HPHV regulator

[ackcheng]

Hi,

While I am searching for a good circuit for tube preamp, I came across Allen Wright's design and also bought his book about super regulator which happens to be a shunt regulator. Then further searching on the web leads me to here

www.euronet.nl/~aespreng/Shuntregv11d/Manual.pdf

A very well organised and describtive manuel on HPHV shunt regulator which also uses the same idea as Allen's super reg

I wonder if anyone had built this before and have any comment on it? I am ready to build one myself. My question is that whether it is possible to modify the circuit to produce 4A at 550V

I noticed that IRF830 has a capacity of 4.5A

Should I reduce R6 to 0.5R, then the max current for the CCS should increse to 2.5v/0.5R = 5A? Would this work if I use a huge heatsink?


Can I parallel a few more IRF830 to get a higher current in the CCS part?
Can I parallel a few more IRF830 to get a higher current in the shunt part without using tube to increase the shunt current?

[ackcheng]

Sorry, Current and voltage requirement should actually be

170V / 4A

[Alain Dupont]

ackcheng,

IRF830

well, 500volts and 4.5 amp and 74 watts is way out the picture even with an external device...

He gives a little explanation at the end...
Ishunt should be at least 0.2* Iload; lets use 40mA (0.040A). Pshunt = 450 * 0.040 = 18W.

in your case 170 volts * 4A ILoad ==> 680 watts ; its Huge
you would need a MosFet capable of 2000 watts dissipation
and a heathsink to dissipate more than 700 watts....
The transfo would be 1000*3 = 3KVA !!!

Better try an other idea...

Could you tell why you need 170 volts and 4Amps ?

Alain.

[ackcheng]

It is an OTL design with 4 pcs of 6c33. 4A is actually the peak current not the RMS.

[ackcheng]

"in your case 170 volts * 4A ILoad ==> 680 watts ; its Huge
you would need a MosFet capable of 2000 watts dissipation
and a heathsink to dissipate more than 700 watts....
The transfo would be 1000*3 = 3KVA !!!"

if the Iload is 4A, doesn't that makes the Ishunt 4 x 0.2 = 800ma?

power that the mosfet need to handle would be 170 x 0.8 = 136W?

[Alain Dupont]

ackcheng,

I am working on a SE 6C33C-C 18 watts ... little world ! {Simplex from Ari Polisois}

In your case if you want dynamics, the power supplies must have a lot
of reserve and a PS 4*200 mA in average would have to pass 3 or 4 times this
in peak... so 800*3 ==> 2400/3200 mA
{Its pure class A and the transfo must have room about 2* to 3* average current}

You understand why there are no regulators on such configurations
as OTL with 4 * 6C33C-B's amplifiers ... but huge 1600 microF 250Volts cap's

For my SE the winding for the B+ will give 270 Volts at max 0.9A {average 250mA}
that's only for 1 tube !
I'll have 2 * 330 VA transformers ; 1 for each monobloc.

Alain.

[ackcheng]

Thanks!

The Iload = 4A is actually the peak current already. Does that mean it is still better to have more researve than the theoretical peak current?

[Alain Dupont]

ackcheng,

No no need to go further than the peak current.

0.8 to 3.0 Amps peak = about 170*3 = min 510 VA to 680VA by channel
only for the B+

42 watts for eah heater = 4 * 42 * coeff 1.5 ==> 168VA to 250 VA

You are already at 930 VA by channel. + Add some more for the driver stage

Did you take a look at heatsinks for about 130 watts ... without a fan
it's huge. Plus the MosFet must be about 5 amps average and 175 watts...


But now, you understand that it's huge to regulate even 3 amps at 170 volts!
Good luck in your project.

Alain.

[Alain Dupont]

ackcheng,

I have been thinking about your project, and my opinion is that a "shunt regulator"
will be detrimental, you need to let pass as much current as possible
otherwise the dynamic will be stretched!

You can start without it, at first ; them insert it in the circuit to compare.

PS: reserve the "shunt regulator" for the driver stage.

Regards.

Alain.

[ackcheng]

Thanks! When i think about it the problem also arise when the 6c33 does not drain the peak current --? All the current will be taken up by the shunt and the power requirement will be much higher!