I don't know what a 600V dust motor is.

Here is an example:
https://ntrs.nasa.gov/api/citations/20130012970/downloads/20130012970.pdf

This paper uses 3 phase, but 2 signals 90 degrees out of phase can be used to make 4 phase. Thank you for your diagram, I will study it as soon as I have the time!

 

Here's a rough sketch of the 'poor man's inverter' I had in mind. Notice how the 12v inputs have their duty cycles shortened a bit so they're never on at exactly the same time:

 

I see an unanticipated issue in using the transistors as the switch in the scheme in post #19. The output is pulled up thru a high resistance, one megohm, and pulled low by a switched on transistor. , which is a quite low resistance. So the impedance at the point is changing from very high to very low. That will cause some sort of problem I am sure.
And I see serious destruction in the circuit shown in post #22, which is that for half the cycle the transistor has 1000 volts base to emitter. There is a problem with that. Quite a serious issue actually. A possible solution would be a well isolated base drive for the switching transistors. But it would not be simple because of the high voltage.

 

#22 is a very simple schematic with pieces missing.

When the input is 0V, Q1=off, R1 turns on Q2.
When the input is high, Q1=on, Q1 first pulls Q2-Base off then pulls down on the output.

 

I will need to redraw the whole circuit and think about it . Really it is part of an "H" bridge in disguise. How could I have missed that??

 

#22 is a very simple schematic with pieces missing.

When the input is 0V, Q1=off, R1 turns on Q2.
When the input is high, Q1=on, Q1 first pulls Q2-Base off then pulls down on the output.
View attachment 333740

This is interesting. Is R2 also 1 meg?

Is this design applicable to a wide variety of HV NPN transistors? Or do the transistors you pick change the resistances you need for R1 and R2?

What is the function of the diode? What current is it blocking?

 

This is interesting. Is R2 also 1 meg?

No. R2 limits the current into the Base.
If the collector current is 1mA. And if the transistor has a gain of 20 at 1mA worse case. Then the Base current needs to be 1/20 of 1mA. (or al little more) The drive signal to the Base might be 0 to 5V or 0 to 12V.

Is this design applicable to a wide variety of HV NPN transistors? Or do the transistors you pick change the resistances you need for R1 and R2?

The gain of the transistor does affect the value of R2. (some)

What is the function of the diode?

Without the diode Q2 does not function. Or not well.
Note the +0.6V and the -0.6V from Base to Emitter of Q2.

 

This is interesting. Is R2 also 1 meg?

Is this design applicable to a wide variety of HV NPN transistors? Or do the transistors you pick change the resistances you need for R1 and R2?

What is the function of the diode? What current is it blocking?

It looks like the diode is preventing reverse voltage bias rather than preventing current.

 

It looks like the diode is preventing reverse voltage bias rather than preventing current.

With out the diode the B-E of Q2 are shorted.

 

With out the diode the B-E of Q2 are shorted.

Certainly that is also true.

 

Do you want a MOSFET that is turned on by voltage not current?

Yes, by voltage and not current.

If you want more part numbers I can find them. There is a class of transistors used in the focus control of TV sets. Most are not made now days but I probably have a box of them in storage.

I might take you up on that sometime! But for now I will buy the STN0214 and give it a try.

Thanks for your help!

 

I will need to redraw the whole circuit and think about it . Really it is part of an "H" bridge in disguise. How could I have missed that??

Yes I accidentally reinvented an H bridge by starting with a full wave rectifier and reversing it. ronsimpson's circuit in #24 would be one side of a possible H, I think.