Push Pull DHT triode

[James D.]

I have re-worked my 2A3pp design after the excellent feedback on the diode bias version. In particular thanks to Brett, I have worked on the operating points for the 6C45 and the 2A3 and these are shown on the schematic. (please ignore the - sign on the 2A3 cathode volts they are +48 not -48)

I've shown this version with cathode bias as an alternative to the battery and diode bias of the previous design. I will build it both ways round to compare.

The design is heavily influenced by Lynn Olsen and Gary Dahls writings - as should be obvious by how deeply into class A the 6c45 is running - both Va and Ia are above the book max values

I guess the 2A3 looks under run for a Sovtek - but I agree with the comments that this valve sounds better closer to 45 operating conditions and it suits the 5K anode load to run slightly lower current.

The iron will be Sowter and Lundhal - probably one amp of each for a shoot out then order a second pair of the winner...

My experience with pp amps is that they can benefit from running slightly unbalanced so the iron is specified with some dc current capability. I usual still try to static balance the stages but I have experimented with up to 20% unbalanced dc and found that it can improve the sound and measure better... Not always but definitly sometimes....

Any comments gratefully received

Ciao

James

[fdegrove]

Hi,

This looks perfect to me.
The 6C45 can take a lot so I won't lose sleep over that.

Here is were you surprise me...very few people know that:

My experience with pp amps is that they can benefit from running slightly unbalanced so the iron is specified with some dc current capability.

I definitely agree with you here and FWIW a few died in the wool tube guys i know have confirmed these findings.
It often makes a huge difference.

Good luck building you obviously know your stuff,

[James D.]

Frank,

Many thanks for the kind words. I had good teachers... I was lucky enough to have a Valve amplifer tutor from about 8. My Dad was an electrician and caught me with my head in an old radio poking around. Realising he had very little chance of stopping my inquisitiveness he sought out an experienced valve man to educate me in the dangers and the pleasures of valve electronics. I had a wonderful time and learned how to design and build valve radios and amplifiers all through my teens. I didn't touch transistors until I reached twenty - then I spent 20 years wasting my time designing transistor amplifers. I'm now 46 and I have finally given up on silicon for audio. Some of us take a long time of banging our head against the wall before we realise it's not going to fall....

My audio valve previous has been with pentodes and beam tetrodes for output valves, sometimes connected as triodes, so all this DHT stuff is new to me - and it's wonderful!

I've designed and procurred studio audio equipment, sat on advisory panels for studio audio manufacturers as well as hifi and I can't believe I have gone back to where I started out! Valve amps and full range speakers!

ciao

James

[analog_sa]

Very interesting. Anyone care to offer theories regarding why the unbalanced output sounds better. I didn't know it and will certainly try to put it to test.
If i had to offer an explanation it will most likely deal with the output iron enjoying the bias.


cheers

peter

[dhaen]

On the up side:
How about the re-introduction of 2H?
(There could be something unnatural in a source having only discordant distortion).
Also, the transformer gets some standing bias.

On the down side:
The output valves' (carefully matched) curves no longer fit each other.

I'm just guessing.

[James D.]

There are several factors working in favour of a mis-match pp pair sounding better than a matched pair (quiescent current match and mis-match) IMO.

1) we only static match the tubes and dynamically they will be mis-matched anyway - so there is a chance that the dc mis-match will result in a better dynamic match

2) If we know that we are mis-matching the pp pair then we allow for dc current in the OPT - the gap linearises the OPT considerably.

3) It increases even harmonic distortion...

4) It decreases odd harmonic distortion....

5) 3 & 4 together yield a better sounding distortion spectrum than a matched pp stage would produce... (this is much more important than 3 or 4 alone)

A good starting point for investigating this is

http://digilander.libero.it/paeng/on...P_Mismatch.htm

These were all known to the men who taught me. These men learnt their theory in the 40s and 50s and were 'expert' at Williamson style and Ultralinear designs. Interestingly they didn't rate the Williamson design that highly as it wasn't very good 'open loop' - they taught me that any amplifier was ultimately only as good as its open loop response. Global NFB was the icing on the cake - not the cake itself... I've found this to be true of all amplifier types.

I would love to hear some more reasons for this as well...

Ciao
James

[fdegrove]

Hi,

James, you summ it up nicely.

The term "mismatching" is a bit of a misnomer here in that, as you describe it correctly, you rematch the output tubes dynamically.

We found this out accidentally...no idea this was documented at all.

Cheers,

[ThorstenL]

Hi,

I have re-worked my 2A3pp design after the excellent feedback on the diode bias version.

Nice Job. I designed something a little similar for a friend. There the 6S45 differential stage drives 300B's, direct coupled.



Sounds rather good actually, especially with TJ "Mesh" Anode 300B's. Drawn BTW by Morgan "Valve Amplifers" Jones and Build by Jon "Electrostatic" Finlayson. As you can see, i had little at all to do with that Amp...

Sayonara

[dhaen]

Mmm,

Been thinking of some monkeying around myself.
Are the chokes "normal"?

Cheers,

[ThorstenL]

Hi,

Been thinking of some monkeying around myself.
Are the chokes "normal"?

The ANode Load chokes are in fact vintage output transormers from and EL84 PP Amp, secondary unconnected. I suspect a custom made, super low capacitance unit with permalloy core would be better still, but decent EL84 PP OPT's can be found VERY cheaply....

Sayonara