Simplistic MosFET HV Shunt Regs

[jaz_12]

You must make a practical decision.

Thank you

 

[Tix88]

Hi Salas, I would like to try your shunt to feed the two mono Aikido PCBs, a CLC filtering system with SS straightening could be right for me? Would I benefit by using Tube rectification? Thank you for your precious answers

 

[Salas]

How many mA will two mono Aikido ask for?

 

[Tix88]

I salas Total Power is 20ma.

 

[Tix88]

Sorry 30ma [emoji42]

 

[Salas]

Set the SSHV2 for 50mA CCS and configure at the output connector two wire mode* as it will be a common reg between two channels the Kelvin mode would not help much anyway. CLC and SS bridge are fine. Use 0.1uF to 1uF decoupling film cap across the regulator's input connector if there will be significant wiring length coming from the CLC. Good luck and let us know.

*What is two wire mode

 

[savu]

Hi Salas, after weeks of using the regulator without any stability problem, this evening the output voltage just started to fluctuate. Now it has aground 10 V peak to peak fluctuation at the output, the output is set at 270V. I have checked the FET and PNP transistors and they all seem ok.
CCS is working fine (I can very the constant current from the pot). Diode D3 is ok also.

Can it be the IRF840?
Is it ok that I have a common heat sink for both DN2540 and IRF840? Does it go into thermal disequilibrium?
Silviu

 

[Salas]

It can be the 840 if it was working under much heat. About common sink if the MOSFETS are well insulated and the heat is not excessive it should be alright. I mean if it is a build's basic configuration problem and not a part's failure it wouldn't work well for weeks in the first place.

 

[savu]

Thank you Salas,

The temperature gets a bit hot... I’ll use a bigger heatsink and change IRF840.

Regards,
Silviu

 

[Salas]

Did you check its insulation to the sink has not gone conductive or measured if it shows few Volts Vgs or not? If not, its certainly broken.

 

[savu]

I checked it and it’s ok. One thing that I forgot to mention... I don’t have any thermal paste in there. That might help a lot to reach thermal runaway. I’ll try to find some paste today in the computer shops around Dublin.

Thank you Salas for your wonderful design and fast replies

 

[Salas]

Read their compositions because some CPU thermal pastes contain conductive elements

 

[merlin el mago]

Salas it's ok that measurement of 38mVAC after the SSHV2 using RMS meter?

 

[Salas]

Better use the scope (with short ground probe tip attachment) to see what that AC indication is about in that installation. EMI/RFI on non-shielded probe wires? Instability? Hum?

 

[merlin el mago]

What position scope 10mV/cm & 0.2ms? Probe x1 or x10?

 

[Salas]

Probe X10 to have bandwidth. Vertical set at the most sensitive, horizontal just go up and down to find something or not, but 0.2mS is a good starting point for wide view.

 

[merlin el mago]

See attached video scope set 5mV/cm & 0.5ms

YouTube

 

[Salas]

Looks like LF instability or even hum maybe. Go 5mS to see better.

 

[merlin el mago]

ok

5 ms

 

[Salas]

5 ms

Shows four squares period = 20mS = 50Hz = hum (can be groundloop or transformer field or whatever, PSU ripple its not as it would be 100Hz, review the build's arrangement, wiring scheme, and wires dressing)