The main goal in putting a system together, is to create an illusion that gives an impression of a musical event, they are here, I am there, etc.. I have no problem with a builder putting together a system by trial and error (which includes the source, amps, speakers, and room), then declares; this is the best illusion I've heard. I can accept, that for any given individual's ears and taste, the truth of that statement.
The problem is when it is claimed to be better than all others for everyone. To generalize for all cases, one has to construct a dogma, and invent new principles from whole cloth, as the generalization cannot be supported by current principles of cause and effect. Absent any testable hypothesis, absurd explanations must follow, including claims of exclusive purity, not only of the system as a whole, but for each individual element. True believers, as always, will remain unmoved by the scientific method.
Sheldon
The problem is when it is claimed to be better than all others for everyone. To generalize for all cases, one has to construct a dogma, and invent new principles from whole cloth, as the generalization cannot be supported by current principles of cause and effect. Absent any testable hypothesis, absurd explanations must follow, including claims of exclusive purity, not only of the system as a whole, but for each individual element. True believers, as always, will remain unmoved by the scientific method.
Sheldon
Very intelligently written, many good points, and even-handed. Gotta respect that.
I agree in a large part with you. But, (1) I have heard these amps extensively, and I am not so sure you have, even though you build a similar topology (2) This amp design got "Best sound of Show, RMAF 2005" by a 7 Moons audio reviewer most of us know and respect.
Jeff Medwin
hey Jeff - many thanks for all the help back in the day on my 5842Q/2a3 transformer coupled amp. I had a whole cook book with tips from you, Ed Warden, Bill Petrowsky and Christopher Paul - think it got ditched during a cleaning - wish I could gather the info again.
Freddi,
Is this Fred Irion ?? WOW !!! How long ago was it that we did this amp and communicated ?
Seems like 25 years !!
Are those DYNA ST-35 Output trannies on the back-right ? If so, quite a NICE sounding part choice. I also LIKE the way you positioned things, including the RCA jack entry point, close to the tube. Now a days, I'd put RCA jack on top of the chassis, close.
Is this amp still operational, and being used? Looks like a dual supply, separate for each audio stage, with a tube rectifier for the front end. Are those two chokes in the rear the 3 HY 24 Ohm DCR ESPEYs I used to suggest people to get from Fair Radio? Wow, they sound so darn lovely. I LIKE the L1/C1/L2/C2 B+ filter approach, sonically, over ALL else I have tried. Write to me, how many years ??? !!
Last edited:
Member
Joined 2009
Paid Member
This thread is a microcosm of the hobby, right here. anybody read the book 'Zen and the Art of Motocycle Maintenance' ?- a story in part about the conflict between the analytic and artistic. A conflict that took root in our society from the time that the ancient Greeks sought to understand the world around them and sought to reduce it to simpler elements. The purity of science and engineering that has demonstrated over again it's utility in helping humans generate new and better tools leaves little room for the approach based on intuition, creativity and emotion. This conflict is less evident in Japan for example where cultural roots are different.
I knew a plasma processing engineer once who had an uncanny ability to develop and debug the recipes (gas mixtures, pressures, flow rates, plasma energy levels etc.) for cleaning and etching semiconductors. He followed an intuition, his explanations of what he was doing sounded like psuedo-science as he struggled to find a vocabulary familiar to others who were very analytical, but the results were nonetheless very good. He 'knew' what he was doing but could not analyze it, break it down into good science and communicate it to other scientists. Yet those who used his recipes were thankful of his expertise.
As it happens, I have looked in more detail at the approach that Jeff advocates for a 2A3 amplifier (communicated some time ago) and I find the majority of what he has advised me to make very good sense from a solid engineering perspective and some of it to be unfamiliar but interesting.
I knew a plasma processing engineer once who had an uncanny ability to develop and debug the recipes (gas mixtures, pressures, flow rates, plasma energy levels etc.) for cleaning and etching semiconductors. He followed an intuition, his explanations of what he was doing sounded like psuedo-science as he struggled to find a vocabulary familiar to others who were very analytical, but the results were nonetheless very good. He 'knew' what he was doing but could not analyze it, break it down into good science and communicate it to other scientists. Yet those who used his recipes were thankful of his expertise.
As it happens, I have looked in more detail at the approach that Jeff advocates for a 2A3 amplifier (communicated some time ago) and I find the majority of what he has advised me to make very good sense from a solid engineering perspective and some of it to be unfamiliar but interesting.
Last edited:
No, I have not heard the amp, and neither have you (unless it has a real problem with a buzzing transformer). The only thing we should hear is the sound reaching our ears from the speakers and room.
Again, I won't argue that what you hear, best satisfies you that the music that reaches your ears, as played through that entire system, sounds most real to you. And it may to others too. I accept your preferences. What I do object to is the insistence that it's the only way. And to be clear; when I cited absurd explanations in my earlier post, it was in reference to the "wrangler's" meaningless explanations.
And by the way, the "wire with gain" crew, can be equally dogmatic that their's is the only way to perfection. Even if they get the physics right, they can miss the point.
In all cases, the human perception apparatus is part of the entire system. Fundamentally, since we don't have a practical way in our homes to exactly reproduce what might reach our ears in a live event, we have to fool the interpretive circuits in our brain. Whatever it takes to do that is perfectly fine by me. And we most likely have different priorities on what elements of the experience are most important or most pleasing to our individual tastes.
So, fine to say "I built this way and I really like the results". But don't insist that every other way is no good. And, if you are going to try to explain cause and effect, then stick to known physics, but be prepared to support your arguments with data. Otherwise it just "my taste is more correct than yours", on which argument is useless.
Sheldon
When you do a DIRECT COUPLE either into a 2A3 or a Type 45 Finals tube, either tube is VERY EASY to drive....but direct coupled only !!
You can not successfully parallel a 12AX7, or any mu of 100 tube, as you will easily hear the skewing of dissimilar tube sections. Most noticeable in diction, clarity of voice, it just never comes into pristine focus....skewed and screwed ( up ). The skewing is audible - it even applies to certain paralleling connections of multiple dual-diode directly heated tube rectifiers in such a amp !! Believe me, I have tried all of this.
There is a long audio history of whimpy high mu ( mu of 100 , under 1 mA. ) drivers DCed into low mu triode Finals:
There is the 1947 Jack Robins / Chester Lipman DC amp, 6SF5 into 6B4G. There is the much later Isamu Asano DC 2A3 amp ( 0.43 mA. ), and the later Nobukazu Shishido DC 2A3 amp ( 1 mA. total - but BAD paralleling of 12AX7 sections ), and modern amps like the Fi, Uthus, and Serious Stereo amps, all using such drivers. Amazing, isn't it ?
Jeff Medwin
Last edited:
hey DrLowMu - yeah "Fred Ireson" from WV - that amp is parallel SE with One Electron output iron (can't remember the Z - maybe 1K6-?), Southeastern interstage transformer, chokes from Fair Radio. Its sat for many years but should still run - think I botched the ground tying the stringed buss to the IEC sockwr ground rather than between the RCA jacks. It sounded pretty good in the day with older lps - Al Jolson, Ernest Tubb were enjoyable on my Edgarhorn System 100 - don't remember CD digital good in general in those days on that amp - lot of that must have been the mastering. I've been Karlsoning for a number of years so sidetracked 😀
From time to time the claim that paralleling nominally identical triodes damages sound crops up. If the triodes are actually identical then this can't do any harm, but real triodes are not identical. My question is this: within any real triode there are, in effect, several triodes in parallel due to 'island effect', and often audio circuits run them in the bias region where this is relevant; as one triode already has more than one ideal triode within it how can paralleling some more make much difference?drlowmu said:
What evidence is there for this? If the paralleling is AC-only does this improve things? If so, does this mean that the real problem is not something esoteric and 'non-EE' but simply poor current sharing so no advantage results from paralleling (but the designer assumed that it did)? Low mu valves would not suffer from poor current sharing in the same way as high mu valves.
The Fi Primer design has been my reference amp for about 10 years now other than a brief period I ran magnepans in the main system. Works good with anything from 90db and up in a 15x15 room but i'm probably getting closer to 4.5W out of it.
I'm using One Electron outputs, and deviated from the stock schematic only in that dissipation is cranked up to about 24W via about a 50V increase in B+ vs. the stock design. I use both Sino and Sovtek 2a3 tubes - I like them both and swap them every 6 months or so when I feel the urge for a change. Running some variant of the 6sl7 in the driver, sylvania 6188 IIRC.
I swap other stuff in and out for periods - st70, heathkit ea3's, cheap tube console pulls, and other homebrew stuff. This all only continues to confirm that this amp is all around pretty solid.
I'm using One Electron outputs, and deviated from the stock schematic only in that dissipation is cranked up to about 24W via about a 50V increase in B+ vs. the stock design. I use both Sino and Sovtek 2a3 tubes - I like them both and swap them every 6 months or so when I feel the urge for a change. Running some variant of the 6sl7 in the driver, sylvania 6188 IIRC.
I swap other stuff in and out for periods - st70, heathkit ea3's, cheap tube console pulls, and other homebrew stuff. This all only continues to confirm that this amp is all around pretty solid.
Sorry Jeff, what you write doesn't match my experience...I tried direct coupled ecc83 to 2a3 last year just for fun. Also 6SL7 since I like octal valves... I paralleled the ecc88 sections and it sounded somewhat better but it was not special at all. My speakers are 94 db too.
Direct coupled 6SN7 sounded superiour in every respect. I don't measure things like THD - just simulate and measure a lot then trust my ears...
My push-pull 2a3 using a pair of e88cc per channel (differential to concertina to CF) which uses 2 coupling capacitors still beat these single ended things. Maybe it was the single ended transformer - I hear you have to spend serious money on decent Iron when you go for single ended. push pull is far more forgiving.
Also - One problem with these direct coupled single ended amps (besides the VERY serious HT voltage) is the mighty BIG cathode resistor, and equally huge electrolytic bypass cap.
But you seem to know more than me. I mostly just trust my ears.
Ian
Last edited:
All 2A3, 6A3, 6B4G tubes (with the exception of the rare single plate 2A3 version), are all comprised of two distinct Triodes in parallel.
Why don't they smear the music?
Why don't they smear the music?
Member
Joined 2009
Paid Member
perhaps there is a difference between paralleling low-mu low-Rp tubes and paralleling hi-mu hi-Rp tubes
There are very few new 2A3's available now that are not single plate. It's hard to find a twin plate Shugang even. They were the budget choice in the 1990's but they don't seem to be available readily these days.
Sovtek, EH, KR, Sophia, JJ, EAT, new Shugang etc are all single plate 2a3's.
For one, you CAN parallel low mu tubes in the finals stage ....they are low mu ( 4 ) tubes. NOT like a mu of 100 12AX7 !!! Main difference.... mu and position ( front end high gain versus finals stage almost unity gain ).
Also, now a days, we use a single plate 2A3 called the JUJ 2A3-40 as the best all around performance - value ever !!
Jeff Medwin
Hello Ian,
First and foremost, I do not doubt for one millisecond what you heard. ' Have no problem with what you told me. The spec of 94 dB is the MINIMUM you need to do tube amps properly. There are HUGE advantages having 99 to 105 dB speakers, if ya wanna play tube audio design game optimally. Think Altec / GPA 604 MLTL or an Altec A7-800 system, tweaked,....perhaps the best bang for the buck.
But JUST FROM YOUR DESCRIPTION above, I would say you didn't build a really good SE DC amp.............. by any stretch of the imagination !!! You just threw stuff together and listened. Not the type of amp builds I am used to hearing !!
Here are a few key points "I" would frankly propose :
(1) A 12AX7 ECC83 is a sensitive tube, with a mu of 100, actually, its a great tube, IF you implement it well. I get a sense from what you wrote, you didn't.
(1a) The front end B+ needs to be LOCKED DOWN and regulated really well, to get MAXIMUM definition and dynamic slam . I use a SHUNT type regulator, located within an inch from my 12AX7's plate resistor. My shunt draws about 15-20 times of what my audio tube idles. This gives tremendously low Z, and a SHUNT regulator is the best sounding type of all of them. Without such a detail, the amp will not sound special, because its not built special, just as you report.
(2) The Cathode of the 12AX7 driver tube requires you to use TWO closely matched paralleled resistors ( Roederstein MK-3 are my favs ) matched to 0.1% of each other, to get maximum DYNAMIC response from the stage, "it just sounds better". I am talking of the Rk value now, of course. Order ten, match two.
(3) The cathode of the 12AX7 requires a MINIMUM of six, progressively smaller uF value, all-film bypass caps typically. The reason for this is because ALL film caps are pure crap, the worst part in audio, and only the better of them when used, will play back music, ( no matter HOW they statically measure ), in a very NARROW range . It takes "about" five bypasses, progressively smaller values, starting with a main film, and the first bypass being a 0.68 uF .....if you want to have it "all" played back to you . ( Open up your wallet !! ) Only exception, is if you use a Richard Marsh Multicap, about 15 uF , and even THAT will need something across it also, a second cap, possibly either a .033 uF or maybe a 0.013 uF DynamiCap, determined by ear.
(4) The 2A3 finals, just like the driver tube in (2) above, requires TWO Rks in parallel, and matched to as close as you can get it...0.1% would be ideal. Percy will match them from his stock, not as close as I suggest, for about a dollar a pair !! The best Rk I find for a Direct Coupled 2A3 is a Mills MRA-12, use two of them matched !!
(5 ) You can NEVER EVER use an electrolytic Rk bypass cap in a two stage DC amp, the Rk cap is directly in the signal path. That is horrible to hear !! Like in (3) above, the Rk needs to be bypassed with a multiple film caps, of the highest possible quality, starting with ONLY a rather small 8 or 10 Uf Film as the main cap, DynamiCap being OK as a general choice. The real fun, is in choosing the next five or so smaller uF value caps. ( Open up yer wallet !! ) Not part of this post.
What I have outlined above, I routinely do with every two stage DC amp that I build for myself, to get the result I would expect. On 99 dB speakers, I find the approach to satisfy me, and can not imagine using any other topology for the rest of my life !!
Less is more, but ONLY if you implement it with thought and great care. We haven't even discussed the wiring choices you used in your SE DC amp, or the parts layout, or the power supply..... ALL is critical and audible. Push Pull is forgiving, SE is brutal if you wanna build it to perform. It can !!
Jeff Medwin
Last edited:
one of two 2A3's i did about 2 years ago....D3a pentode drives the 2A3
Nice !!
You might put your RCA jacks on the topside of the chassis, such that there is 2 inches or less wire span, between the RCA jack and the grid of the D3A !! It is a better design practice, and will usually sound better. Cheers.
Jeff Medwin
- Status
- Not open for further replies.