The Nemesis 813 PP Amplifier

[Paul Barker]

813 not as refined as gm70 whether because it’s a true triode I don’t know. There is barely any difference between 4 or 5 entirely different versions of 212. Copper gm70 is that cooks that it could be presented with 212’s in a blind test and the blindfolded person could easily confuse it for a 212. In today’s world the graphite gm70 will have to do, but if you don’t need the extra power I’d recommend the GEvt4c.

[izzy wizzy]

I have defeated my nemesis. However in getting it quiet, I think the sound has suffered and a bit too sterile for me now. It is a good starting point and I appreciate that some thought it sounded good at Owston.

Thanks for the suggestions re the output valves. To develop this I think the route will be GM70 as they seem to get a vote for sounding better than 813s. I didn't know about GM70s when I started this thing all those years ago but I remember thinking when I did see them, they would have been my choice. To feed this thing in the future, it is also the economical choice. Plus no top caps which lowers the danger issue a bit; getting burnt being the remaining topsides problem.

I've always fancied the VT4c but cost is prohibitive; not only the valves but a new OPTx as well. The GM70 should be fine with the OPTx I have.

They are also lower current to heat so in the range of Andrew's VCCS supplies slightly modified.

Think I'll be happy enough with graphite GM70s. I'm not into super exotic valves. I'm hoping the gain won't change much. It might be a bit less but I have oddles at the moment so hopefully not a problem.

I have Andrew's boards for the 4P1Ls fils. The rest is fairly OK for now I think. Sure I could go for a seperate supply for the front end but the RC decoupling is simple and I kinda like that.

Any hints on buying GM70s welcome. Do these look OK https://www.ebay.co.uk/itm/Tube-GM-70-G ... SwQsZc-7yH

Unless anyone here wants to get rid of 8?

Cheers,
Stephen

[IslandPink]

Blimey. I've just looked on ebay - seems I've missed the boat on coppers...

That's very very NOT OK, they were only about £120/pr until even 6 months ago.
I wonder what Yuriy can do ?
I would buy at least another pair.

[simon]

Yes, they've doubled almost overnight it seems. I emailed Yuriy last night but no reply yet.

[Paul Barker]

Yes Stephen that link to gm70 from Russia brilliant deal. Think you’ve chosen well taking into account your constraints.

People sitting on some GEVT4c’s I would advise use those.

[izzy wizzy]

As I've said a number of times, in getting the amp quiet and ready for Owston, I lost something and it made the amp very steely, bright and shrill in the mids to top and it lost some bottom end heft as well as flattening the dynamics. I only listened twice since getting back and each time it didn't last long as it wasn't good.

So in getting ready for the revised build, was seeing if I could reuse the filament transformers with the 4P1Ls at 2.1V off Andrew's VCCS supplies with 5VAC in. Thought I had better model the revised supply in PSUD. Oh my! The current supply was an oscillator creating a major amount of HF hash. Removing the final RC of 3R and 0.22u fixed it in the sim so I did that in the amps. And I'm back in business. They sound nice again and all the issues above that appeared with the DC vs AC on the driver have gone away.

Cheers,
Stephen

[pre65]

A good result.

[izzy wizzy]

One more thing before this becomes a GM70 amp, I got the 4P1Ls lit up with Andrew's VCCS supplies. Oh my! They are seriously good. I can only describe the improvement as transformative. Sat down to run through the usual cuts from records to get a feel for it but sat transfixed through to the end of the first record. Back to enjoy some more.

Cheers,
Stephen

[IslandPink]

I know - when testing these I started to think that maybe a lot of what I heard as differences in various SE amps might have just been the filament supplies.

[izzy wizzy]

It certainly counts for a lot. Although I'm still prejudice to SS in my amps, this device goes a long way to showing there are some SS things that bring out the best in yer valves without having the issue I find so annoying. Quite the opposite in this case. Listening some more last night made me realise some distortions and rubbish on records that I thought were part of the record just weren't there or diminished so much I didn't notice. There's not one area top to bottom, soft to loud that isn't better and therefore the music more enjoyable. It even seems more in time. And I can play it softer to get the same pleasure if that makes sense.

Cheers,
Stephen

[simon]

Nice. Very nice.

I've been fighting the urge to replace 12GN7 with 4P1L because of the added complexity but I think I'm going to have to give it a go. Sorry for being lazy and not looking back in the thread right now but is your OP and circuit detailed already?

[izzy wizzy]

Nice. Very nice.

I've been fighting the urge to replace 12GN7 with 4P1L because of the added complexity but I think I'm going to have to give it a go. Sorry for being lazy and not looking back in the thread right now but is your OP and circuit detailed already?

Details of how it is now-ish are on this page http://www.audio-talk.co.uk/phpBB3/view ... &start=300

Except now the 4P1Ls are lit as above with the VCCS supply.

For all things 4P1L, look at Ale Moglia's blog. About 230V 30mA I think or thereabouts. Sockets are a PITA. The chinese ones are awful but affordable which I have and then there are others at way more expense which I'm now considering as I think they are keepers. I loosened up the socket pins but maybe too much on one of them. I feared for the valve trying to get them in and out as they were. You also have to consider microphonics as some are way more than others. I've yet to select for this but may have too. Mark has them in his 300B amps I think.

Cheers,
Stephen

[simon]

Thanks Stephen.

What sort of gain are you getting from them as triodes?

[IslandPink]

I suspect about x10.
As a pentode ( as I use ) it's around x22 without a cathode by-pass cap and x32 with one.

[izzy wizzy]

mu is about 8 according to Ale http://www.bartola.co.uk/valves/2012/06 ... de-curves/

As I use it transformer loaded, I probably get about x8 or thereabouts as a triode.

Edit. Just looked in the model. As a diff stage, I get 18dB so x7.9, A single triode CSS or inductively loaded will be about the same.

Cheers,
Stephen