Simplistic MosFET HV Shunt Regs

[Salas]

Is IRF840 having its source on ground's side? Difficult to picture bcs sandwiched. That's voltage section for instance.

 

[buzzforb]

Metal backing to inside of pcb putting source at ground.

Am i right that no voltage is found across outputs because the current is not finding ant impedance that would cause Vdrop. I have larger than 500R pot. Could that be causing problem?

I have it set to read 500r across pot.

 

[Salas]

To develop voltage, the JFET's current should flow to the 68K resistors through the LED. The trimmer sets that current. Is its middle leg connected to its side leg?

 

[buzzforb]

Yes. I will try another. Is it possible toreverse a pot? Never considered that?

 

[Salas]

The trimmer must travel from zero to max between Jfet's S pin and LED. If you reverse, still working as rheostat though, it should increase Vo anticlockwise.

 

[buzzforb]

last worrisome post for the night. Could the 840 have been blown in the intial smoking and be causing problems? It is the only thing that remains unchanged. I even changed the LED. Still just getting 20mV across output, LED and Jfet Vds. 160mV Vbe of bottom bjt.

 

[Salas]

Very possible 840 is out. Should give about 4V Vgs if healthy.

 

[buzzforb]

Aha! Naturally it was the last thing. It was actting as short?

 

[Salas]

Could be. If it will work by changing it, possibly.

 

[vgeorge]

To be sure, and in the end saves time, is to take out all the semis and check them. Zeners and leds sometimes, could go south also.

 

[dadedady]

Because your pcb is like Salas but reversed, take out all fets, trans., zeners, led and solder them on the coper side. Try again. Perhaps you have change the orientasion.

 

[buzzforb]

It was the 840. Apparently it too was fried in the original smoking. For all who smoke this thimg, zi ended up burning thr pre filter resistor, the two ccs fets, q1 gate resistor, and 840. I think vref section was ok and probably protected by 840 that got fried. My guess was right about the 840 acting asopen circuit, passing all current rather than acting as voltagr controlled resistor as it should. Thats also why there was a small voltage at outputs and across fet and LED. The Vref section was getting current, but the 840 was keeping it from getting enough to function and ultimately control current through 840. Circle of life! Ended up with max Vrop across 5k test load of 199V, which works out to about 40mA. I assume final adjustment will take place when hooked up to tubes. Thanks for the help!

 

[AndrewT]

Lets talk theory of operation perhaps understanding better will help me solve it.could you explain functioning of the entire Vref area including bjts and jfet.

The best sense I have read in a long time.

You can build the CCS section alone and test it.
You can then add on the shunt regulator and short the output to test the combined assembly.
Remove the short to test the Shunt.
Add a load to further test the shunt.

At all these intermediate stage be very careful to only apply sufficient input voltage to test the section that has been added. eg ~30V will test the CCS alone.
The same 30V will test the CCS+Shunt with the output shorted.
etc.

 

[buzzforb]

The best sense I have read in a long time.

You can build the CCS section alone and test it.
You can then add on the shunt regulator and short the output to test the combined assembly.
Remove the short to test the Shunt.
Add a load to further test the shunt.

At all these intermediate stage be very careful to only apply sufficient input voltage to test the section that has been added. eg ~30V will test the CCS alone.
The same 30V will test the CCS+Shunt with the output shorted.
etc.

It sometimes hard to do in the middle of a build. You just want to get it done and listen.

 

[Salas]

So you made it in the end. Congrats.

P.S. The 840 should have acted as a short, not open circuit. Maybe a fuse double the mA of projected CCS after the rectifier is a good protection against faults, at least for the filter parts.

 

[buzzforb]

P.S. The 840 should have acted as a short, not open circuit. Maybe a fuse double the mA of projected CCS after the rectifier is a good protection against faults, at least for the filter parts.

That is what i was trying to say. It aint working until the tubes are glowing!

Thanks for the patience.

 

[king4joy]

Tea-Bag:

Please count me in for 1 SSHV2 board soon as it is available, dude....

Do you take payment by PayPal??

Do you stock the semi and other parts for the SSHV2 as well??

 

[alphaigor]

sshv and srpp

Hello
Please ,can someone tell me if only one SSHV regulator can be connected to SRPP configured circuits of both channels ,without sonic degradation or each channel must be connected on dedicated SSHV regulator ,one Salas for left and one for right.Until now I have built dedicated regulator for each SRPP channel and didn't try single regulator for both channels.Anode followers are working great on single regulator but I have no experience in supplying both channels in SRPP configuration.Maybe such question is already answered on the forum but I can not find it.

Thanks

 

[Tea-Bag]

King4joy - I have you marked down for one.
I will be ofering Semi's at a minimum. The rest of the stuff-will wait for more feedback once the GB thread starts.

 

[Salas]

Hello
Please ,can someone tell me if only one SSHV regulator can be connected to SRPP configured circuits of both channels ,without sonic degradation or each channel must be connected on dedicated SSHV regulator ,one Salas for left and one for right.Until now I have built dedicated regulator for each SRPP channel and didn't try single regulator for both channels.Anode followers are working great on single regulator but I have no experience in supplying both channels in SRPP configuration.Maybe such question is already answered on the forum but I can not find it.

Thanks

There is nothing special to SRPP. You make one regulator and you just try. In general we can't predict anything without any special info on each application, and we guarantee nothing.