Rectification: solid state Vs. valve.

[andrew ivimey]

IMHO that there really is no necessity for valve rectification to be sonically better than solid state rect.

Valves look better and perhaps there is an aesthetic design principle, i.e. if the amp is valve, the PSU should be too.

Now a **** SS rect. is going to make and amp sound worse, but a well designed SS rect. surely, is not.

And one point on favour of diodes is that you can get a much higher HT out of them than with valves.

However, valves do not like big capacitors. Big capacitors can be very useful in helping a more constant supply of current from the PSU and you can use relatively whacking big caps with solid state that valves won't put up with.

OTOH, and this is where I am asking the people who know hugely more than I do.

Do large caps (with diodes) actually put more stress (resistance or draw of current) than smaller caps??? My feeling is that this is so.

Diodes (FREDs or your bog standard silicon diodes) seem to do this anyway.

It seems to me that lower values of caps (or even no caps at all!) are more desirable than adding more microfarads.

[BOONDI]

Hi Andrew, all.....

I have been thinking about this for a supply that I was re/designing for my headphone stage. There are advantages to large caps such as lower ripple. The disadvantages are higher inrush current to charge the cap at switch on. There is another school of thought that says multiple smaller caps should be used in parallel to get the value needed. This has the advantage of lowering ESR and inrush current demand. It is also claimed to be better sounding. Knowing you, I suspect an experiment or two.......

[Darren D]

I made a PS supply for my direct coupled SE 6B4G (2a3) amp with just one choke and a single 10uf cap.

There was no hum at all. I started with three chokes and three 10uf caps which were removed one by one.

Looking back I'm still not sure why it worked so well, but it did.
Lovely amp, only mared by blowing the output valves at an alarming rate. From memory I was going through one about every 3mths.

In comparison my cap coupled PX4 is still using the same valves 3 years on.

But the original point I was trying to make is, experiment, you may surprise yourself.

Regards
Darren

[andrew ivimey]

Interesting points Darren, of course PSU problems aren't just horrid HUM, caps and chokes cut out nasty noise / hash too, affect curren etc etc.

Adrian, I am tired of experimenting (winsome worn out smile) a) doing silly things and b) when I hit a vein of gold, am just reinventing someone else's wheel & c) if not actually mixing metaphors, certainly combining them.

On HUM, I have been very surprised at what gives hum and what does not, just how much you need to cut out hum and just how little; so much so I can't find rhyme not reason.

So, when I got home I reduced the caps hugely on Cheapnissmk7; from cap/choke/cap = 200mfd/7Henry/470mfd to 30mfd/7Henry/60mfd. No Hum sure, but a very different tone - not balanced, or not as I like it.

The HT will have changed, and I haven't measured that. Is this tonal imbalance the amount of mfds or the HT (or both)?

[Richard]

Hi Andrew

I've come to believe that whilst the rect type will contribute to the supply "sound" what follows the rect may be more important in a hifi amp where we really should be looking for a near-perfect supply.

The better SS diodes are now very quiet switching. A couple of years ago a few of us listened to 4 types and there really wasn't much between them. They were all feeding a single bulk cap for a pp stage. Conversely swapping valve rects nearly always has a big difference in sound.

I also think that as no supply is perfect some people will always prefer one type to another. This is far more accepted by guitar amp users who simply accept that their supply and rect type changes the sound and use it as part of their instrument. Some will even use a variac to run the amp at the point they like.

Rich

[NealG]

One other aspect to consider with large cap supplies vs small cap/valve rectifiers is the current pulses that large cap supplies have on the mains supply affecting anything else on the same supply. Small caps with a large choke will pulse far less and have a more continuous draw...

[Paul Barker]

Yes and of course quality of mains, not entirely off topic, is a massive contributing factor.

There was a guy talking to us at Heathrow who maintained he has 70v potential on his incoming earth.

This would sound impossible, so I asked some pertinent questions and he it turned out replied competently, he is a sparkie it turns out.

So there are all sorts of things going on in that swc that enters your house.

the supplier has done nothing to help. His neighbours are probably blissfully unaware.His explanation is some sort of factory influence nearby on the same supply.

[Richard]

Davy may have an update about his KEL84 noise for us soon along these lines Paul.

Rich

[andrew ivimey]

All thought provoking stuff, thanks. I have limited myself in size with the present spud amp as I am trying to keep Cheepniss v.cheap. And then I want superb quality and of course a dozen upgradewishedfor tweaks occurr to tortured mind and then I need to spend more.

Well, I'm not, not yet anyway but this severely limits what I can do to the PSU. At present it is all built into one of those Maplin ally boxes that would easily accomodate say WAD pre or phono.

If I am going to have something the size of a two stage SE amp, with all quality metalwork I may aswell have 45, 2A3, 300b, and I already have!

Spud/Cheeniss has to be spud sized and magnificent, cheap and reliable - what a design brief!

.... though there is always room for lateral thinking ...

[Ianm2]

well sort of, yes, large caps will stress everthing that goes b4 it, ( you have to think where its coming from ) ie, the rectifier, and the mains tx, as they are drawing much more current, so that translates to heat.
Its actually quite amazing that these small rectifiers, some of which can take 50 amps.

The basics are fairly simple, but to get into it, requires degree maths.

I suppose if you have a purist se amp, it makes sense for a valve psu, there is something rather elegant, simple and right about all valve.

however diodes are still simple to implement, and don't suffer from cap limitations as much, and psu 'sag', which is a lot of the sound of a distorting guitar amp, an awful lot happens in an amp when the rails sag, and its not easy to find a full treatise, harmonics, compression, distortion and all that...

in the billies I had, still have a very soft spot for those, I tried a 2 diode rectifier, and didn't like it at all, it took away all the lovely valvey signature, but I didn't drop the volts etc., so it probably went really high, I suppose like for like, I should have made the same volts, I know a little more now, so may be able to try it again one day properly.

if you look at most of rankin's amps, they are valve psu's, whilst he can be a tad dogmatic, he has done loads of amps, and this is one area I am sure he would have put diodes in if they were better., by the by, I think some of his commercial ones may use diodes, but not sure.

anyway andrew you're a doer, so keep on doing and finding which you prefer, that's all that counts

oh and btw, I am pretty sure a lot of the problems with valve amps, buzzing and all that is due to ss rectifiers