Two diode split supply?

[Skorpio]

If diodes sound bad because of the turn-off/on, will two diodes sound better than four?

It would be possible building a spilt/double supply with only 2 diodes.....

Has anybody tried that?

 

[ifrythings]

I'm not sure if two diodes will make less noise than 4, but if you want the same current from the 2 as from the 4 then I would think the higher current would make more noise.

You could always put a snubber network across each diode or just a cap and that should shut them up.

 

[BobEllis]

If I understand your plan correctly you are thinking of a CT transformer with the center tap grounded and a single diode in each rail.

This will work, but probably not as well as you'd hoped. Since the diode only can conduct in for half of each cycle (rather than two diodes conducting a half cycle in a bridge), your nominal DC voltage will not be the 1.4x VAC you get with a bridge, but around .6 x VAC. On top of that, your ripple will be higher at 60 Hz. You'll need a lot more capacitance to bring the ripple in line what you'd get with a bridge rectifier.

All in all, there is no free lunch. You'd just be trading diode noise for other problems.

 

[Skorpio]

Correct....but on the other hand I can draw more current out of the transformer.....and the loss i voltage could be an advantage.

Then I am only trading bigger trafo for more capacitors (double) and are getting less switching noise.

A step towards tube rectifiers...?

 

[lndm]

A step towards tube rectifiers...?

I agree that much of this effect can be simulated with resistors in series with the ss diodes, in particular, damping ringing in the power supply components, even when the diodes are still spiking. Use in conjunction with snubbers for best effect.

Most valve diodes will have a larger resistance (50 ohms or whatever) but I have heard drastic improvements with ss diodes and resistors as small as 1.5 ohms. In this case, going much larger had apparently passed the law of diminishing returns for that circuit.