Simplistic MosFET HV Shunt Regs

[iko]

Didn't leave it powered long enough to measure, but I used R1 = 15R

 

[Salas]

What is the Vdrop of your LEDs?

 

[iko]

What is the Vdrop of your LEDs?

In total for the three LEDs I measured almost 6V.

 

[Salas]

You did run 200mA through the CCS. Add a 6.8R in series with your 15R. With given LEDs, that will tend to run around 140mA. See if it holds better, or there is a mistake somewhere else. Measure the drop across the set resistors combo, so to check if you are in the ballpark for CCS current.

 

[iko]

Thanks salas! You've earned much beer already! If you ever come this way, or I your way, you get free drinks

 

[Salas]

Canadian Club I always liked.

 

[iko]

Me too! And there's plenty of it here.

 

[iko]

Don't know whether the Maida is good for 120mA... see post Nr. 138 in this thread.
Mine is working very well at 55mA and gets barely warm to the touch (heatsink of IRFP450 that is).

Edit. just saw that Salas replied already

stixx, thanks. If the resistor between the mosfet and lm317 is removed, I think the Maida should be able to do quite a bit more current. It works in simulation.

 

[Salas]

Hmmm, wondering if it will get the on transient well over many operation cycles. Its the most sensitive part of the Maida there.

 

[iko]

I got a weird behaviour with the hv shunt. The non-standard components are:
* irfpc60 instead of the 840
* 2n3906 instead of the bc560b
Everything else is just as the schematic. For the output a 1k resistor. I hooked the regulator to a variable power supply, and started to raise the voltage from 0 up. If I raise the voltage slowly, most times it goes up at the output as well, just fine. But if the input voltage goes up quickly, the output voltage stays stuck at about 15-17V. Any idea why this would happen?

 

[SpreadSpectrum]

stixx and ikoflexer,

I have a maida that powers a stereo kt88 push-pull amp, 200-500mA at 460V. It does indeed work. Not as well as the shunt, I am sure, but efficiency would be a huge problem here.

 

[iko]

stixx and ikoflexer,

I have a maida that powers a stereo kt88 push-pull amp, 200-500mA at 460V. It does indeed work. Not as well as the shunt, I am sure, but efficiency would be a huge problem here.

That's very good to know, thanks!

I replaced the 2n3906s with a couple of 2n4402s and the problem seems to have disappeared. Next, let's see what it does in the amp

 

[iko]

Just curious, which version are you using with the kt88 amp? Have you made any changes to it?

 

[Salas]

Just nice to see that this shunt does not blow with parts I have never had used on it.

 

[iko]

Just nice to see that this shunt does not blow with parts I have never had used on it.

Well, let's see, a lesson not to pay attention to American Idol when your wife says "watch this" Last night were sacrificed an irfpc60 and an mpsa92 feeding 256V into the output of the shunt Furthermore one irfp9240, one irfpc60, and one mpsa92 were burned on the altar of Shunt when trying the jfet trick.

 

[Salas]

Did you slip a probe?

 

[iko]

Did you slip a probe?

Just after making sure that the shunt works ok while on the bench supply, I stuck it into the real amp in between the tubes and psu. The problem was that I was paying more attention to the TV and wired the shunt in reverse, input to output and output to input

The shunt is working now, but I cannot say how well yet, ran out of time last night.

 

[Salas]

Lets see how it sounds with lower hfe PNPs and much higher Ciss shunt Mosfet that you use. Did your trafo hold at 140mA in the end?

 

[Salas]

The shunt is working now, but I cannot say how well yet, ran out of time last night.

Try for your shunt and after it will work OK, it normally lends a tone quality that you will never give up.

 

[iko]

Lets see how it sounds with lower hfe PNPs and much higher Ciss shunt Mosfet that you use. Did your trafo hold at 140mA in the end?

Yes, the trafo does hold with the proper circuit. Initially the problem was a fried irfp9240 which just wasn't working right. Sound wise, the initial impression is good.