Simplistic MosFET HV Shunt Regs

[Stixx]

Some says tube rectification will cause the input voltage variation to *any* regulator may result in it dropping out of regulation. That's why SS rectification is more preferable.

I am not Salas, but I can assure that all Salas shunt regs are working perfectly with tube rectifiers.
This has nothing to do with solid state or tube... a tube rectifier drops a certain voltage , so do take that into account.
Get PSUD2 (free) and start fiddling.. just choose your Vin at least 20V higher than your Vout and you won't have a problem...
Regulating to 400V might be borderline and has been answered numerous times, but I'll leave that one to the Master.

 

[buzzforb]

Due i need a load resistor totest sshv2 and what value?

 

[vgeorge]

If you don't know your target load, 15k is a safe value. Be sure to be more than 5w, better 10w.

 

[euro21]

Testing dual SSHV2.

Raw supply: 240V ct - AZ1 mesh- 5.4uF -10H (100R) - 47uF -3H (80R) - 220uF -> 254V
20mA CCS, 7mA load (#26 preamp), 220V.

Noise below 1mV (in the other side 2.5mV)!

 

[Salas]

Post some scope screens if you can. You measure on preamp's output, or on the rails? Maybe you got some more 50Hz on the 2.5mV channel? Twist output pairs.

 

[euro21]

My measuring was on dummy load (30K -3*10K 5W) with dual channel (Kenwood VT-176) AC millivoltmeter (330uV-100V).

 

[Salas]

Ah, OK. Move the mains cable away maybe. Nice proto board work. Do you use SA1381 and SK117? How's drift?

 

[euro21]

Yes, SA1381 and 2SK117GR (3.7mA, 3.92mA).

 

[Salas]

OK, fine then. Use 27-30mA CCS if you can. So 20mA goes to the reg alone. Which one of two trimmer you had substituted for fixed resistor? I think I see less than 4 trimmers on the pics.

 

[buzzforb]

With variac turned all the way down and only 1.5V on the rail, i already have 22mA of cuurent across test point on sshv2.

 

[Salas]

You should have current, CCS works from low voltage all the same.

 

[buzzforb]

If i increase the rail voltage, the vdrop across test resistor increase in proportion. At ver low rail voltage, i am already dropping 500mV across test point.

 

[Salas]

Go slowly and see if it stops growing appreciably at a point. Back up the current trimmer a bit before that.

 

[buzzforb]

Was able to back down the drop across test point, but with inly 16V out of prefilter, i am dropping almost 12V across 220r resistor on prefilter. 500mV across R1, 3.5 across r2 of sshv2 is yielding 160 mV across r5. Diode ,easuremnets reflect gate voltages.

When i first fired it up, i fried r1 along with 220 on prefilter. Did i fry the q1 in the process?

 

[Salas]

Probably yes.

 

[buzzforb]

Ok. Done for the night.Thanks.

 

[Salas]

Check it out well tomorrow with a clear mind. After enough got it going, its not impossible. There must be a bug somewhere.

 

[buzzforb]

Yup. For the gate resistor to q1 to fry, i cant imagine q1 not being toast. Changed out resistors and zeners just in case. Will do q1 tomorrow. Considering that at such low voltage the fets are passing do much current, it is the only posibility i can think of. Hey, wheres that gun?

 

[Salas]

Stay away from guns. Such Q1 stuff happens in best of families even.
Check all your self made PCB's connections against schematic with a clean mind, use a 47uF capacitor directly across DCinput after you had exchanged all parts suspect as hit.

 

[Salas]

*Make sure also you did the 4 wire output connections to test load well BTW. Very important.