|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
101.183.12.24
In Reply to: RE: 2A3 Options posted by maxhifi on December 21, 2016 at 10:23:30
To my mind there is only one option.
If you are running 2A3s near full rated dissipation and you would like a similar but more "refined" sound than the JJ, I would suggest the EML 2A2-S.
The EML will last a good long time at near 15W dissipation. It sounds more even and resolved than the JJ; it's dynamics are not quite as great - it is not a overtly propulsive, but seems honest to me. The build quality is better than the JJ.
Cheers,
91.
"Confusion of goals and perfection of means seems to characterise our age." Albert Einstein
Follow Ups:
Do us all a favor, and ask Jac what his favorite dissipation level is on his EML 2A3s, where it lasts long and sounds good to him.
I will be surprised if its 15 Watts plate dissipation. At such high dissipations, near the maximum dissipation rating, the 2A3 always sounds stressed there, because they are stressed, thermally stressed.
Post up here his response.
If anyone has to run the 2A3 that hard, for it to sound good, a DIYer likely has multiple problems elsewhere in his DIY gem / amp, typically in this order, power supply, wiring, layout, parts choices.
Be cool to hear Jac's response. Jeff Medwin
It's loafing, should last a good long time, and sound pretty darn nice at 15W. If it doesn't, the recommended use and 28W rating would be misleading.
I'm well aware of how Jac likes his tubes used - I've communicated with him several times and he outlines it on the EML site. I wouldn't be suggesting it if I did not have some understanding. For example, I would not run the 17W 2A3-M at 15W... more like 10W.Cheers,
91.
"Confusion of goals and perfection of means seems to characterise our age." Albert Einstein
Edits: 12/23/16 12/23/16
So it looks like there are a couple choices of 2A3 for 'high power' (two watts or so) operation. This is good news as I intend to re-build my DIY-HIFI Supply 'Billies" which already accommodate the 2.5V mesh plate 300B's from China (microphonic). WE-300B's from Kansas are getting to be close to 15 years old and have no getter left but still play well.
Link below:
According to Dennis Fraker, any 2A3 from China is inferior, "trash", due to poor metallurgy. It MAY possibly sound OK when new, but it will soon turn to trash over time, due to poor metallurgy.
The best two 2A3s to use are the JJ 2A3-40, and the EMLs, period.
The JJ is, by far, the best value of the two, and what Dennis designs for and supplies in his 2A3 amps in 2016. He ran JJs at RMAF this year, see Herbie Reichert's reviews for Stereophile, of his listening experience.
Jeff
I didn't say that it USED them.
The early TJ Globe 300B's were micro-phonic, later ones not so much. I have an early pair but they have not been in the amps for about 12 years.
"The JJ ...and what Dennis designs for and supplies in his 2A3 amps in 2016."To clarify, Dennis prefers EML 2A3-M for best sonics. Read down to the last couple of paras of the attached link.
Cheers,
91.
"Confusion of goals and perfection of means seems to characterise our age." Albert Einstein
Edits: 12/23/16
Yes, I was fully aware of that post. Here is a link to "the" most recent regarding this discussion.
Merry Christmas !!
Jeff
A person could make themselves dizzy tying to follow... Thankfully, I have (only) a passing curiosity.
Have a great Christmas Jeff!
Cheers,
91.
"Confusion of goals and perfection of means seems to characterise our age." Albert Einstein
Dennis said " First of all, most NOS 45's, 2A3, 300B, etc., have a rectangular
plate structure, and usually a rectangular grid structure.
So far, so good. Now, what happens when we insert a non-symmetrical
part into this desirable arrangement-- we place into it a filament structure that is made like a cheap toaster-- a folded-up filament
mess that is "M" or "W" shaped. "
My question is,
What is the shape of the electron cloud?
Just because the filament is "a non-symmetrical part" doesn't mean the cloud is "non-symmetrical" and keep in mind the electrons do not go from the filament to the plate. They go from the electron cloud to the plate.
If the electron cloud fills the space between the filament and the grid in an even way then the fact that the filament is "a non-symmetrical
part" won't matter.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
The point that is fundamentally misunderstood by many is that the filament shape determines the effective anode area.
dave
It is my understanding the coated filament of the 45, 2a3 and 300b produces a much more generous electron cloud than that of a thoriated tungsten filament.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
The article I pulled those from was from 1929 which is the era of the 50.
dave
Page 11 shows the "M" filament of a 211 utilizing almost the entire plate.It's to bad that the book doesn't show the "effective anode area" for a coated filament tube.
Thanks for the link.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 12/25/16
I think we may be interjecting a bit too much into the accuracy of the sketches. It does appear clear that the anode area is directly related to the filament structure. In the drawings I referenced there are curves around the corner which seems slightly more plausible than the sharp angles shown for the 211. I do strongly suspect that the edges are much more softly defined and and the "boxiness" of the drawings represents the average behavior over an area.
The point I was trying to put forth was that the filament appears to define the "active" plate area and while 100% "anode activity" may be achievable, I am not convinced it is a figure of merit that we should solely base out tube choice decisions on.
dave
Can't help but suggest a serious consideration as to why some of us prefer the sound of 'round plated' tubes?
well.... since you are asking for wild ass hypothesis...
a typical "box plate" is planar about the cathode and has a grid and plate structure on either side effectively creating two distinct tubes in parallel. The Cylindrical model doesn't have the two discrete structures but a single one.
Now one can argue that in the planes in the "box plate" versions may (can) not be perfectly parallel leading to differing characteristics along the planes surface and the same goes true for the cylindrical version but the big difference I see is that the box plate model still has two distinct structures when viewed from the top and the cylindrical model will only have gradual change over the surface.
dave
;-)
A lot of these arguments fall apart for this very reason. The biplate 2A3 is bad because it is essentially two discrete tubes in parallel like the 152TL. By that same logic, the 45 is bad because it is also two tubes in parallel. The thing that finally sold me on this was comparing the EC8010 to the EC 8020. Both have the same Mu but the EC8020 has double the gm which translates to 1/2 the Rp. If you look at the tubes the EC8010 is simply an EC8020 with only one plate rather than a plate on either side of the cathode / grid structure so it can be argured that the EC8020 is simply a pair of EC8010's in parallel :-)
dave
Nice Dave, Nice.
Jeff, with an EIMAC 35T on my desk as I type this.
"I am not convinced it is a figure of merit that we should solely base out tube choice decisions on."
I agree and was just trying to point out that the electron cloud is much larger than the filament itself.
Some seem to think that the filament itself needs to cover the whole plate area or the tube won't sound good (or something like that).
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Thanks for the Christmas wishes!! It means something to me now-a-days.
I got saved, at a local Southern Baptist church in September 2015, so this, at my age, is only my second Christmas as a true believer in Christ.
Amen Bro'.
Jeff Medwin
I see that the state of the world hasn't changed since I got the JJs in the first place... it's good to know I made the right choice to begin with. The EML looks like a very nice tube but is too much of a stretch cost wise for my budget and the quality of my amplifier and supporting equipment.
The JJ-2A3-40 is already by far the most expensive tubes I have ever bought, I think I will follow Jeff's advice and improve my amplifier a bit in other ways. The power supply could definitely use some upgrades.
My recommendation hasn't change from last time either. Electro Harmonix Gold Grid. They sound good, and won't break the budget. I also still recommend the ALK crossovers for your LaScalas. Then your 2a3 amp will work perfectly with your speakers.
Merry Christmas!
------------------------------------------------------------------
Big speakers and little amps blew my mind!
Merry Christmas to you too!
Good to see you're still around here.
Nice to see an an endorsement for the EH tubes too.
... go for it!The JJs are nice tubes - if you don't want to spend more there is no need to consider anything else.
Consider what you read. Experiment, and trust your ears. No one has the formula for an amp that sounds ideal to all music lovers. Don't allow yourself to be harangued or herded. Keep in mind what you want to achieve - both sonically and with regard to your learning - then develop your approach.
Keep us informed of your adventures.
Cheers,
91.
"Confusion of goals and perfection of means seems to characterise our age." Albert Einstein
Edits: 12/23/16
Would be fairly easy to change from cathode to fixed bias. Anyone know what the max Rg is for the JJ? I have 150k at present, it seems marginal for such a big tube in fixed bias.
Run the JJ-2A3-40 at 11.2 watts dissipation. Ideally, this is with 255-260VDC across the tube.
Direct couple will drastically outperform cap-couple, and transformer-couple is second best. Use Lundahl.
LCLC is the only way to go with your PS. The "L's" don't have to meet critical inductance, but they DO have to be connected directly to the rectifier D.C. output.
A.C. Ripple should be looked at, but sonically, is a less-important factor in S.E. operation. The object is to get ripple as low as is practical without adversely affecting musical expression and flow.
Power for the driver tube must be as clean as possible, but once again, you don't want to compromise musical performance in order to "overkill" it-- be responsible to musical flow-- ignore numbers except where you truly need them to accomplish a certain thing.. Then, understand WHY you are doing or allowing a certain set of numbers.
The JJ-2A3-40 is capable of near-perfect performance if it is used right.
The EMLs are the same plant's more expensive variety. If an amp is set-up for either kind, it WILL NOT sound optimum on the other tube style.
I find it easy to get the SAME performance from either type-- the methods used are DIFFERENT from each other.
Decide what you're going to run, and then make the amp perfect with THAT tube ONLY. You can still run others-- if you set the amp up right, the others won't sound as good as the set-up tube, no matter what else you use.
The bottom line is that your amp can only sound as good as the man who set it up...
--Dennis---
LCLC implies a critical inductance first choke.
Isn't lClC (little l) a better representation of what you are doing?
Kind of like when one adds a small cap to a critical inductance choke filter to increase the voltage a little bit?
Those are referred to as cLCLC with the small c denoting that the first cap will not make the filter act totally like a cap input filter.
Just like your first choke is too small to make the filter act like a real choke input filter. (each diode staying on for it's full half cycle)
Hey, it's just a suggestion to keep it real and not confuse anybody.
I wouldn't call a cLCLC filter a cap input filter and you shouldn't call what you are doing a choke input filter.
Yours acts much more like a cap input filter than it does a choke input filter.
There's no reason not to be honest about it.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
In another post you said,"First of all, most NOS 45's, 2A3, 300B, etc., have a rectangular
plate structure, and usually a rectangular grid structure.So far, so good. Now, what happens when we insert a non-symmetrical
part into this desirable arrangement-- we place into it a filament structure that is made like a cheap toaster-- a folded-up filament
mess that is "M" or "W" shaped."My question is,
What is the shape of the electron cloud?
Just because the filament is "a non-symmetrical part" doesn't mean the cloud is "non-symmetrical" and keep in mind the electrons do not go from the filament to the plate. They go from the electron cloud to the plate.
If the electron cloud fills the space between the filament and the grid in an even way then the fact that the filament is "a non-symmetrical
part" won't/shouldn't matter.edit, by "fills the space" I mean the entire dimensional area.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 12/25/16
Hello Dennis,
Please sir, what do you mean with this phrase ..
" they DO have to be connected directly to the rectifier D.C. output. "
Thanks.
Jeff
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: