PP 6l6 class ab...advice needed

[marantz1]

Hello Folks,

I have a pair of 7k opts and a bunch of 6l6gc/gb and quad of kt66... I would like to build a pp class ab amp with about 25 to 30 watt of power.

What will be the appropriate b+ voltage or plate voltage? I.looked at the datasheet and the 7k opts might work but I don't know what the appropriate plate voltage be.

Thank you..

 

[artosalo]

With fixed bias 6L6GC's some 420...450 V will be fine.
With UL-connected output stage some 30...35 W can be expected.
With pentode connected the power is higher. I have only used UL and it works great.

 

[Zibi]

https://translate.google.pl/transla...ar.com.pl/audio/diy/concertino/concertino.htm
https://translate.google.pl/transla...r.com.pl/audio/diy/concertino/concertino1.htm

 

[Miles Prower]

Hello Folks,

I have a pair of 7k opts and a bunch of 6l6gc/gb and quad of kt66... I would like to build a pp class ab amp with about 25 to 30 watt of power.

What will be the appropriate b+ voltage or plate voltage? I.looked at the datasheet and the 7k opts might work but I don't know what the appropriate plate voltage be.

Thank you..

From the STC App Report:

V_PP= 360V
V₂₂= 270V
V_GK= -22.5V (fixed)
I_Pq= 88mA (both)
I_2q= 5.0mA (both)
I_p(max]= 138mA
I_2(max)= 16mA
R_L= 6600R (P-2-P)
THD= 1.8%
P_o= 26.5W

7K isn't too far off the specced 6K6 load, so that'll work. You can actually get a bit over 30W if you include grid drivers that can take 'em just a bit into Class AB₂. (I measured 32W @ 1.0KHz with 30W Hammond OPTs)

You can use the same working voltages and get 47W if you cut the load resistance to 3K8 (P-2-P) and increase the grid drive to 72V_p-p if you use a source follower as the grid driver. Nor so sure if that's worth it as how much louder is 47W as compared to 30W anyway?

The thing to keep in mind is that the 6L6-oids like to make a lot of high order harmonics, so that they don't sound as good as that THD spec would suggest. The type's originator, O. Schade, recommended feeding back 10% of the plate AC to the grids with parallel NFB. This will definitely help tame that pentode nastiness, and just a bit more gNFB will take the "edge" off.

If you had an OPT with multiple secondary taps, and the right taps, you could also use cathode NFB instead of parallel NFB.

The type also likes to make snivets and RF oscillations, so both screen stoppers (1K5/0.5W C-comp) and plate stoppers ( N= 10; #18; ID= 7/16" space wound ) are suggested here. Parallel the plate stopper with a 100R/2W C-comp (or four 470/0.5W C-comps in parallel) and install inside the coil. Connect as close to the plate pin as possible.

 

[Eli Duttman]

Two very well proven topologies are Mullard and Williamson. The H/K Cit. 5 is a superior execution of Mullard topology. The Heath W5M is a reasonable example of Williamson topology, but I'd definitely use 6CG7s, instead of non-linear 12AU7s. Both designs use 6L6 family tubes.

Mullard style circuitry is less prone to trouble, when implemented with so/so O/P "iron". Williamson style is vulnerable to low frequency instability. Study the phase shifts and you'll see why. If you have doubts about the "iron" you possess, go Mullard.

 

[marantz1]

hello Folks,

Thank you all for the response I will report back on what i will build. I am pretty sure I will have more questions coming....

Best regards

 

[artosalo]

Below is the Dynaco-style circuit I have used quite a lot. It is simple and with a good OPT the performance is also good.
I have tested this triode-pentode front end with many different output tubes and output transformers.

This drawing is with EL34's but with just adjusting the bias and GNFB this will work perfectly with 6L6 etc.

 

[Miles Prower]

The Vixen

Since 807s and 6L6s are electronically identical, 6L6s are a drop-in for this design.