going to do development sweat, I get the prototype to keep, free.

I think I can handle it as well. Though I have about the same motivation. I pretty much got this as a spare in case I do something stupid with my Pfeiffer.

I printed out the schematic and I think I have everything figured out.

I was looking at a couple other designs. Is the rotor in the pump a permanent magnet or more like a regular induction motor?

-Jerry