Waiting for more resistors. But had a final run through of component values and pretty sure this will be right.
The prinicples copied from Jeremy Epstein many years ago are 1/ the same b+ without decoupling so out of phase noise from driver meets out of phase noise from output valve. In the “darling” article it was preferred on audition. 2/ The level of the output stage to enable direct coupling appeared around the same time in Jeremy’s circuit first followed by Thorsten and others as it became the “popular thing” at the time around the milenium, a fertile time for ideas. Dont be frightened to build it like this the direct coupling used was very popular in the milenium and nobody came a cropper. The only capacitors here are paper in oil, no nasty electrolytics in the signal section. Only a 1uF microwave cap in power supply before choke and 2 stacked Black Gates after making 100uF as they are 100 x 2 sections on top of each other. The noise cancelling nature of the two valves shall mitigate ripple.
The heater supply wants leveling around the cathode voltage even though it tolerates +300v -100v you have to add peak ac voltage to dc voltage potential so on peaks it would stress the heater/cathode interaction. But we dont have to make it exact.
The only elements to limit frequency response are bypass caps 3.7hz and output transformer. Unknown, but used them before and found worthy. Finally, input capacitance may require a low impdance source for hifi quality at hf.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
Adendum “ out of phase noise from driver meets out of phase noise from output valve” better said power supply noise effect on first stage is out of phase with effect on output stage. Though I can’t go deeper into this subject as the two valves stages have differnt gain. Similarly the concept of the dominant single ended distortion 2nd harmonic mitigation of the output stage by the first stage similarly.
This morning editing the prior post in this post, the idea flashed into my head, that these concepts only apply if like the late Sukuma San you employ the same valve for each counter distortions, be that power supply noise or distortiin. Then also you’d best use same loat mechanism et al.
So I’m not invested in any cancelation sucksess of either in this design.
I tried to employ valves of similar 2nd profile but ended up back with E88CC for reason of other factors.
Just possibly all cancellation concepts in this design are voided by dissimilarity. One day I may research the Sukuma examples in a lesser form.
Meanwhile Aikido concepts have not been tried, mainly because I dont want to build a monster with more valves.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
Posted by Paul Barker (A) on July 12, 2003 at 13:57:01
In Reply to: Re: Naked Monkey, Half Monkey, Full Monkey so what's the best posted by eduard on July 12, 2003 at 10:26:35:
An interesting thought about the chared power supply is that if the output valve ht falls due to suddenly increased current the ht available to the driver will rise, because the same increased current will draw more voltage across the larger of the two resistors. Therefore at the instant the power valve runs out of puff the drive signal to it is boosted.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
OK First turn on good job I used the variac because with the lowest tap on the transformer for my estomated 530v I need the variac at 88% of my mains. So thats a work in progress. Its a smaller variac than the blue big boy I brought to meetings. So variac is fine not bug and heavy and no noise like a power station.
Music is quiet on full volume but equal channels for that mstter two channels for a change! No noises pops or bangs no smells. So wiring which I tripple checked before variac.
But its not efficient enough. So I’ll try bypass the cathode resistors of the voltage amp first and if necesary add a stage. But hay ho. Nothing ventured nothing gained. No disasters or shorted output transformer this time !
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
Vision is an obstacle nowadays. I suspect its natural aging. But its made me realise the next build, on the bench I’ll solder flying leeds on to the valve bases first not deep in the recesses of a chassis. Numbered leeds would improve the situation also. Components on tag strips like Linn and others. More search lights aswell. LED lihts are hopeless for human vision. Note to self, buy a stock of halogen lamps, if anywhere in the world they are still stocked.
How you boys cope with surface mount solid state is a mystery.
Another thought is just use octal valves in relay bases. Ive often had this thought.
I’ll keep this one in its chassis. The final build in this one chassis.
So I suppose the next build will have to be about 20 watts. So onto the GM 70 monoblock.
Im glad I didnt build the one watter. Those are for guys with efficient speakers. All that chasing one watt stuff is much more expensive, when you cant just go out and top up youre pension.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
OK Ive spentthe morninghead sratching about the gain and the difficulties and riska of direct coupling, or just adding a stage cap coupled. This has led down the road on a wobbly path culminating with the decision to keep the same stages but cap couple. Forsake the resistor bias with feedback, I have some 47 uF caps polypropylene which will -3db the now 460 ohm cathode resistor of E88 cc at 10hz (considering my loudspeakers that’ll do.
with a new load of 84k parallel with a grid choke so Im calling that 84k gets me a Actual gain of 32 from a mu of 33. The b+ changes to 494v via variac.The 6550 grid now sits at -44 pk which brings the sensitivity up to 1.37 peak and from a cd player which is all I have to test it on, should put out 8 watts and I can now see how it stands up to driving the quad 57’s funnily enough I am a child of 1957, and also the Kef Corelli’s.
I know Ive been on a mad journey, and it was a journey too far juggling the direct couple involvement. While in my ear is ringing Nick saying to me, “what are you trying to fix?”.
So Ive got some 2 uF coupling caps 550v paper in oil in the signal path now, and another bypass cap.
As the amp stood there was no audible power supply noise from my speakers.
b+ now just under 500v I can put the two Black Gates in parallel not series (aslong as I screw clamp the variac position.) to get 400uF instead of 100uF and you will now be able to hear a nat fart.
I’ll post the revisions pictorally when Ive rebuilt it. I will no longer have to worry when something will cause a valve outage. And I wont have to stand the heaters at the grid potential.
But it was sweet sounding if a bit quiet (under stated) No point if its only for quiet! Im still glad I went up in idht’s until I found the one that stands a chance of matching my speakers. Because it sounded as good as any other triode connect dh beam.
Definatly more power will wait for the gm70 monoblocks.
Unless I wind a phase splitting choke out of my Nickle cores, parafed, to build a class A push pull output stage as an add on to this amp first. Say 16 watts.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
The black Gates are still configured on top of each other because this power supply is silent and it means I can use the supplier for higher voltage future projects.
With variac on max B+ is 428 v so we arent up to full potential intended b+ of 494v. But the power transformer is on the lowest of 3 tappings but as it sounds beautiful punchy and loud enough I’ll just enjoy it today. Havent even flipped signal section over to measure stages. For another day to change transformer tap and measure it up!
Just enjoying friends!
This is a keeper. I needed a victory’ about time!
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
The transformer is made with Tappings @ 625-0-625 (where I connected it) 700-0-700 and 725-0-725. I tested it with no load and my measurements show that is 5% regulation, which is a normal figure for a vintage double C core potted transformer.
I didnt flip the signal section over to measure current. But connected to 625-0-625 tap choke input slightly boosted by a 1uF pio first cap. We deduct 10%. Which means when I plugged into raw mains of 240 I also lost 134v through a parallel pair of potato masher rectifiers 5R4GY whose filament voltage is 10% low at 4.5 volts. And through the single choke direct resistance of 230 ohms. So estimate 582 mA current drawer from the amplifier, if we dont add in the dc through opt as of any significance.The power transformer secondary is rated at 500mA. The output valves I estimated as 200mA the ECC88 two halves of 5.5 mA each (I vaguely recall.).
So some head scratching required.In theory if I use the 700 -0-700 tap it ought to bring b+ to 496v plugged into full mains voltage, even with the 582 Ma current drawer if that is true. Not forgetting the voltage drop in the rectifiers hasnt been studied.
Id rather not have to use the variac so next move is to tap at 700 v flip it over turn it on straight from mains and test all resistances in the amplifier to see if anything there identifies excessive current due to stupidity in the designer, which is quite likely. Drop across opt first port of call, each side. Next drop across 80kohm load resistors for e88cc. Turn off imediately if anomoly found. Think and test again after recalculations. Expectation being alteration of bias required. I dont think this will kill any valves. But it will demonstrate their endurance.
Imhope you can learn from this process.
There could be dc current fault in the power supply through a bad winding in main transformer or the choke, or through a capacitor or two.
Which would mean a very long head scratch of which power supply to destroy next.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
By the way the pio caps are 160v rated and in normal working should cope with bias voltage 48 with ease. You may have noticed I have a clamp meter so could find current flow it should only be one of the three capacitors on that channel. Replace it with one of my 47uF polypropylene caps ith higher voltage rating and were on. Finding said current would also eliminate the purpose of switching valves.
So might all come good rapidly. But Ive just got home from the heating job and going to Crown Spar next. Cooke tea for us two and 90 year old uncle we feed every evening. Then in morning accompany uncle to hospital 7.45 preop for general anaesthetic.
So no I wont progress the amp until Friday, when also the heating man is collecting me to wrestle with his blocked heating system again. A pro bono job but he tips me £20. Probono because he and his wife only get £100 each a week I have told them about pension tax credits but being dyslexic I cant help with it. So Im donating my skills instead. I have told them to try Age Conscern et al. Who will surely sort it out. They have no savings and have no idea how they live at all. 84 or 5 and 86 87 in June. Poor buggers. Abused by the system no Idea what they ouht to get, just been taking it on the chin for years. The system is broken, its no wonder Im a so socialist. I turn up to houses in this state in my daily work. Its a broken system.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
