Current limit for HV Power Supply

[pedja089]

It was late when I replayed and didn't look in to schematic too much. And now it late again, so sorry in advanced if I miss something.
You are right, for P channel configuration is correct. Your design look like LDO. And that is inherently unstable. So watch out for that.
Left schematic for NPN will not work as expected. Collector base is reverse biased. And it act as diode.

So that is reason mosfet is cut of. Voltage drop across diode is much lower than Vgs threshold.
NPN current limit is tricky to implement here, because collector must be on higher potential than base...

PNP circuit should work. Or to be precise must work. I suspect that pin out is wrong or you didn't use PNP transistor. If pin out is wrong, T2 isn't correctly biased. I would start with transistor tester, or create simple circuit with PNP just to be sure is it what you think it is and pin out is correct. Try to disconnect base and see is it working. Measure voltage across 1.5R R8(current sense resistor, I don't see clearly part name) with and without base connected.

Also just to remind you that you are measuring output current and current thru R1. So take that in account. With 22K current is 12mA, so across R8 is only 20mV, not enough to bias BE junction, but check voltages current and resistor values. I'll also suggest that you move high side od R1 to 280Vin so it wont affect current limiting.

[Ioannis]

Well, the NPN as you described, and having collector in lower dynamic than the base, makes sense why it does not work.

But the PNP version does not. Double checked, changed transistors, still nothing...

Thank you for the time you spent on this.

Ioannis

[pedja089]

Can you try it without base connected? Measure voltages measured above?
It will be grate to get some pictures, T2 part no, layout etc... If you don't want to share public, send me links to PM.
Anyway it must work, and good luck.

[Ioannis]

Will do that and report back.

The final circuit will be pretty much as the above posted, though it needs a lot of refinement before it gets on a shiny pcb. Mainly for safety of the circuit and users.

Thanks,
Ioannis

[amgen]

alternate place for I limit circuit.....

[Ioannis]

I have such an option for the PWM and the PIC to stop operating in case of a SC.

But this takes time and the Pass transistor will be burned in the mean time. That is why I have to place the PNP transistor too on +V line too.

I found that the NPN indeed does have the B-C junction working as a diode blocking the Mosfet.

Also that the PNP transistor was defect...!

Soon will test for the PNP to see how it goes. On another test circuit it was promising as expected.

Ioannis

[pedja089]

That solution sound very familiar to me
I have done same thing on 42V Communication PSU. I used ADC in interrupt in background to monitor V and I, it was fast enough to not cause damage. Anyway I added hardware current limit, and it works for years without problem.