PSU for Gainclone, regulated with center-tap

[coolbeer]

As I am planning a gainclone build I need a powersupply, I am using the transformer from a Sony TA-VE610 at 240W.
The transformer delivers 30-0-30v AC on the main rail, giving me about 42V after rectification and smoothing. That's a tad much for a 3875... So I need to drop that to about 35v, this is where it is getting funky.

As we have a transformer rail with a center tap, we need to use one positive and one negative regulator(planning on a 317/337 combo), since we are going to be playing music really loud, the amps are probably passing the 1.5A by a long-shot, so we need to have the 317/377 combo drive a power transistor.

I do have a few 2SC2625's, should be ok in this setting. And I have made a schematic that I don't think contain too many faults, but I'm sure they are there.

If someone could look over this and yell at me if I did something utterly stupid I would appreciate it very much.



Have a go at that and try not to laugh and point too much

[stormsonic]

LINK

[woody]

Might I subjest building the raw power supply and measure it first. If your
transformer is 30v 0 30v the voltages after the bridge will probably be ~40v.
Yes 30 x 1.41 = 42.3 but you lose ~2v in the bridge rectifier. If you still want
to lose a few volts a pair of cap multipliers using mosfets might be the way to
go. For this type circuite you might look at the the Zen articles.

[coolbeer]

I was always planning on running the gainclone on a regulated supply, I have been reading up on it and it seems it will sound more "controlled".

Been reading about zener diodes and driving transistors with them instead of 337/317 combo, the only thing I'm unsure about is the negative rail, is it basically the same as the positive? My guess would be reverse collector/emitter, is this correct?

Thank you for your views/inputs, really appreciate it

-
Kolbjørn

[AndrewT]

you MUST NOT use a dual rectifier with a centre tapped transformer.

[AndrewT]

Originally Posted by woody

.....transformer is 30v 0 30v the voltages after the bridge will probably be ~40v.
Yes 30 x 1.41 = 42.3 but you lose ~2v in the bridge rectifier.

Hi Woody.
the 30Vac transformer has regulation, possibly around 6% for 240VA. If it's an EI it could be even higher.

Let's assume the transformer is a 240:30 6% regulation for the sake of this example calculation.

Input voltage varies from 216Vac to 254vac.
Maximum output voltage is 254/240 * 30 * 1.06 = 33.65Vac.
Minimum output voltage is 216/240 * 30 = 27Vac.

The maximum rectified voltage is [3.65 * 1.414] -0.5V (if the diode current is low) = 47.08Vdc.
If some quiescent current is drawn to power the amplifier then expect the maximum supply voltage to be about 46Vdc. This is not recommended for the National chipamps.

Be careful to offer accurate advice, 40Vdc estimate is way off.

[AndrewT]

Originally Posted by coolbeer

I was always planning on running the gainclone on a regulated supply, ...............
Been reading about zener diodes and driving transistors

Have a read of the Nelson Pass website. He describes in detail how to measure and design with BJTs and FETs and shows a typical Emitter Follower PSU using a string of Zeners as the voltage reference.

[coolbeer]

Thank you for saving myself from myself on this part of the forum aswell Andrew

I was wondering why I never saw any designs involving center taps and dual bridges, guess that kinda answers it...

My next attempt, without the 337/317's this time, and yes, the R1/R2 value is not calculated and probably an horrific idea.
What I'm most interested in is whether the negative rails transistor and zener is actually correct, in my mind things should work backwards on the negative side(but then again, who ever claimed that my mind was anything near right )

To boldly go where no one dared to venture before, and try things that have no chance in finnmark to work!
-Kolbjørn

[AndrewT]

Originally Posted by coolbeer

My next attempt, without the 337/317's this time, and yes, the R1/R2 value is not calculated and probably an horrific idea.
What I'm most interested in is whether the negative rails transistor and zener is actually correct, in my mind things should work backwards on the negative side

better.
Have you read PASS yet?

Delete the output resistors.
The output voltage will Zener voltage minus the Vbe drop of the output transistor.
The Zener voltage and the Vbe drop will both vary with ambient temperature and self heating temperature increase and will also vary with current passing. But, it is still much better than blowing up your chips.

The resistor value must be calculated for the various output currents and input voltages and Zener dissipation. Leave that value to last.

[ikoflexer]

Originally Posted by coolbeer

I was always planning on running the gainclone on a regulated supply, I have been reading up on it and it seems it will sound more "controlled".
Been reading about zener diodes and driving transistors with them instead of 337/317 combo, the only thing I'm unsure about is the negative rail, is it basically the same as the positive? My guess would be reverse collector/emitter, is this correct?
Thank you for your views/inputs, really appreciate it
-
Kolbjørn

IMHO, a poor regulated supply will be worse than an unregulated one. This is a power amp, so aim for a psu with low output impedance so that the amp gets its current with little change in the input voltage. That's why many people report "controlled," "tight," "dry" sound when they use a regulator. I wouldn't touch the 317 series for this amp. Just my 2c.