0
\$\begingroup\$

I have an application where there can be considerable inductive flyback current into the supply rail, to raise it above the level naturally outputted by the AC-DC converter (48 V). Namely, the full rated output current (~10 A) can reappear as flyback current gradually decaying over up to several seconds.

We want to add some shunt OVP to the rail (a switching shunt regulator), but we are uncertain about a suitable trigger level for it. Namely, I am uncertain how to interpret the output OVP ratings for some supplies I consider. I already inquired with the manufacturer but got no answer so far.

E.g. VMS-550C-48

enter image description here

E.g. RSP-500-75

enter image description here

A few short questions all related to the How? and Why? of the OVP:

  1. If the output voltage is above the nominal 48 V, but below the values listed in the tables above, does that mean, that the PSU is still "on" but merely not switching anymore? Or at least that it has gone to its minimum possible duty cycle ? Surely the voltage feedback should force it into such a mode.

  2. They both mention, that the output shuts down, when the OVP is reached. However, I can't imagine how this can possibly be any more "off" than situation (1.) above. The output rectifiers must withstand this voltage regardless of the input being on/off. And current will continue to flow into the output. So what is being shutdown and why?

  3. Do the voltage ratings mean that the OVP trigger point can be between 48-60 V for the VMS-550C-48 and between 58.4-68 V for the RSP-500-75? At least the first one seems unlikely.

  4. Does the description ("latching", "re-power to recover") mean that I have to cycle the input voltage for the supply to provide DC power again, once the OVP actually triggers? That would imply that our OVP shunt should limit the voltage to below the OVP trigger voltage.

\$\endgroup\$
17
  • \$\begingroup\$ Have you considered using TVS Diodes to limit these induced Voltages? If youre dealing with short enough pulses some power TVS diodes can handle hundreds of Amps in a simple SMD package. \$\endgroup\$ Jan 27 at 9:50
  • \$\begingroup\$ @PatrickFiedler we are pretty set on a (hysteretic) switching shunt regulator with a trigger point derived from a reference to get the necessary voltage accuracy and sharpness and energy handling capability. The flyback current can last for seconds and so TVS sounds too gambly. \$\endgroup\$
    – tobalt
    Jan 27 at 9:56
  • \$\begingroup\$ I see, in that case I agree with your assessment, you definitely need a more elaborate solution. \$\endgroup\$ Jan 27 at 10:11
  • \$\begingroup\$ there can be considerable inductive flyback <-- how much? It's a numbers game. \$\endgroup\$
    – Andy aka
    Jan 27 at 10:21
  • 1
    \$\begingroup\$ Excellent. Then a threshold somewhat above 48V (max tolerance plus some margin) and below 60V (OVP min plus some margin) should be exactly the triggers you're looking for. \$\endgroup\$ Jan 27 at 11:05

1 Answer 1

2
\$\begingroup\$

I am uncertain how to interpret the output OVP ratings for some supplies I consider.

If, by internal or external influence, output goes beyond a pre-set value then a pre-defined protection kicks in. The behaviour and the levels may vary but they are, generally, determined by the market/customer requirements. The PSUs you are linked have medical safety approval, the medical market are generally happy with OVPs no more than 140% because these PSUs are meant to be used with inductive loads as well such as dentist equipment, surgery equipment, adjustable patient bed, etc.

Answers to your questions:

  1. If you apply voltage to the output from a source which has way lower output impedance and higher output voltage then the regulation will most likely be lost. The overall behaviour may vary from design to design.
  2. Basically, PFC and DC-DC regulators except bias (if any) will stop switching. The output rectifiers must withstand this voltage regardless of the input being on/off true. And current will continue to flow into the output. No, unless the output configuration is a bridge-rectifier the output rectifiers will block current flow inside regardless of the topology.
  3. The trigger point depends on design. If the output rises slowly then the OVP may trigger slightly above the expected limit. That's why you see a range instead of a stone number.
  4. Latching means that the PSU stops indefinitely. You need to remove the input power, wait for some time, and re-apply the input to clear the OVP situation. That would imply that our OVP shunt should limit the voltage to below the OVP trigger voltage. True. You'll need to implement a separate, independent protection.
\$\endgroup\$

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.