Simplistic MosFET HV Shunt Regs

[Salas]

Hi Salas,
great, not shabby at all! Finally something in the ball park.What detail of component or module you think is most critical for increasing this parameter?
What would be the max. expected value (PS still nice behaving) for current slewrate capability assuming 2Adc at 160V?
Thanks,

Those questions involve a full scaled up power components redesign to be answered

 

[vgeorge]

Rectifier hybrid Graetz AZ1 + two UF4007 (Between power transformer and the AZ1 one 100 ohms resistor.

Now only one 125VAC, attached Vout pic.

This arrangement looks safer.
Let us know how it goes. Too much voltage drop before.

 

[merlin el mago]

Now I remember the reason for 250VAC is when used the VR150 in place of SSHV2

 

[merlin el mago]

Max. CCS 23mA Max. Vout 107V dummy load 4K7

 

[Salas]

That Vout computes with that max ccs and load. Problem is more ccs is needed with that load. Use a test load representative of final load. Is Q2 ok? What's Q2's VDS & VGS?

 

[merlin el mago]

Yes I also understood that I need more CCS, I calculate the dummy load 4K7 to get 150V 30mA to simulate the 26 preamp.
VDS 0.254
VGS -0.210

 

[merlin el mago]

Black probe lead always connected to source.

 

[Salas]

It takes about 1.5-2.5V VDS & VGS in this cascode. Maybe Q2 is problematic. BTW remember to use input connector raw dc wiring decoupling cap when experimenting.

 

[merlin el mago]

Atlas DCA55 reads DN2540 as OK
1uF?

 

[merlin el mago]

Have I to change Q2?

 

[Salas]

Look to gate Zeners first. Take them out and leave their spots blank. Q1 must develop about -1.5V Vgs for 40mA which is Q2's Vds also as applied in the cascode. Q2 should have about same Vgs in its own turn.

1uF decoupling is OK for avoiding basic chance of test rig oscillation.

 

[Salas]

Also use a 100-200mA fast blow fuse inline with the raw feed to input connector or the lab PSU's CC mode protection at same mA when testing the CCS in low voltage.

 

[merlin el mago]

Look to gate Zeners first. Take them out and leave their spots blank. Q1 must develop about -1.5V Vgs for 40mA which is Q2's Vds also as applied in the cascode. Q2 should have about same Vgs in its own turn.

1uF decoupling is OK for avoiding basic chance of test rig oscillation.

Attached pics with measurements

 

[merlin el mago]

100mA fuse gone.

 

[merlin el mago]

200mA fuse also gone.

 

[merlin el mago]

Vin 9VDC battery
Q1 VGS 8.6V
Q2 VDS 0V

 

[merlin el mago]

PSU Vin 44V
Q1 VGS 26V
Q2 VDS -13.8V

 

[merlin el mago]

PSU Vin 44V
Q1 VGS 26V
Q2 VDS -13.8V

I forgot to tell you that before this measurement I changed the two gate resistors of DN2540

 

[merlin el mago]

I measured this morning the two DN2540 and soldered to the PCB are gone.

 

[merlin el mago]

D3 gone