|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
209.137.226.171
In Reply to: RE: Filament bias and DHTs transformed my system. posted by andy evans on August 28, 2012 at 23:57:34
Wow! You hit one out of the park this time, Andy!
I will take the time to comment-- maybe even more can be applied here:
(1) DHTs for everything. NO argument whatsoever. Only problem I have is how do you find DHTs with an amp-factor of 100, so you can Direct-Couple a two-stage?. I still think that your approach WILL result in very good fidelity. I'm really splitting hairs here-- well, almost. Two stages tied-together is a great thing! It acts just like ONE TRIODE if all is done right.
(2) Filament bias. Right-on. There is no better way to run a DHT. BUT--IF one buys the best caps, resistors and is VERY CAREFUL-- he can get cathode bias to work. It IS difficult. There are some advantages to it-- the bias provided by the cathode is signal-related, it's not fixed. I think you can get great results either way. The SOUND of each is different when each is optimized. Which do I prefer? EITHER!
(3) Transformer coupling all the way through. Well...... it CAN sound very excellent, BUT-- it does add some signal delay and some rolling-off of both highs and lows like all iron does. How much? How fat is your wallet? The better the iron, the better it gets. NO argument-- however, you have yet to hear really good D.C.-coupling. If you can get to RMAF 2012, do it. We'll have a ball with this discussion if you do. It will prove a great learning for us both.
(4) "hate" bypassed cathode resistors. I did too until I realized that I had to do it VERY right. The parts cost money, that's for sure. But don't hate it. Get the best resistors-- be sure to use two in parallel on the cathode. These must be within 1/100th of a percent of each other, so buy a few. If you don't get this close, you'll have two signals there. If you do, the two paths will make only one sound like a Tin Can (by comparison-- it's not THAT bad!). I suggest perfecting this before jumping off the cathode bias supply cliff.
(5) Truly, I think this forum has really excellent representation, and that includes the people who argue with me. There is real talent here. I would appreciate that argument not be politically motivated. The object is the best sound-- not the glitziest arguments, graphs, charts and proofs!
Thanks, Andy. I like your post-- a lot.
---Dennis---
Follow Ups:
"IF one buys the best caps, resistors and is VERY CAREFUL-- he can get cathode bias to work. It IS difficult. There are some advantages to it-- the bias provided by the cathode is signal-related, it's not fixed."
Can you please explain how a bypassed cathode resistor causes a signal related bias voltage?
"These [cathode resistors] must be within 1/100th of a percent of each other, so buy a few. If you don't get this close, you'll have two signals there."
If the cathode resistors are bypassed, there is no signal (AC) flowing through them, just DC. The signal (AC) flows through the caps, the DC flows through the resistors. You have two DC paths but (in your case with 6 bypass caps) 6 signal (AC) paths.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
"If the cathode resistors are bypassed, there is no signal (AC) flowing through them, just DC. The signal (AC) flows through the caps, the DC flows through the resistors. You have two DC paths but (in your case with 6 bypass caps) 6 signal (AC) paths."
Agreed, but a far better question is WHY the capacitor {bypass for resistor in cathode bias} used has such a marked influence on sound?
{I use Polycarbonate}
The Mind has No Firewall~ U.S. Army War College.
"I use Polycarbonate...."
I used to use a pot of best Russian caviar bypassed with a dozen snails in garlic, ten frogs legs a la maniere de Borodin, a tub of pate de foie gras, a half dressed photo of Brigitte Bardot and a couple of bottles of vintage Bollinger champagne.
I had to give up in the end - the highs were wonderful but the lows were dreadful.
Not sure what you are saying Andy,
everything is a balancing act, like a Chef with a recipe,
or pushing on a balloon.
There is no magic bullet.
Most people know the order is Teflon, Polystyrene,Polycarbonate, polypropylene, polyester, etch with film tin/aluminum superior to metallized.
The best is no cap, The best is no interconnect {Bill from Audioquest was Right}
So was Theodore Sturgeon
{90 percent of everything is crap}
Everyman aims for his own sparrow, you choose the composition of the arrow within the quiver of your mind.
The Mind has No Firewall~ U.S. Army War College.
Hi Dennis,
Maybe the emphasis should not be about meeting God but avoiding the Devil in all his incarnations.
I don't think we'll ever see the end of discussions about how to bias a tube and how to couple stages together. It's pretty basic and unavoidable.
Compromises all the way. There's no decent DHT with a mu over about 15, and I'd even say 10 generally speaking. So that adds stages, no getting round that. I'm sure there are ingenious solutions like solid state hybrid arrangements and in time they'll come out of the woodwork.
Filament bias is probably the worst system except for all the others - like they say about democracy. It's only practical with a few tubes and requires ultra clean DC supplies and quality wirewound resistors. But given that It works like a charm.
Direct coupling is a no-brainer except for the added design problems that throws up. Back to cathode resistors and bypasses. Not a road I'd ever go down unless I was forced kicking and screaming to do so. I tolerate it in my 300b output but only for want of a better solution.
That's it really - everything we do in the end is because we can't think of a better solution and end up with the least bad compromises.
Maybe we should discuss how to build the "least bad" system rather than the best one. I'm sure that's what we all end up doing in practice.
Andy
"Back to cathode resistors and bypasses. Not a road I'd ever go down unless I was forced kicking and screaming to do so. I tolerate it in my 300b output but only for want of a better solution. "
One word, UltraPath.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
http://db.audioasylum.com/cgi/m.mpl?forum=tubediy&n=11390
Hi Andy,So, specifically from the above post, on your 300B's Rk, use two resistors in parallel, matched to that tight Ohmic tolerance spec Dennis suggests, and tell us what you hear !!!
IT will sound better than one Rk, IF perfectly matched.
You should have a SHORT return path to ground on that Rk, I like two inches or less of wire to a star ground, and USE the resistor leads and body length to comprise part of the distance to ground.
Can this be simulated, well I am not sure. Are Rks in parallel, implemented as described, audibly better ?? :-) OH Yes.
Jeff Medwin
Edits: 08/29/12 08/29/12 08/29/12
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: