This is all about output transformers. I wound 1k 8 ohm OPT’s 20 years ago and never used them.
Single half valve output stage into 1k is 15% 2nd h @ 4 watts
Parallel 1k 9% 2nd h 7 watts (lower than double power because Ive held current to 100mA per channel.
I chose 5687 for sound intersepted with 6EM7 power section to drive the parallel outputs better, and hopefully cancel some of the 2nd h of the output.
For once everything is an actual triode.
This amp will go downstairs for TV viewing/general listening to Apple Music on resident KEF Correli on stands.
I wound the transformers @100h to make a bass amp. Doesnt seem relevant any more but I need an amp downstairs and apart from the Tributes these are all there is left and theyre no use for any other valve.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
The SRPP is a dark art, and I dont profess to understand much of it at all. But I’ve gleaned a few elements. The bypassed kathode resistor is more distorted as found by folk using the 5687. Very few experts comment on the gain calc and those who do, use long complexed formulas. So Ive decided to guess until I can build it and test.
So to play safe I have taken the mu 17 and applied 0.66 to conservatively expect (A)ctual gain of 11 for that stage.
The A of the 6EM7 stage is simple fits my spread sheet triode gain calculator so : 5
So the two stages ought to drive the Peak requirement of the 6AS7 dual section use triode from a 1.5v peak source.
The 2nd stage and output stage low pass calculations of the cathode bypass caps are both 4.2 hz. The grid choke is in ac terms classed as infinity it is amorphous so there is zero low pass roll off from that junction. (5687 into 6EM7) the Hammond interstage designed for 10k should perform better at low pass with the 6EM7 thank with a more often used higher impedance tube such as 10y. The purpose of the 6EM7 power section in my thinking, is its reasonable mu permits me to not use the more popular higher gain tubes in the SRPP, allowing me to prefer sonics of the valve choice from my own experiential learning over 25 years. I respect though that once you configure them in SRPP that may not apply. So if not, at least Ive encompased the necessity of gain within that constraint of tube choice. Sorry but Ive never found the 6SL7 to sound adequate. Thin sounding. Im old fashioned about novel tubes, OK Ive been using E88CC to good effect so maybe my prejudice is wrong. Still dont like high gain tubes and only use them if the gain is required and there is no other way.
Low pass. I remember Morgan saying to me in Vry’s flat “low pass is additive”. I think going for two limiting stages @ 4.2hz keeps me safe. I know people calculate cathode bypass at 1 hz. But as long ass the additive pass is 10hz I’ll be OK. The OPT is excessive at bass by my decision in making it. This isnt usually as ideal at hf as the ideal decision in winding a transformer is a little compromise at hf and a little compromise at bass. With that approach the human listening to it will be satisfied quite easily. Becoming obsessed with one area will compromise another. So these transformers I obsessed about bass. It shall remain to discover the overall sound.
So there we are I dont think my big toe is lodged in my mouth. Its worth building.
Rectifiers are potatoe masher version of 5r4 two of, one whole tube per phase as that is required.
Choke input from 660vct. Choke textbook 10h as per 5r4 data sheet for choke input. All caps (edit, in the power supply) industrial polyprop I like the sound and cant convince myself that Black Gates sound superior. They sound different, thats all. Not unhappy with them, neither 7nhappy with industrial polyprop. Not noticed any better sound from what are propably the same caps wrapped in black with gold writing.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
There is a glance at the late Sukuma San in the 6EM7. I dont get all the gain I need in stage one so I couldn't use the 6AS7 to drive the 6AS7, and I think the 6EM7 is a worthy stand in, which will hopefully inject sufficient 2nd out of phase to mitigate the 9% contributed by the output stage. Its really such a high figure SE doesnt make since, the IT should phase split and same output valves push a de pull with the 2nd h at the final hurdle.
But I am using up what Ive got. Not buying what I havent got.
There is another fix though, I could SRPP the output stage (2 x b+ though) parafeed to same transformer. Maybe if its a total disaster…. Watch this space.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
A better solution I have seen work well with 6c33c, would be to apply feedback around the output stage by inserting the secondary of the transformer under the cathode of the output valve.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
I have a gut feeling the 6EM7 is the dominant sonic contribution so Im not expecting the excessivelty lush sound SE 6AS7’s spade out in buckets which is the marmite contribution. I see it more as a impedance matching tool fed by a amaising sounding triode 6EM7.
Obviously others would use a MOFO in the output stage.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
A better solution I have seen work well with 6c33c, would be to apply feedback around the output stage by inserting the secondary of the transformer under the cathode of the output valve.
Ooo, is that 100% positive feedback? Halfing the stage gain to 1? So I’d have to use more stage one gain. Back in goes the E88CC.
Or is double b+ (which isnt hard I have a 600v power supply on a variac already) SRPP output stage Double output valves also required for same power but no issue in this house locating more 6AS7’s when I once built the Rosenblit original Glass Audio OTL which by the way sounded rudely phenomenal soundstage filled the room corner to corner in every dimension.
Well back on planet earth, if its just return ground of opt secondary to cathode of output valves losing half output stage gain I could try that first and add gain earlier if its a winner. Thats most simple crazy thought.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
Ooo, is that 100% positive feedback? Halfing the stage gain to 1? So I’d have to use more stage one gain. Back in goes the E88CC.
No, first, positive feedback will increase gain, you phase it so its negative feedback. The feedback ratio is whatever the voltage ratio of the output transformer.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
Right, so if Ive understood the feedback, my course of action is build without, audition to see if the good old 6EM7 mitigates the excess lush of the standard 6AS7 se amp. Then clip a crock mouth leed from opt to cathode accept half voltage out resulting in a cubic loss of audio, but hopefully still enough to show me a worthy sound. Then adjust resisors of stage one and swap out the 5687 for a 12ay7 or a ecc88 base and heater adjustments as appropriate. Bearing in mind I’ll have to check the 6em7 operating conditions to pass on the extra gain. Should be ok though. If not, out comes another valve lick a dht i.e. 10y but maybe just a triode connected EL38 which by the way is how I imagine lestening to music in heaven would sound. I love it so much I dont actually see the point trying the EL36.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
It won't be as much as half voltage out because its not 100% negative feedback. Also, the output of the transformer is in series with the cathode circuit not across it (which I suspect you mean with the clip lead mention).
Oh, not what I saw in my mind. That makes a lot more sense series connecting that way, and the fact of the transformer ratio mitigating the volume loss.
Seems a very good way for the situation I am in, which may not require any drastic changes. I like it on paper.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
Not forgeting a permanent 100% fix of the 2nd is push pull which can take the form of totem pole
I would doubt it, unlike a push pull output stage where 2nd can be canceled, any form of a totem like stage will not do the same as the two valve see different views of the load, The lower valve is driving it from its anode and the upper one from its cathode.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
OK then this feedback option is required due to the 9% whhatever it takes off that excessive distortion is wanted and a little assistance from the 6EM7 stage could contribute enough to notice, and as you say not a lot of gain lost. So definately the amp is expected to sound improved compared to the first drawing of the circuit.
Thank you Nick for you're contributions.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
