Simplistic MosFET HV Shunt Regs

Jump to Latest
R

reubster346

Member
Joined 2004
Hi All
Reaching out for some assistance.
I've built 2 x SSHV2 Regulators using the GB PCB which will be partnered with the TubeCad PH2 Phono preamp

Each channel of the preamp draws about 10ma, the target B+ is 300V.
The power supply filtering on board the PH2 PCB is RC RC: 1.5K 50uF 1K 50uf. The caps are PP
Both channels of the preamp are working perfectly via the raw DC supply and each channel pulls about 10ma.
The signal ground is common between the 2 channels.
I've separated the B+ supplies to each channel

The regs work beautifully on the bench with a 27K Dummy load resistor and an input of 350V.
The target current is set to 30mA (300mV) and is rock solid and the B+ varies by less than 1%
Heat sinking is more than adequate, after 1/2 hour they are barely warm to the touch.

The problem comes when I connect to the PH2 Preamp.

When I connect the reg to the PH2, I am using a variac to wind up the voltage. When the incoming DC has reached ~150V things are looking OK.
- The LED is lit
- You can measure 300mV at the test point.

As I wind the voltage further up, the IRF840's go "pop"
I've burned out 6 under experimentation and following the failure, a resistance measurement between the Drain & Source pins measures in the range of 2 - 5K ohms.
It seems to be a very repeatable mode of failure.

In each instance, I've always pre-confirmed that IRF840s are insulated from the heat sinks before power up.


I'm at a loss to explain this, but am thinking the issue may pertain to the fact that at the point of failure, the DC heaters may not be fully warmed up - therefore the tubes are not pulling any load.

Or otherwise, the PH2 filter Capacitance (100uF) is too large and is causing inrush.

The S0 & F0 and S+ & F+ pairs are twisted and connected to the PH-2 + & - input terminals. The run is ~ 20cm.
When not connected to the reg - you can measure that each pair are joined at the PH2 PCB
I've use this wiring to the raw DC so am confident there are no issues here

I'd really appreciate some support on this - as I'm out of ideas.
Cheers - Reuben
 
Salas

Salas

diyAudio Chief Moderator
Joined 2002
Paid Member
Hi

Surely no small components are broken too? Please inspect. Inrush at the load should be limited by the CCS anyway. Let it come up instantly without the Variac. Maybe a long time stress develops between what the shunt voltage Vref tries to achieve and what is meanwhile available.
 
Salas

Salas

diyAudio Chief Moderator
Joined 2002
Paid Member
Wishing you best of luck. See about using it in two output wires mode too. By connecting force and sense locally at the output connector. Just in case there's something weird going on with the channels common ground referencing. Let me know when ready for new tests.
 
Salas

Salas

diyAudio Chief Moderator
Joined 2002
Paid Member
Surely can go lower Vout with smaller value voltage reference resistors (R9, R10). That one is easy. But to provide 1A there must be found Depletion Mosfets (Q1 Q2) of such current ability (enough IDSS). Maybe there are some available with at least 150V spec also? I don't know.
 
You must log in or register to reply here.
Top Bottom