All P-channel Regulator For TSSA Split Supply?

[Jay]

Originally Posted by RJM1

The problem that I see with this arrangement is that the negative rail and the negative regulated supply do not share the same ground.

Hmmm... correct, they don't share the same ground. Why is that a problem? (+v changes related to ground = ground changes related to -v)

[RJM1]

Let’s say for an example that you have an amplifier with output rails of +-45 V and you want to regulate a +-15V supply from these rails. On the positive side you would have +45, +15 and ground, On the negative side you would have -15, ground, and +30V only letting you use -15V for the output rail.

It’s not a problem if you are using a single negative supply and you do not want to use the unregulated supply. But with center tapped transformers and dual unregulated supplies the unregulated +- voltages are referenced to ground so you cannot float the unregulated negative ground.

[jan.didden]

Originally Posted by Jay

What is the problem of using the same transistor regulator for both positive and negative supplies?

Mosfet complementary pairs rarely have good complementary characteristics, but such regulator as attached is rarely used. Why?

Are you suggesting that in some way it would be good to have exact complementary devices in a + and - regulator??

Anyway, what you want would work with shunt regs. Shunt regs are two-terminal circuits so can be used in the + or - position without modifications.

Jan Didden

[Jay]

Originally Posted by RJM1

But with center tapped transformers and dual unregulated supplies the unregulated +- voltages are referenced to ground...

I'm not sure if I understand you, but from the picture in post#1, there will be 2 secondaries, and they are not center tapped. Will there still be a problem?

[tiefbassuebertr]

Originally Posted by janneman

Are you suggesting that in some way it would be good to have exact complementary devices in a + and - regulator??

Anyway, what you want would work with shunt regs. Shunt regs are two-terminal circuits so can be used in the + or - position without modifications.

Jan Didden

Yes. An example you will find by the commercial phono RIAA head amp PP2 from NAD - go to the second attachement by post #5 about
Head Pre for Denon DL 103

[Jay]

Originally Posted by janneman

Are you suggesting that in some way it would be good to have exact complementary devices in a + and - regulator??

Yes. Something I'm missing?

Originally Posted by janneman

Anyway, what you want would work with shunt regs. Shunt regs are two-terminal circuits so can be used in the + or - position without modifications.

High current shunt regulator??? I haven't used or even seen one. Please give me a reference to a shunt regulator capable of around 2A. May be not an efficient solution but I'm after a good sound, so why not.

May be I need to regulate the front end only with a shunt regulator (with or without a capacitance multiplier), in line with Nico's suggestion (I forgot that in my circuit I lowered the front end voltage rail, which has good effect, with RC filter).

But I really like the above regulator in single ended circuits. If RJM1 or somebody else give a "no-go" with the same regulator for both + and - rails, I will use IRF640 and IRF9540 as used by Ayre V3 amplifier (I have plenty of IRFP460 but only a few IRF640/IRF9540 pairs so I tried to avoid using them in power supply. But I think I have no use for them in an output stage anyway).

EDIT: In the Ayre, the output is not regulated with the IRF640/9540, and I remembered Samuel Jayaraj comment about current hogging in paralleled mosfets, so... I wish the post#1 circuit has no issue...?

[jan.didden]

Originally Posted by Jay

Yes. Something I'm missing?

I don't know - I wasn't sure you were serious about your priority to have perfect complement fets in the + and - reg.
I guess you were...

Originally Posted by Jay

High current shunt regulator??? I haven't used or even seen one. Please give me a reference to a shunt regulator capable of around 2A. May be not an efficient solution but I'm after a good sound, so why not?

There's lots if circuits around, if you don't want to design your own. Google is your friend.

jan

[tiefbassuebertr]

Originally Posted by Jay

Yes. Something I'm missing?

High current shunt regulator??? I haven't used or even seen one. Please give me a reference to a shunt regulator capable of around 2A. May be not an efficient solution but I'm after a good sound, so why not.

As you can see in the NAD PP2 schematic, the shunt regulator consists of two parts:
1) the constant current source so as
2) the zener diode

If you want a 2A shunt regulator, you must design an appropriate current source for 4A (have a look to the CCS from the single ended amp "ZEN" from Mr. Nelson Pass) so as an appropriate power zener diode (normal zenerdiode include Sziklai-Darlington or MOSFET e. g.) also for 4A.
By normal operating you need 2 A for your amp and 2 A for the "Power Zener" itself.
Please note:
In case of failured amp, your power zener must be able to work with the full current of 4A.

[Jay]

Originally Posted by janneman

There's lots if circuits around, if you don't want to design your own. Google is your friend.

I couldn't find it.

Originally Posted by tiefbassuebertr

As you can see in the NAD PP2 schematic, the shunt regulator consists of two parts:
1) the constant current source so as
2) the zener diode.

Can you show me how to do it for split supply +/-45V capable of at least 2A? I have built many zens, and have googled as well, but have no idea but the current source in post#1 which seems to have "con" from RJM1 and Janneman.

Of course, I'm not a new EE graduate who likes to design sub-quality circuits when much better circuits are free from experts. Hence my preference will go to shunt regulators used in respected amplifiers (not my own design, even if I could). But I cannot find such regulator in my collection of amp circuits, so I guess it is not feasible?

[Nico Ras]

Originally Posted by Jay

I'm not sure if I understand you, but from the picture in post#1, there will be 2 secondaries, and they are not center tapped. Will there still be a problem?

There will be no problem. It is like using two batteries in series and V/2 is the virtual ground reference.