Audio Asylum Thread Printer
Get a view of an entire thread on one page
|For Sale Ads|
In Reply to: RE: I have gotten into computer audio just a bit posted by tube wrangler on February 06, 2017 at 11:12:16
Without making any judgements whatsoever about sound quality, it looks to me like your 2A3 monos and Don Garber's (FI) 2A3 monos are based on (but not identical to) the same 2A3 schematic.
This amp, and ALL of its progeny, drives from the original 1929 Loftin - White two stage direct coupled amplifier. You can search on line and find other examples, the original Isamu Asano amp, and Nobu Shishido's amp, from Japan. In Joe Roberts' Sound Practices magazine, Spring 1994, we have the superbly-detailed article my Ciro Marzio, and Christiano Jelasi, on this amp type. We had Craig Uthus doing the Moth Amps, and Don Garber doing this topology with some of the Fi Amps.
The very best by-far implementation is the Dennis Fraker Serious Stereo DC amp. No tube amp, SET, in audio can hold a candle to Dennis' on any direct A-B comparison.
Please NOTE : It took me the past nine years of attending RMAF shows, to fully evaluate and appreciate what it is. Each year, the setup is different, and results in-the-room are different. But after nine years of attending, I think I have a better handle on the build, than anyone other than Dennis himself.
Herb Reichert heard it, and got it, right away at this past 2016's RMAF. I was in the room, just the three of us guys, when Herb listened very attentively. Dennis' very latest amp build, and its' performance is fully the equivalent of the newest Ford GT, or a new Ferrari, or a new Maserati.
" The truth will set you free. "
Well, at least this one is D.C. coupled, and Thank God it has no Grid- (music!) STOPPERS! Tubes don't oscillate just by themselves-- crappy wiring makes them oscillate). Grid stoppers STOP MUSIC.
This schematic is very basic, it is missing 4 additional chokes, local feedback R7 is a joke, C1 should be about 9 Super-High quality capacitors, ditto for C2, C3 and C4. R3 and R4 must be ONE resistor-- not two in series. This resistor had better draw about 15X or more-- than the 6SF5 plate current.
R5 is critical-- it should be a Caddock TFO-20 or better quality.
C4 should not exist-- use Choke-Input here-- it SHOULD NOT meet
"critical" requirements-- the lower the DCR, the better, the physically smaller the choke, and the larger the wire in it, the faster and more dynamic the amp is.. It is used as a retard-coil instead of critical inductance-- to scrub-off unwanted R.F., noise etc. The rectifier MUST see this retard-coil BEFORE it sees a capacitor. There must be a careful balancing act performed between lowering ripple, and listening to the amp.
One CANNOT build a S.E. amp like 1/2 of a Push/Pull. It will be a total Wuss-- nearly all are!
The 5Y3 rectifier is a total dud. A 5U4GB will run this thing with FAR better dynamics.....
A LOW DCR power transformer that is about 3 to 5X times the power needed is also basic to getting this thing to start making music. Chokes must be
high-current rated or they will rattle or try to saturate.
The 6B4 tube is NOS-- and like all of them except the venerated RCA Single-Plate of old, cannot even begin to compete with modern tubes.
Dynamics WAY too low, distortions WAY too high....
Various R.F. chokes are also needed throughout the power supply systems, and are needed on the driver-tube's filaments.
Since this type of amp is sensitive to induced distortions of all kinds, it is necessary to use only the best, most expensive parts. The thing will Wimp-Out on bandwidth if you ignore it-- for instance, it
won't have shimmering, deep and extended highs, but a good Push/Pull will....
Layout is super critical, also, NO component can be directly bolted
to the chassis, and that includes even decorative wood or aluminum panels!
This means you have to invent ways to mount components where they
work at their best. You can't allow any wires, tyraps, chassis parts,
or other components to be parallel to-- or anywhere near anything that is
carrying power or a signal. Remember- since it is Single-Ended, it can't
cancel induced distortion or hum. If you arrange your circuit so that it
DOES have some distortion-cancellations-- then it will also cancel parts
of your music that are similar. People who can't hear extended highs won't even know it's happening, but your trusty analyzer will, and so will you-- over time.
Over time is the clue. You'll never get tired of listening to it if it's really good. My own situation is unique-- I have really good music
inputs, and I also use these kinds of amps for Home Theatre. I demand and
expect Big-Time dynamics and explosive lows-- something people normally use subwoofers and subwoofer amps for-- or at least they Bi-Amp.
I won't tolerate this because I learned long ago that all the amps in
a Bi-Amp or Tri-Amp situation MUST be the SAME amps, and must employ the SAME bandwidth capabilities. Not only that, but all wiring to the drivers
must also be the exact same stuff, and the exact same lengths.
Interconnects and signal-shapers? Ditto! So why bother?! Just make one
amp that can truly do ALL of that itself, and get ALL of it right. Sounds WAY better that way! Not only that, but you won't tire of it-- ever.
And that's really the biggest challenge-- can you answer all of them?
Will you please explain technically how a grid stop resistor stops music?
And please just stick to the technical facts, I'm not interested in your subjective opinions.
Have Fun and Enjoy the Music
"Still Working the Problem"
Anyone who says tubes only oscillate as a result of bad wiring doesn't know this subject matter. You might as well ask a child in a sandbox to explain molecular cohesion.
Buy Chinese. Bury freedom.
2 Things that I totally disagree with, One is the DC filament debate
Both DC and AC can be done wrong IE(HF AC Heating) but when DC is done right IMO its a tough act to follow I promise you
The other is "RF" stopper/damper resistor, anything can be done wrong using inferior A musical resistors, but when done correctly there is nothing added or taken away.
But I digress use only if you have very hi GM tubes 6dj8 comes to mind
There are a lot of amplifiers using these hi gm tubes and they have harsh false detail of sound, prob because of the HF oscillations.
There is so much RF flying around today that IMO almost all tube amps should utilize them.
I engineer my amps with half theory and half "human engineering" my ears tell me much more then my test equipment ever will that's for sure :)
Think positive thoughts and be kind to everybody
You mentioned you use SET amplifiers for home theatre. Since first hearing about this idea (from you), I reconfigured my home theatre from a totally decent Denon/Triad 5.1 system, to a 2.0 stereo system using my far from perfect 2A3 amplifiers, and high efficiency speakers with no subwoofer.
The result is OUTSTANDING, this was excellent advice, I advise anyone with SET amplifiers and high efficiency speakers to try using them to watch movies. You won't believe your ears how much better movies sound this way.
I had this mental block that movies require hundreds of watts of solid state power, and subwoofers to be enjoyable. It is simply not so, the 2A3s make maybe 3W per channel and it has all the dynamics I could ask for and then some. I think this is one of those things which needs to be experienced instead of talked about - I for one never believed it could be the case until I tried it and now I will never go backwards.
And this isn't just a TV, I have a front projection setup, with a 120" Stewart screen in a purpose built windowless room.
A good 2A3 tube is an oddball. While a good 45 can give you lovely voices, some real bottom end, and can out-detail your pet kitten, a good 300B-- even the best (EML) just doesn't tell the tale like a 45.
The 300B that people "love" is simply drawing too much plate current. That means the output transformer has to get physically larger. TOO MUCH larger for ideal S.E. operation-- it kind of falls-off of the normality (for reproducing a real sense of presence-- at a live event) chart. (fictitious to be sure, but plenty real if you LISTEN to it...). Whoops! There goes a lot of your bandwidth, jump-factor and easy-to-drive capability. So much so that it begins to sound euphonic, not truly real.
BUT! The lowly 2A3 can really do it all and have more power than the 45.
2A3 tubes, of course, come in many kinds. There are the NOS BiPlates-- which always lose tons of coherent musical information, but can sound "fat"-- even pseudo-impressive.. But alas! It Ain't Real, Dude!
Then, you can try to round-up a pair of old RCA Single-Plates. These are
fairly rare, and good ones usually fetch about $1000.00 for a pair. Good?
Not really-- they're just too old-- the materials suck, compared to today's thicker glass, higher vacuums, and far better metallurgy. They ARE, however, musically coherent. You get that because of a geometrically-correct filament structure and a Single Plate.
Enter the modern EML mesh-plate and the JJ 2A3-40. The EML mesh is a
geometrically correct filament structure with a plate system that combines
pieces of both solid-plate and mesh-plate constructions into one plate structure. Best of both worlds? I have them, but prefer the solid-plate version, or the old AVVT Pure Woven Mesh. Just sounds more real to me....
The JJ 2A3-40 is something that people think is a 300B. It is NOT-- but it does use the plate assembly and glass bulb from their 300B. And why not? Works just fine. The rest of it is pure 2A3-- the filament and the grid structure. BOTH make the tube FAR superior to any 300B-- a fact that JJ admits to on their own website.
The 2A3 accidentally (?) happens to fall into "ideal" plate-current territory for a near-perfect S.E. output transformer. Not too big, not too small. And just the right amount of inductance. Cool! That area is 40-to-50 ma. The 2A3 tube LIKES that....., and it LIKES the usual 2.5K loading, although with the lower plate currents, one could go up to 4K or so-- that would amount to listener's preference and little else... measured distortions drop with easier loading, but so does low-volume dynamics. So which REALLY has the lowest LISTENING distortions? Nobody
measures THAT, I'm afraid...
Other advantages include easier-to-drive grid, and a 2.5 volt filament
which encourages balanced A.C. operation of the DHT filament-- this gives you a complete musical picture (with an unavoidable small amount of hum), rather than a partially rectified-out smothered, sanitized "cleaned-up" false picture of musical input... that operating Direct-Heated filaments on D.C. produces. This does not imply that Indirectly-heated tubes shouldn't use D.C. heating-- that works perfectly, but the 2A3 is a DHT.
There are 4 volt (filament) tubes that are DHT, and people also like those. There are even some 2 volt models out there in NewOldStock.
Why bother? The newer JJ 2A3-40 is such a screaming sonic and reliability bargain that it shouldn't be ignored....
but, there are a few things that I can absolutely guarantee are not being measured:
1. Transfer Efficiency
2. Jump Factor
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Post a Followup:
Post a Message!
This post is made possible by the generous support of people like you and our sponsors: