So hopefully this will be a design thread and a build thread. My plan is to develop a small amp with 4x6v6 SE UL outputs and a EL84 pentode input stage. So far I have ideas on what I want to do but no actual schematic.
Starting from the output stage. Will be using the Edcor CXSE25-1.25K output transformer with the 40% UL taps. The 6v6's will be biased at 250v, 16v bias and 37mA Ip and 4ma Is. In UL mode each 6v6 outputs about 3W. 4x 6v6 in parallel at the recommended 5K load will work just fine with that Edcor OPT.
Each 6V6 will have its own ~400 ohm cathode resistor, cathode cap, and also grid resistor and coupling cap from the EL84.
The EL84 will be run in pentode mode. No cathode bypass cap or run with LED / Diode bias. There may be enough gain here to run some "schade" feedback around the 6v6's too. The EL84 will have to be able to drive 100K.
That's as far as I have got so far. Not sure on Power Supply yet and not sure on specific values for the EL84.
The power supply will be using all Edcor iron.
I am thinking out aloud and will probably want gain of 32 from the input stage. 500mV input for full output.
Any comments and input will be greatly appreciated.
Starting from the output stage. Will be using the Edcor CXSE25-1.25K output transformer with the 40% UL taps. The 6v6's will be biased at 250v, 16v bias and 37mA Ip and 4ma Is. In UL mode each 6v6 outputs about 3W. 4x 6v6 in parallel at the recommended 5K load will work just fine with that Edcor OPT.
Each 6V6 will have its own ~400 ohm cathode resistor, cathode cap, and also grid resistor and coupling cap from the EL84.
The EL84 will be run in pentode mode. No cathode bypass cap or run with LED / Diode bias. There may be enough gain here to run some "schade" feedback around the 6v6's too. The EL84 will have to be able to drive 100K.
That's as far as I have got so far. Not sure on Power Supply yet and not sure on specific values for the EL84.
The power supply will be using all Edcor iron.
I am thinking out aloud and will probably want gain of 32 from the input stage. 500mV input for full output.
Any comments and input will be greatly appreciated.
I have found that I can reduce output stage gain to about 3, using plate to grid feedback, using 6P6S (which is almost a 6v6)
So then with a 5k load, I need 30Vrms to drive the output to 100Vrms, and giving a good THD result of about 1.5%.
I am running them in triode, so the power comparison isnt valid-
FWIW I couldnt get them to better the THD in tetrode mode, at the same output, even if the power output is better.
This is not to say, that you couldnt do better !
Anecdotal stuff, maybe useful to you, maybe not.
I havent tried a pentode input stage, so I cant help there - my only thought is to wonder why the selection of EL84.
Wouldn't something smaller, work better? (Honest question, as I dont have an answer)
So then with a 5k load, I need 30Vrms to drive the output to 100Vrms, and giving a good THD result of about 1.5%.
I am running them in triode, so the power comparison isnt valid-
FWIW I couldnt get them to better the THD in tetrode mode, at the same output, even if the power output is better.
This is not to say, that you couldnt do better !
Anecdotal stuff, maybe useful to you, maybe not.
I havent tried a pentode input stage, so I cant help there - my only thought is to wonder why the selection of EL84.
Wouldn't something smaller, work better? (Honest question, as I dont have an answer)
do you remember what your component values were?
I am just on the fence on whether any feedback is really necessary for a 6V6 SE UL output stage?
I am just on the fence on whether any feedback is really necessary for a 6V6 SE UL output stage?
So input stage should be an EL84 running as a pentode.
Maybe:
250v b+
32mA
-9v bias
4K load
Vp 120v
250v screen
Should get ~30 gain
Maybe:
250v b+
32mA
-9v bias
4K load
Vp 120v
250v screen
Should get ~30 gain
Rin was 22k, plus grid stopper of 3k3, Rfb 220k/270k
(Again this is in triode, so gain is lower to start)
From brief experiments I used the same FB, zener regulating g2 at 200V, from memory I got closer to 3W output, with 2nd harmonic about -20dB (which in my opinion wasnt good enough, so I reverted to triode set up)
IMO, individual RC bias networks for each of the "finals" is insufficient. The matching between members of a quartet can be good, but it will be imperfect. Highly undesirable "current hogging" is quite possible. Combination bias, with individual trim pots., along with individual RC bias networks, will allow things to be set "just so".
Since the EL84 is being considered for "small signal" duty, I suggest the Russian 6Π15Π (6p15p) be considered too. Regulating g2 B+ with the 0A2 gas discharge tube disposes of its fragility and sockets correctly wired for the 6Π15Π accept EL84s without incident. "Rolling" opportunities are available.
Since the EL84 is being considered for "small signal" duty, I suggest the Russian 6Π15Π (6p15p) be considered too. Regulating g2 B+ with the 0A2 gas discharge tube disposes of its fragility and sockets correctly wired for the 6Π15Π accept EL84s without incident. "Rolling" opportunities are available.
If they normally drive into 5K what would be a problem with 100k? Does using an EL84 make any sense gain-wise compared to a high gain VA driver that would allow whatever FB you want to employ?
Would an EF86, 6SJ7, or a C3m be a better option?
I was trying to keep to current manufactured tubes. The EL84 is very widespread and is not a bad driver tube (see the Legacy amps).
In your opinion a matched quad of 6v6's plus a say a 25 ohm unbypassed cathode resistor (plus the usual cathode resistor / cap combo) on each tube won't be enough?
Would it be easier to ditch the 6v6's and say jump to a KT88?
I was trying to keep to current manufactured tubes. The EL84 is very widespread and is not a bad driver tube (see the Legacy amps).
In your opinion a matched quad of 6v6's plus a say a 25 ohm unbypassed cathode resistor (plus the usual cathode resistor / cap combo) on each tube won't be enough?
Would it be easier to ditch the 6v6's and say jump to a KT88?