Holton's active regulator

I've been thinking to power my Pass X series DIY project's front-end with Holton's modified regulator at http://www.aussieamplifiers.com/regulated.htm

I need 400mA and +/-47V out, and I plan to put in 55..60V

I have mailed Holton twice and not received a reply. First time was monts ago, second time weeks ago. I don't know what is wrong with email, but luckily it seems he is reading these groups at least now and then.

There are some possible opinions
- Holton's regulator is overkill for this application
- It's just fine for this
- A regulator of another kind would do better
- It's not good enough

Would anyone put two of those in series, ie. one feeding another? Would that give anything in return?

-Kimmo Sundqvist
Voltage regulator

The revised version looks like a pretty good design. The schematic
is a little hard to read but it looks to me like the changes are:

Removing the resistor in series with the 470pF compensation cap.

Increasing the output caps from 10u to 100uF

Increasing the emiiter resistors from 0.11 to 0.22 ohms

Chage to MJE340 and MJE350 for driver transistors

Changing the zener circuits and filtering.

I have not bulit it but it likes like a decent design and uses good transistors. I would try it out. The design does not have short circuit protection so be careful not to short the outputs.

I agree, this is a pretty good regulator.

However, I don't like the change to the zener references. The new circuit uses a transistor to act as sort of a min-regulator itself and add a bit of temperature compensation to the reference. First, you won't need super temperature stability for an audio circuit. Low noise is far more improtant. The problem I see here is that the zener current is too low - under 1mA. This makes the zener voltage rather "soft" (on the knee of the curve) with a higher equivalent resistance. I would much rather see 5mA to 10mA on the zener for more stability and less noise.

You could also try adding a "speedup" capacitor across the feedback to improve ripple rejection. Another way to achieve this is by putting back the resistor in series with the 47pF loop compensation. But I prefer the cap in feedback. Beware, any changes can affect stability.

The last thing I would suggest is to add more of a resistive load on the output. This will push operation more into class A. Right now it is almost pure class B.

But overall the circuit is nice and can be scaled easily to your application.

Jim Hagerman