audio-talk

Nick wrote: ↑Sun Feb 12, 2023 4:25 pm Well it was just the triode and cf that the ecl would do. No big gain as the same number of envelopes ends up getting used. I a bit confused though you saying the 807 doesn't need that heft when you are using a cathode follower anyway.

That is so that the ECC81 has an easier time in the face of the feedback because the CF sort of relieves the ECC81 of the heavy lifting. At least I think that's what it's doing.

There was a lot of controversy in the past with people (chiefly those on DIYAudio) arguing that the bare ECC81 or any triode for that matter was unsuitable in the face of plate-to-plate feedback coming from a power valve. They suggested that a pentode driver such as the 6AU6 as used by Gary Pimm in the push-pull Tabor would give better results.

The CF seems to work for whatever reason, giving far cleaner sound (to my ears) in this kind of amp, than the ECC81 did on its own in past RH type amps I built.

Yep it's just you seem to be saying "you don't need a cathode follower where I am using a cathode follower". A pentode cathode follower would make the first triodes life even easier. But tbh it's the feedback that's giving it the most work to do.

simon wrote: ↑Thu Feb 09, 2023 9:25 pm DIYHiFiSupply?

What happened to Alex Kitic? I recall Paul sending him stuff as a collaboration but don't remember anything after.

Alex got a job in a different industry and is very busy. I chatted with him online not too long ago. His last amp was a RH6080 headphone amp in 2018.

Thanks for the update

I’ve decided that this design by Kitic works really well. Sounds great with my pair of Hammond 1627SE output transformers shoehorned onto the top plate: They only just fit, leaving a 1cm gap between the shoulder of the 807 bulb and the end bell, but there are no heat problems.
It really could do with a bigger chassis. The one I’m using is an old 17 x 10 x 3in one from 2008. A 17 x 12 x 3in chassis would be more suitable.

Cressy Snr wrote: ↑Sat Feb 18, 2023 8:41 am It really could do with a bigger chassis. The one I’m using is an old 17 x 10 x 3in one from 2008. A 17 x 12 x 3in chassis would be more suitable.

You have a PM.

The intention had been, that I wasn't going to do any more work on this particular amp. However, the anxiety has been giving me a hard time lately and there's nothing like the smell of solder and the risk of electrocution to concentrate the mind and compel one to be present in the moment, at least for a short while.
So I got to thinking about the 6B4G mono blocks I built around 2011 and the great sound that came out of them and thought it might be a good move to incorporate a voltage regulator in the power supply. All last week I planned it out and finally today I got the job done:
To say it made a difference would be an understatement! It's funny how unreliable the aural memory is. Blinking' heck, if this is how the mono-blocks sounded back in 2011, then Gawd knows what I've been buggering about at since.
Alex, originally intended this amp to be a 2A3SE hence the presence of the cathode follower driver. If at some point, I were to do it as a 2A3 then I will have come full circle. Crazy man.

For the sake of completeness, here is the new power supply: It was knocked up with the components I had and nudged and cajoled into performing properly. Had to use SS rectification to get enough HT to keep it in the zone.

And the audio circuit with all the voltages measured:

Both my pairs of used 807s have seen a great deal of action and I've noticed, over the last few days, the sound deteriorating, turning a bit fuzzy and vague. This is unfortunate as I have no more 807s.

Also it has been difficult to find anything in my stash of old used ECC82 and ECC81, that were not past their best. I did manage to find a NOS Brimar 13D9 with two decently matched triodes in it, and giving a healthy output from both sides. The best ECC82 I have, with enough welly for the CF position is a new production, Electro Harmonix ECC82EH.

Alas, with no more 807s at my disposal I've ended up completely redoing the output stage; sockets, heater wiring, the lot, in order to accommodate standard 6L6GC types. A right pain in the arse, with much cursing and burned fingers!

I have four new production Tung Sol reissue 7581A and two of these are now in service as the output stage. The only NOS valve left in the audio circuit is the Brimar 13D9, but needs must.

I also took time out to fix the value of the potential divider to the grid of the error amplifier in the series pass regulator. I don't like pots in those kinds of situations with DC sitting on them. So the grid bias of the error amplifier is now fixed, resulting in 277V of output from the regulator, giving 258V at the plates of the 7581As. Perfectly serviceable in this circuit.

The bias on the 7581As has been changed to -14.5 volts by the addition of a 47R series resistor to the existing 150R cathode arrangement. This is as close to the book values for operation into a 2k5 load as I can get. This mode of operation gives, in tetrode mode, the lowest 2H distortion out of the three suggested scenarios on the 7581A data sheet.

Revised schematics:

Why not feed the grid CCS from the regulated side to reduce dissipation in the CCS and give it a cleaner supply. Or even just use a resistor as you know the supply is quiet.

Nick wrote: ↑Fri Mar 03, 2023 4:31 pm Why not feed the grid CCS from the regulated side to reduce dissipation in the CCS and give it a cleaner supply. Or even just use a resistor as you know the supply is quiet.

Thanks Nick, I would never have thought of that in a million years.

This is the state of play now:
It's very black. I'll get the feed to the g2 VR tubes sorted as per Nick's advice over the weekend, probably with a resistor. It'll be one less set of components to worry about.
Needless to say, now it has some half decent valves in it, the amp is a lot happier.
I had to refit the PC fan guard mounting arrangements in order to put the standard octal sockets back in place. They do look nice, but it was an absolute b'stard to do. Had to take out a section of the outer rings at the rear, to avoid them fouling the end bells of the output transformers.
Also I could do with a couple of better rectifier diodes than the IN5408s I have in at the moment.

OK, so here is the modified schematic: As I said earlier I'll sort it all out over the weekend. Should be good.