I designed a 50,000 volt power supply, that I am having problems with on grounding the HV side of the output. The HV appears to be injecting HV into my logic sections through the ground, damaging the drivers & square wave generator circuitry from time to time.

My requirements are a DC voltage that has a selectable output of +/-. Therefore, I have to ground the secondary side of my Fly-back transformer. My power supply's regulator and logic are all tied to a common ground.

I have been unable to locate any papers or info on how ground or isolate grounds for high voltage DC outputs relative to the logic or pre-driver analog sections of the circuitry.

Also, I should add that everything works very well without mishaps or otherwise, when the HV output is floating.

Schematic below:

enter image description here

My last thought was to isolate the grounds circled above. But if I do this, it may affect my 70V regulator's performance (I don't know).

Regulator section of schematics below:

enter image description here

Logic section of schematics below, ignore BJT:

enter image description here

To: Kyle B.

Yes, I have the HV ground return going back to the bridge rectifier’s GND. And it’s true all schematics does not show actual wire routing. I agree with the star method of returning grounds to their particular voltage source GND. I have used this in many designs on stepping motor drive circuitry in the past, and it works very well, even with active currents locking the motors on point.

But I have never worked much with voltages over 220 AC/DC. I don’t like using fly-back transformers (FBT) as high voltages source, but they will provide a semi-constant voltage to the load that I need. FBT are also a source of harmonic noise that has to be dealt with using chokes, snubbers, etc. I would rather use something like a Mark’s Generator, if it could keep a constant voltage to the load. Still working on all this.

Thank You Kyle B. for your thoughts, if you think of anything else, that would be welcomed.

  • 1
    \$\begingroup\$ Can you share a photo showing the physical arrangement of the transformer and the control logic? \$\endgroup\$
    – The Photon
    Commented Sep 10, 2020 at 20:19
  • 1
    \$\begingroup\$ You might be experiencing "ground bounce". All currents require a complete loop. As you have very brief but high current pulses running through your circuit, the transformer pulses are passing through "ground" trying to return to their source. Consider the ohmic value of the wires in this loop and realize these current will, due to E=IR, cause some non-zero voltage loss, thus causing ground to "bounce" up in voltage. At a minimum, you should be employing a "star-ground" scheme. I.e. any wires that carry high-current pulses should NOT also be carrying current which operates the logic. \$\endgroup\$
    – Kyle B
    Commented Sep 10, 2020 at 20:21


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.