DIY Transistor Amp Using NVA Boards

[pre65]

Alas we will never know . I miss Richard .

Did Richard ever admit to being "clever" ?

[Daniel Quinn]

Pre your a stupid twat .

[pre65]

Pre your a stupid twat .

[Cressy Snr]

I deployed three power supplies today and everything fired up on the bench without incident. Rail voltages checked out perfectly and everything looked fine, but the DC offset at the output was unacceptable. It’s back with two power supplies now and none the worse for the experience. Nevertheless, the three PSUs were a fun experiment to do.

I think I’ve taken the boards as far as my abilities will allow. Madness will not be happening

[shane]

[Pedant mode]

Pre, you’re a stupid twat.

[/pedant mode]

[pre65]

[Pedant mode]

Pre, you’re a stupid twat.

[/pedant mode]

Dennis, very clever, but no idea of speeling, or punctuation.

[Cressy Snr]

Power supplies , something mr dunn could agree with.

But 3 instead of bigger va ?

In theory the 3 power supply setup should prevent each section talking to the next via the power rails, but I found a limit with these boards. RD told me in 2019 that three power supplies were possible but it could get very hairy. In the end he told me it had been more trouble than it was worth.

I was pleased it worked without incident as I’d arranged the capacitance values for each power supply to increase, the nearer to the output stage they got. This meant the charge time increased from input to output so that the sections came up one after the other at power-up. This prevented nasty noises being passed to the speakers at switch-on.

Indeed this worked as the theory said. I used a sacrificial pair of cheap drivers to test the thing but as I said earlier, the DC offset was unacceptable. The sacrificial drivers were sucked way back off their rest positions by the -300mV offset, and leaving them too long would not have done their voice coils any good. In any case, the sound with that degree of shift over would have been absolute crap. ‘Hairy,’ Richard said, ‘hairy’ it was. And as he also said, ‘more trouble than it was worth.’

OTOH twin power supplies off big transformers are the bees knees with his circuit and of course Paul and Tomasz use these in the top-end NVA options. What I’ve got now however is no longer Richard’s bare bones circuit but the 2 PSU scheme principle holds just fine even with the ‘Frankenstein’ job.

[simon]

That's an interesting post Steve, with very little experience of solid state I hadn't thought about inter-relation of multi-stage PS with SS rectifiers. Valve rectifiers have some benefits .

Looking closer at your 3 stage PS the amount of capacitance decreases the closer to the front of the amp which on the face of it is contrary to the aim of reducing ripple. I'm guessing that 6800uF is adequate for ripple, even in the first stage, and the increased capacitance in the second and third stages is for charge storage for instant power rather than ripple? Or is there something else going on?

[Cressy Snr]

Aye, the big capacitance value in the output stage power supply is for energy storage giving instant power when required. Class B amps like this one place big demands on the power supply. Less is perfectly adequate for ripple reduction, but sag would become an issue if the storage was compromised. That’s what I’ve read anyway.

Additionally the amount of capacitance has to be balanced against inrush current and the size of ripple current spikes. There is an amount that runs into diminishing returns, where the improvement is not worth the cost of the caps.
You also get half-wave rectified versions of the audio signal on the power rails in class B amps and the more welly you give it the worse it gets, so it’s worth keeping the power stage on its own exclusive supply rails. All good stuff to try and deal with. But get it right and the sound is really rather good.

[simon]

Yeah that makes sense. No benefit from a sectioned PS with say a small R between caps? I appreciate with low voltage and high current this will "cost" voltage. It's a bit of a different world from valve amps.

And for my extra merit thicky at the back of the class question might there be any benefit from a regulated supply? Not sure how much current the amp draws, and so how difficult the regulator would be to build, but maybe it wouldn't sound any better any way?

[Cressy Snr]

Most authors I’ve read (Elliott et al) say that a power amp is not worth regulating. However they also tend to agree that in solid state active preamps and phono stages, regulation is a must.
I suppose one could do a sectioned power supply with RC filters, I don’t really know how much this would benefit a solid state system with brute force cap banks.

[Cressy Snr]

I’m going back to three power supplies. However the third one will simply be used to power the built-in phono stage. Why I didn’t think of this before ripping the thing out is a bit frustrating, but it won’t take long to put it back in.

[Cressy Snr]

On second thoughts: I can’t be arsed.

[Daniel Quinn]

It should be noted new NVA as diverted from Richard , in that they advocate higher quality (price ) toroidy transformers instead of higher va .

All of there amps have lower va than Richards .

[pre65]

It should be noted new NVA as diverted from Richard , in that they advocate higher quality (price ) toroidy transformers instead of higher va .

All of there amps have lower va than Richards .

And what do you deduce from those comments ?

Is it possible the current NVA amps are "better" than those made by Richard ?