Why does this HV circuit oscillate?

[merlinb]

Can anyone explain to me how this circuit oscillates?

I have been experimenting with controlling high-ish voltage using a MOSFET controlled by an optocoupler (6N136).
Feeding clean DC into the driver side I can control the conduction of the MOSFET, but at most settings the high-voltage output oscillates wildly (see screencap).
I can kill the oscillation with either a capacitor between output and ground, or a capacitor between gate and source. But I'd really like to understand the mechanism here. Why does it oscillate, and how does the oscillation manage to exceed the 368V DC input voltage without any obvious reactive components??

Yellow trace is the voltage across C2.
Blue trace is the voltage across R1.

[jonpaul]

Hello great question: The very widebnd FET can easily form an oscillator from stray L and C.

1/the parasitic oscillation tank is formed by the PCB and FET inductances (perhaps nanohenries) and th self capacitances.

2/ Add a series gate damping resistor 100-1000 1/4 or 1/2 W VERY close to the FET gate terminal.

3/ These circuits are sensitive to the load, try with a resistive load say 2-10W non-inductive R.

4/ Be careful with 380V bus, you or your equipment can get nailed by the bus cap stored energy.

CAN BE LETHAL!

See the excellent FET app notes on parasitics, eg from Intl Rectifier.

Bon chance,

Jon

[xavier60]

Driving the MOSFET like that keeps it in it's linear region. A charge pump could be used to produce a higher voltage for the Gate to bias it well above its threshold.

[T3sl4co1l]

Use SFH6345.

Tim

[ocset]

Also, your voltage hasnt overshot by much....maybe it is a scope/probe anamoly.....maybe its an old uncalibrated diff probe you used for one measurement?
I am sure you realise that 240VAC mains peaks at 339V....so maybe your 368vdc measurment is out...is the scope  fully calibrated?

If you want low power controlled voltage from 368vdc to zero...then you could rig it up with a LNKswitch from power integrations if you wanted.
Or just make your self a little flyback converter using  any current mode controller of your chcice.

Or if you just want linear regulation, and your power is low, you could put in a series PFET and an error amp by way of an opamp and control the PFET gate with the opamp to give you your wanted (variable) vout.

But also the 368VDC you see could be due to the light loading, (low damping)  and the fact that your isolation transformer has high leakage inductance, and your 50uF is peak charging to the peak of the leakage inductance ring.

Attached is an easy zener based output voltage regulator along your lines (LTspice sim and pdf schem attached..LTspice is free download)

Also attached is an opamp based regulator, where you can vary the vout by varying the control voltage into the error amplifier.

If your load really is just a pure resistance, then you can use a current regulator to push x amps through it and give your wanted vout...such current regualtors  can be quite low in component count.

[Zero999]

A BJT opto-coupler is not the right thing to use to drive the gate. As drawn, the MOSFET is operating as a source follower, meaning it will never fully turn on, have a high voltage drop and dissipate an excessive amount of power, when passing a decent current. Use a photovoltaic opto-coupler, such as the TLP591B, which will turn the MOSFET fully on.
https://www.mouser.co.uk/datasheet/2/408/TLP591B_datasheet_en_20190624-1132742.pdf

How fast does it need to switch? The circuit attached to the original post and my idea are both quite slow, with the MOSFET taking several ms to turn off. The turn off time of a photovoltaic MOSFET driver can be reduced by adding a transistor to speed it up.

[merlinb]

2/ Add a series gate damping resistor 100-1000 1/4 or 1/2 W VERY close to the FET gate terminal.

Hmm, I tried a 100R gate stopper but it made no difference. I then tried a 15R drain stopper too, and that made no difference either! I feel like there's more to this than just the usual 'mosfets are squirrely'.

[T3sl4co1l]

I think no one caught on to my cryptic but pointed comment earlier, so it's probably time to explain what that means.

6N136 has notoriously poor CMRR.

The problem is the internal wiring, lack of shielding, and base brought out to a whole pin.

Any capacitance to nearby (AC) ground, causes positive feedback into the base.

This includes the LED, or the power MOSFET's drain, etc.

Basically, by biasing it with a pull-up resistor and mild turn-on, you've made a high voltage astable multivibrator.

By changing to a low capacitance, no-base-pin, well shielded device like SFH6345, the symptom will disappear completely.

A keen reader might find the comment perplexing enough to read the datasheets and compare the two devices, and perhaps not find the critical parameters in play here.  Indeed, they do not specify any capacitances to the base.  The inference is the CMRR test, which while done at a seemingly respectable dV/dt edge rate, is done at a suspiciously low step of only 10V, whereas other optos are done at a much more practical 1000V or thereabouts.  (Likely, they chose this because it's simply not enough injected charge to turn the transistor on/off in a logic-level test, and so the dV/dt really doesn't matter much at all.  It's one of those specs you need to read with a jaundiced eye and figure why they're telling you what they're telling.)

As for practical solutions, this is not one.  Definitely, use a proper power supply circuit, with gate protection, current limiting, and voltage control or whatever it is you need at the output.  Nothing worse than slipping a wire, momentarily shorting out your hard work and pfft...BANG, up it smoke it all goes.

Tim