Why is my MOSFET dropping so much voltage?

[arnoldc]

Help

Here is a schema of a high voltage regulator I found on the net.

I wonder why am I getting only 100V at the output

The zener string gives 211V reference, but I can't get the proper output I am expecting- 200V

How can I get this right?

[arnoldc]

Darn, I simulated this thing with LTSpice and it should give me around 200V.

Grrrr.

[x-pro]

Originally posted by arnoldc
Darn, I simulated this thing with LTSpice and it should give me around 200V.

Grrrr.

So it should. However here is a list of possible problems:

1) The gate of the MOSFET needs protection - a 10-15V zener between gate and source, otherwise the MOSFET could easily be broken - resulting in a high gate current and an additional voltage drop on 1M resistor.

2) 47uF capacitor could be leaky, again reducing the output voltage.

3) 1mA current through the zeners may be not sufficient for a particular zener type you are using.

Check DC voltages on each capacitor and it will be clear where exactly you lose the voltage.

Cheers

Alex

P.S. - perhaps it would be reasonable to add a protection diode (i.e. 1N4004) across the current source transistor in reverse polarity and a series resistor to the drain of the MOSFET to reduce the dissipation and protect at least a bit against short circuits and overloads.

[arnoldc]

Hi Alex, thanks for the reply.

I'm using BZX79C 51V zener diodes, they should be 1W type but I have to check my Farnell order for those.

I'll put in your suggestions to the circuit and see what happens.

Thanks again for taking the time.

Cheers!

[Eva]

The leakage current drawn by the 47uF capacitor is probably causing too much voltage drop across the 1Meg resistor. Try 100k instead.

[arnoldc]

Eva, you're absolutely right!

Last night I was making measurements to check against the simulation on LTSpice (attached).

I noticed that the simulation says I should get ~200V after 1M (Node7 on model), but I'm getting only 100V or so.

So I was wondering why it is so.

Does that mean I have a crappy 47uF capacitor? I actually changed it to another one- 100uF and got the same results- 100V.

[Eva]

You have either a leaky capacitor, a leaky MOSFET or both.

I would insert 1K resistors in series with the 47uF capacitor and with the gate of the MOSFET to be able to measure where that current is going (100V/1Meg=100uA, 100uA*1K=100mV).

[arnoldc]

Hi Eva, I will try your suggestions tonight.

Thanks!

[arnoldc]

It's bad MOSFET

Thanks to Eva and Alex!

[x-pro]

Originally posted by arnoldc
It's bad MOSFET

Thanks to Eva and Alex!

Don't forget to put a 10-15V zener between the gate and the source for a replacement MOSFET - otherwise you'll end up with two bad MOSFETs... also could be useful to reduce the capacitor value on the gate down to 10-22 uF and the resistor value from 1 Meg down to 100K. The time constant on the gate is too large.

Alex