Super Regulator, collecting the facts

[jackinnj]

Originally posted by Elso Kwak
Is that an advantage?

I was just stating the facts. and as we know the HFE can really bounce around a lot amongst a group of transistors.

[peranders]

Originally posted by peranders

Jack, has your layout left your computer yet? Prototype is made?

Anything to show?

[jackinnj]

PA

just got back from France --

I did some HV boards -- from an old (1982) Audio Amateur design and distributed at my cost (see group buys) -- but these super-reg boards are still in the cyber-stage.

[Onvinyl]

Hi,
has someone tried and connected the superreg in a way that the pass transistor is longer a pass but a shunt?
I would love to hear if so.

The shunt trans. must be a pnp-device for the positive rail then, but how to connect the sense + feedback lines?
Rüdiger

[peranders]

I see no reason why you couldn't build a shunt regulator with the opamp fed from the regulated side.

Why do you ask? Anything special you want to use it for?

I have a hunch though that the PSRR will be _much_ less.

Why don't you simulate it and see if you can make it work.

[Onvinyl]

Hi,

Originally posted by peranders
I see no reason why you couldn't build a shunt regulator with the opamp fed from the regulated side.

Why do you ask? Anything special you want to use it for?

1) curiosity, see http://www.diyaudio.com/forums/showt...711#post135711
2) For the output stage in my phono (version of la boheme). I have a super reg (not ALW, but Jung-Didden 2002) for the output stage. It is a huge improvement compared with batteries. 'Holographic' sound and much more precision in stage, you really can tell where the instruments are placed. There is an issue, however: the bass is weak and fluffy, compared to an discrete output stage in borbely-style with simple regs.

Why don't you simulate it and see if you can make it work.

Because the soldering iron heats up faster than the computer boots
No, because I'm not familiar with those programs. I tried, but it just ates time. I might draw an conceptual sketch tonight, and get some opinions here.

have a nice day,
Rüdiger

[jackinnj]

Originally posted by Onvinyl
Because the soldering iron heats up faster than the computer boots

you need a slower computer or faster soldering station.

[Onvinyl]

Originally posted by jackinnj

you need a slower computer or faster soldering station.

Don't you mean the other way round?

Rüdiger

[peranders]

Originally posted by Onvinyl
No, because I'm not familiar with those programs. I tried, but it just ates time.

I really recommend that you test it. Many here have experience of LTSpice.

I have simulated three project with success, QRV-05 diamond buffer, QRV-06 current feedback amp and the last SST-01 softstart and for this LTSpice was _very_ helpful because the design is very unlinear and unpredictable. The simulation was nearly 100% exact!

You know you can also get distortion from the program!

You draw the schematic and when you are ready, just hook up a signal generator at the input then put an oscilloscope probe where you want the see the signal. Try as much as possible to get the simulation models for the parts you are going to use.

[Onvinyl]

Hi,

Originally posted by peranders

I have a hunch though that the PSRR will be _much_ less.

Why? Are you referring to the bootstrapped op-amp arrangement?

However, the first real useful web-source concerning shunt regs (because it names design goals and tells us how to calculate relevant values), that I found just today, seems to imply that a top-notch shunt can't be achieved by simply swapping the super-regs pass-transistors from npn to pnp and await wonders.
older shunt reg. article by walt jung
Maybe the error-amp could stay, but we need a very high Hfe transistor arrangement.
Rüdiger