Turbo pumps

[johnf]

Chris
Not quite. You have to remeber that not only are the drive phases 120 degrees apart but so are the motor windings so you end up with the three pases setting up a rotating magnetic field this makes the rotor rotate in motor mode it is always lagging ie slipping a bit but if you ramp the motor speed down with the controls then the rotor looses energy by truning into a generator ie it is leading and dumping energy back into the controllers DC bus

[chrismb]

So is that a 'yes' to connecting 120deg separated switching outputs, 1 2 and 3, straight to A B and C (howsoever the currents flow)?

[johnf]

Yes
and if it turns the wrong way interchange any two wires

[chrismb]

OK.

So what I am considering trying is to rig up an oscillator driving a Johnson counter (which drives a set of FETs) the oscillator being under the control of a comparator also connected with the output of my pressure gauge and a manual set level. Once it is syncd to the motor, then I would presume the slow pressure changes would mean the pump will then change speed so as to maintain the set pressure, without a gate valve being necessary.

Is this a plan, or a misunderstanding of something important?

[Jerry]

Basically the drive is very similar to a standard variable frequency drive except this drive goes much higher in frequency. Most standard VFDs on the market top out around 400Hz. The motor in the pump is just a three phase motor. Nothing too special, just a lower voltage (Leybold says 150v). Most VFDs can be set this low so they could be used to test the motor if nothing else.

About coupling the speed of the motor with pressure, I dont think it will work. The delay between spool up or spin down is going to be so large that closing the loop and getting a stable pressure will be darned near impossible. Thats why MKS and others make motorized valves.

[johnf]

Jerry
I tend to agree with you
We found a bug in the software on our inhouse turbo pump controller the other day
If you unplug the mains it while it is spinning then power up again the S/W assumes that the pump is @ rest. If you push the start button there is a terrible noise as the rotor spools down to 300hz drive with a few over volt messages from the DC bus appearing on the LCD. All of this over a few seconds, remember the pump was running at 70,000 rpm and has spooled down to 3600rpm over about 5 seconds.

another S/W release is about to happen !!!
PS pump survived but I do not think this is good for the bearings.

The only thing that saved the pump was the time constant / integrator in the VCO line from the micro ie analog electronics

Chris
your idea is Ok as long as everything is in sleepy time ie changes happen over minutes /10's of minutes

shorter than that get a throttle valve

[Jerry]

Thats one nice thing about VFDs. The better ones have a feature where they sense how fast the motor is running from back emf or something and then can match that frequency and start up from there.

[Joe Jarski]

I tried running a Varian 3 phase pump on a VFD once, with zero success. I kept getting faults related to cable size and length. My assumption, not being electronically inclined, was that the rotor inductance may be to low to smooth out the PWM pulses or something of that nature. That may be totally wrong, but I do know that the VFD didn't like the turbopump motor. Just a FYI on one of my "it should've worked".

[johnf]

Joe
Jerry
Chris
yep
These motors are loosely coupled compared with normal induction motors
Chris I would advise using the circuit I presented for the output drive, As Jerry has suggested you can achieve the required frequency input in what you are happy with Micro, analog, sig gen on the bench.
If what had happened @ work with a closely coupled motor there would have been a very ugly noise / interesting metal throwing as the motor self destructed.

To answer a query I was going to add, that that if you power up a winding With a little DC then you can sense the frequency /RPM that the motor is running at.
In this case you would power up two out of the three wires with a DC current and read the Freq of the non powered windings.
Other pumps have harder magnetic material in their rotors so they self generate ( see Doug's post)

All for now

[chrismb]

Cable length, hmm!?

My V60 controller comes with a few metres of a control lead - permanently fixed into the controller. I was going to shorten this so that I can interchangeably use the plug for other turbos, but might I take it from this comment that the cable length is a design feature of the controller and should be kept the same?