Use for 5K:8 PP OPT?

[6f6]

Perhaps a quad of 6v6 might make a good amp- assuming that the tranny can handle the current.

[soundbrigade]

Soory Dsavitsk, for borrowing your thread but:

What do I do with a 5K SE tranny?
VT1091

I have a pair of EL156, a bunch of GU50, another bunch of EL500 and some 6V6....

[Shoog]

Still, the 6AS7 sounds as a good idea to! Shoog has built a quite simple, completely differential, two stage, amplifier with them, with apparently quite good results.

In my 6080 PP amp I am using mains toroidals for the outputs, I took some 18V+18V:120+120 which I unwound the secondaries to 6V. I run them with the secondaries in parallel with 4Ohm speakers for a nominal impedance of 6,400Ohm. I have tried them with the secondaries in parallel, and the output falls off massively and the signal distorts badly, thats with the impedance at 1,600ohm - which should theoretically be spot on for the design. I really don't know whats going on with this, possibly my memory fails me and I didn't unwind them in the way I thought. But on the basis of that a simple PP 6080 with 5K outputs might work well.

Shoog

[dsavitsk]

Back to revisiting this. I drew up a possible schematic -- basically trying to use as many parts as I have on hand. I think the only thing I need to buy are a pair of power transformers. These will be monoblocks.

Anyone have any thoughts about the wisdom of the design? The choke is rated at 120mA, so the 6080 runs at 55mA per side and the 6N1p is at 5mA per side.

[dsavitsk]

Here's the PS sim.

[Shoog]

Looks promising.
You need a a means of current balancing.
What driver voltage are you expecting.
Are you thinking of those Lundahls for the front end?

Shoog

[dsavitsk]

Originally posted by Shoog
You need a a means of current balancing.

How important is this? Is a reasonably matched tube not going to be enough? The transformer could certainly have a cap inserted and be used with a pair of CCSes as a parafeed design. But, most push-pull schematics I've looked at that use this output stage don't seem to worry about this beyond matching tubes:
http://www.electra-print.com/docs/6bx7pp0001.pdf
http://nutshellhifi.com/Amp-pix.gif

Originally posted by Shoog
What driver voltage are you expecting.

I'm not exactly clear on how to do this calculation in PP -- I've only ever really worked with single ended designs. But, I figure that the 6N1p can take 2V p-p at each side (it is biased up 1V), and that will be just shy of a 60V swing on the grid of the 6080, which is about where the 6080 will clip. There should be about 110V on the plate of the 6N1p which means that the CCS will clip right before the tube depending on how efficient it is. I'm not sure how this translates into push-pull, however.

Originally posted by Shoog
Are you thinking of those Lundahls for the front end?

That's the plan. I had forgotten all about this project, and then had my memory jogged that I had these OPTs in a drawer.

[ErikdeBest]

Hi Dsavitsk

From own experience I can say that the two triodes of the 6080 are hard to get matched, and the single resistor in your schematic is about the worst auto-bias method there is
Broskie's explanation

Erik

[Shoog]

Sounds as if your about right on your driver calculations.

Unfortunately the two sides of a 6080 are notoriously badly matched - they never really needed to be in their intended role. I have used a suggestion of Brian Becks whereby I have LM317 CCS's in the tails of each triode. This takes care of current balance. I then have two large caps between the cathodes which effectively turns the pair differential. The cathodes of the caps are tied and referenced to earth via a 1meg resistor to keep them biased.
This has worked really well for me, and they have stayed in perfect balance until such time as they drift to badly and then they start to oscillate. Simple, differential and adjustment free - what more could you ask for.

Shoog

[dsavitsk]

Originally posted by ErikdeBest
Hi Dsavitsk

From own experience I can say that the two triodes of the 6080 are hard to get matched, and the single resistor in your schematic is about the worst auto-bias method there is
Broskie's explanation

The problem for me is that I am really cap phobic -- pathologically so. I will go to silly lengths to get caps out of the signal path, and Broskie, unfortunately, does not share this particular pathology. I have a pretty good stash of 6080's, but assuming I can't find one with reasonably matched sides, I'd almost prefer to do something like the attached. The bias resistor in this case works more or less like fixed bias with the signal current not going through it. My DAC and preamp both use a version of this circuit, so my experience is that it works reasonably well.

There is also an added benefit that I need not drop so much voltage in the power supply, or get it so clean, and can thus use a different driver (if it comes to it) that wants a bit more on the plate -- 12at7 or something. For the added cost of more CCSes, I'll save on PS caps.

Edit: though I guess the down side is limited swing on the output stage -- might be a problem.

Originally posted by Shoog
I have used a suggestion of Brian Becks whereby I have LM317 CCS's in the tails of each triode. This takes care of current balance. I then have two large caps between the cathodes which effectively turns the pair differential. The cathodes of the caps are tied and referenced to earth via a 1meg resistor to keep them biased.
This has worked really well for me, and they have stayed in perfect balance until such time as they drift to badly and then they start to oscillate. Simple, differential and adjustment free - what more could you ask for.

Do you have a link to a schematic? Sounds interesting.