Simplistic MosFET HV Shunt Regs

[Telec]

I was about to use the SSHV2 in a tube preamp, but there is a huge ripple (~0.3 V) ripple in the amp output. I measured the same ripple in the regulator output with the dummy load resistor. Any clues where the possible source would be?

My amp consumes 34 mA at 270V and I have set the total current to 54 mA

 

[Salas]

Decouple the input just in case there is some oscillation at the CCS due to incoming wires inductance. Use 0.1uF-1uF right across the raw DC in connector. If its not this we will think of further stuff to check.

 

[Telec]

Decouple the input just in case there is some oscillation at the CCS due to incoming wires inductance. Use 0.1uF-1uF right across the raw DC in connector. If its not this we will think of further stuff to check.

Thanks, Did try that (0.33 uf MKP on raw input) but the ripple is still there clearly audible.

 

[Salas]

Have you seen it on the scope? Whats its frequency? Is it an oscillation or 100-120Hz ripple?

 

[Telec]

Wont have access to scope at the moment. But i would say from the sound of it that it is 100 Hz and sounds a bit like sawtooth, so I suggest it is the original ripple from CRC filter. Which is 200 uf + 2 Kohm + 200 uf

 

[AndrewT]

Thanks, Did try that (0.33 uf MKP on raw input) but the ripple is still there clearly audible.

MKP is usually low esr.
That will cause oscillation rather than snub it out.
Higher esr, or added R in series with the, decoupling capacitor/s is/are more likely to snub ringing.
We are seeing many recommending X7R for HF decoupling

 

[AndrewT]

Wont have access to scope at the moment. But i would say from the sound of it that it is 100 Hz and sounds a bit like sawtooth, so I suggest it is the original ripple from CRC filter. Which is 200 uf + 2 Kohm + 200 uf

f = femto
F = Farad

 

[Salas]

Check that every semi works first. Q1 Q2 Q3 must be showing some VGS (DC small voltage between pins G and S). Q4 Q5 about 0.6V VBE each. Q6 must not show a short between D S when the system is powered down.

 

[Telec]

f = femto
F = Farad

Yes, and there is a mistake in kilo ohms as well. It should be kohm...

 

[AndrewT]

kilo ohms is not ambiguous.
Yes,
K = Kelvin.
k = kilo.

 

[Telec]

Check that every semi works first. Q1 Q2 Q3 must be showing some VGS (DC small voltage between pins G and S). Q4 Q5 about 0.6V VBE each. Q6 must not show a short between D S when the system is powered down.

Q1 = 1.300 V
Q2 = 1.255 V
Q3 = 3.16 V
Q4 = 0.559 V
Q5 = 0.558 V
Q6 = 0.143 V ( no short between D and S, resistance about 90 ohm)

 

[poty]

And how about the voltage across the CCS?

 

[Salas]

At least 10V drop it needs. The semis readings look OK. The CCS should be stopping 300mV ripple easily anyway. It takes to scope it now so we see if its oscillation which fills everything up with harmonic noise spray.

 

[Salas]

* Check that Q4 Q5 have their printed face looking towards R8 and Q6

 

[Telec]

Thanks, Will check these

 

[LuvValve]

Remember you should jumper S0 with F0 & S+ with F+ locally at each output connector and then use thick enough force type wires only. Good luck.

I did that and still pops. I feel more likely the caps are somehow leaking because if I mute the preamp only once and quickly un-mute before I switch the power amp, I don't hear pop. By mutting , the relay shorts the output to ground. This preamp has 2.2uf Mcap Silver Oil cap at output.


BTW, it makes very noticeble differrent when running the reg in 2 wires mode. 4 wires mode is so much better to my ears.

 

[Salas]

At least it was an opportunity to compare SQ in a certain build between wire modes.

I remember there was some talk maybe years ago that SIO mustn't be used near hot tubes because its oil will react and there will be some hundreds of hours useful life remaining only. Don't know if that was ever proven, but I had serviced a BAT preamp once that it was doing weird things and it was due to DC leaking Jensen PIO packs in the end, which I replaced. That does not mean that you would not look in to the power transient hypothesis also. Digital oscilloscope in single shot mode should help a lot.

 

[Salas]

P.S. Are there coupling capacitor's output to ground reference high value resistors in that circuit (say 500k -1 Meg) ? Those are to avoid some pop situations relative to charged up coupling caps connecting and disconnecting.

 

[Telec]

Thanks, Will check these

Turns out it was a ground loop after all. Thanks for the good advices though. Nice to have such a helping community in case of problems

 

[Salas]

Alright, good news.