Easy EL34 amp

Well after four years...a thread resurrection. :eek:

Hi Kev, nice work on the amp! What bias do you run the 6v6?
Primo, it would have been whatever bias setting would cause about 60% plate dissipation. (from design center max). I don't have those notes any longer to know exactly what that value was. No particular reason for the 60% plate dissipation...not too hot, not too cold.

My current amp that runs 6V6's I think is biased upward of 75% to 80% of design center max plate dissipation. What set that bias point was (of course) the screen voltage chosen, which was chosen to properly load the output stage given the primary impedance of the transformers. But also, because that's where the lowest THD was measured at full power output. It's a bit of a different animal, being pentode connected and using Dave Gillespie's EFB output stage regulator. This amp uses Heathkit AA-100 iron, and is easily the best sounding amp I have scratch built to date. I'm confident that what makes that amp great is the output stage regulator and the output transformers, or more specifically, the HF extension offered in those output transformers.

Interesting you have feedback from the screen. I wonder since you feedback from one side only, it might increase even harmonics and give it a more present sound.
Don't know. I don't have an FFT analyzer. The feedback tap from the screen is actually not a new idea. Dynaco used it in their Mark III, and probably others. I kinda copied it from that. Works very well on some transformers in taming the winding resonances and not well at all on better wound iron with very little winding resonance.
 
I was reading about using low primary impedance to lower THD, I notice you use quite high primary impedance of 6.6K for a pair of EL34. Usually I see people use 4K to 4.5K primary. Any reason?

Primarily because my goals are different than yours. Generally the niche I look for in designing/building a tube circuit are that it is accessible to a lot of people. That means low cost, easy to use, stable, low maintenance, sounds good, and looks decent. For power amps it means no preamp required, stereo chassis, simple circuit with fewest tube count to achieve the power output I want. That led me to the 5-20 style topology as the bread-and-butter of the designs I make.

I've actually never built a Williamson type amp simply because there are too many tubes for the goals I had in mind. My goals carry with them some inherent compromises too, such as dealing with limited gain and that thorn in the side Miller capacitance when you hang a volume pot off of the grid of the first gain tube. I've tried various incarnations of frontends: mu-followers, cascodes, SRPP's, as well as pentodes. I've tried various incarnations of inverters, and the differential pair is the winner for me. I've tried CCS's as sources and sinks on the voltage gain stage and diff pair stages made from both tubes and transistors. I've tried traditional voltage regulators in the power supply. It all worked as advertised, but none of that helped the amp sound better in my opinion, so I stopped adding them.

With my current knowledge, this amp exemplifies the best I currently know how to do. This is my current and only power amp. It's a little 15 watt 6V6 push pull amp. It embodies all of the sweat and tears of years of learning, as well as one new problem I've never encountered before, but that is now solved and goes into my arsenal for future builds. This approach will scale up to about 35 watts per channel with a beefier power supply, bigger output tubes, and commensurate output transformers. I know this approach is not for everyone. But this recipe works for me.
 
Great iron those AA-100 output trans & they are the right impedance for EL84/6BQ5 as Dave Gillespie used them with those tubes. They might be about 7.5K , if my memory is correct

which is ideal for many tubes!

I tried them on my Eico HF-12's and they were much better on the highs and even noticeably better on the lows!
 
There is no uprated dissipation 6V6 like there is with the 6L6 family. The 6L6 family plate rating spans nearly 2:1 from the metal 6L6 to the last of the Sylvania tube production that used the 7581A plates. The regular -GC is 30W of course.
cheers,
Douglas
Hey Doug, did you forget the 7408. It makes a little more power than the regular 6v6.
 
There may be more variants of the 6L6 than any other tube, being the first sheet beam kinkless tetrode (1936). The 807 /6BG6G is an example , the 807 a popular pre-ww-2 transmitting tube and the 6BG6G an octal based 807 version produced for the first postwar magnetic scanned US television receivers.
 
There is a 5881, 5932, 7027 version of the 6L6 type also.

https://en.wikipedia.org/wiki/6L6

7581 is mentioned on a tube seller's site also as Tung Sol sells a new version of it.

5932 must be very rare as I've never read about it before! Always something new to learn1
 
Last edited:
Back
Top Bottom