Alternative to 6386 tube

[PRR]

> At very low gain (CV = -10 ... 11V), the transfer courve shows the opposite kind of nonlinearity: Its an "S" with its ends bent _upwards_.

Yes. That surprised me too at first (in simulation), but it is understandable.

With grids at -10V, gain is very low for small signals. But with say 5V peaks, one grid swings from -15V to -5V, the other grid swings -5V to -15V, both around the -10V control voltage. When a grid is at -5V it has VERY much more gain than at -10V, so the peaks "pop up". Opposite of most amp designs where peaks go flat.

I tried cathode resistors and current sources. Best result may be with a small (like 50Ω) common cathode resistor or a few K in the plate, but this is not much different from just grounding the cathodes.

This is your main distortion source when in deep gain-reduction, so it limits the maximum level you can feed it. 10V peak is really too much for most small tubes; 1 to 3 volts is all they can take without getting seriously bent.

 

[zebra50]

[quote author="Swedish Chef"]I agree!
I built a 436 to begin with without trimers and it sucked hard! :grin:
Sorry I omitted to say that i AM using trims on each stage
chef[/quote]

...and of course I put trimmers on mine and took them off again because I preferred the sucky sound of it without! I auditioned about 10 6BC8s and didn't notice much difference in behaviour between any of them - all equally good/bad.

 

[jhaible]

[quote author="PRR"]> At very low gain (CV = -10 ... 11V), the transfer courve shows the opposite kind of nonlinearity: Its an "S" with its ends bent _upwards_.

Yes. That surprised me too at first (in simulation), but it is understandable.

With grids at -10V, gain is very low for small signals. But with say 5V peaks, one grid swings from -15V to -5V, the other grid swings -5V to -15V, both around the -10V control voltage. When a grid is at -5V it has VERY much more gain than at -10V, so the peaks "pop up". Opposite of most amp designs where peaks go flat. [/quote]

Oh yes - this makes a lot of sense! Thanks for this hint.

JH.

 

[Swedish Chef]

[quote author="PRR"]> With grids at -10V, gain is very low for small signals. But with say 5V peaks, one grid swings from -15V to -5V, the other grid swings -5V to -15V, both around the -10V control voltage. When a grid is at -5V it has VERY much more gain than at -10V, so the peaks "pop up". Opposite of most amp designs where peaks go flat....

This is your main distortion source when in deep gain-reduction, so it limits the maximum level you can feed it. 10V peak is really too much for most small tubes; 1 to 3 volts is all they can take without getting seriously bent.[/quote]

That makes alot of sense...hmmm.

 

[Rob Flinn]

Swedish Chef wrote:
I agree!
I built a 436 to begin with without trimers and it sucked hard!
Sorry I omitted to say that i AM using trims on each stage
chef


...and of course I put trimmers on mine and took them off again because I preferred the sucky sound of it without! I auditioned about 10 6BC8s and didn't notice much difference in behaviour between any of them - all equally good/bad.

I`m kind of with you on this one Stewart.

From what I remember about my experiments building my 436c, I came to the conclusion that if you want a "silk purse" sounding vari mu, then this isn`t the best one to play with. I put trim pots in mine, but didn`t really find it made it more useful. I think its great at what it does, but the sta levels I`ve built have always been more versatile in use. Its fair to say the sta level is more complicated, the parts are more expensive, & of course you need the 6386! One thing with the sta level is that the trim on the 6v6 o/p stage, actually trims everything before it to an extent. OK its not matching quite as well as individual trims on each stage.

As for a 6386 substitute maybe the RCA Ba-6a is the way to go, using 6sk7, with dual trims. The 6sk7 is really cheap to buy & there are plentful quantities around. Also, because they are not dual devices thay are a lot easier to match even before you`ve played with the 2 trim pots.

 

[weiss]

I know i'm bumping an old thread here but i need to find a substitute for two 6386s in my 95' vari mu. would the 6BC8 or 6BA6 be equally usable or do i have to go for the JJ 6386? I have no real experience with tubes and their various characteristics..

 

[pucho812]

JJ Tessa tubes makes 6386.  Its nice and the price is far about 120usd per tube

 

[weiss]

JJ Tessa tubes makes 6386.  Its nice and the price is far about 120usd per tube

yes, found one for 100€. and what about a WE 396A (2C51/6N3P) NOS? approx. half the price, would i notice any difference? or do i lose a lot of magic and open sound?

 

[gyraf]

A good friend of mine recently bought and traced a rather big batch of brand-new JJ6386's (to gain knowledge about the 670's workings) - and found nearly no resemblance with the handful of NOS real ones that he had for reference.

As long as you can get the originals at (comparably) decent prices, I would stay away from new production ones.

Jakob E.

 

[VertigoRecording]

A good friend of mine recently bought and traced a rather big batch of brand-new JJ6386's (to gain knowledge about the 670's workings) - and found nearly no resemblance with the handful of NOS real ones that he had for reference.

Hi Jacob,
I totally agree with his findings. The new JJ 6386 tubes are useless and do not perform well at all in a vintage, non-recreated, non-tribute Fairchild 660 or 670. In all fairness, some of the very early releases of the JJ 6386 tubes worked pretty well. I have a friend who has a set from maybe 10 years ago, or whenever they were first manufactured. At this point they have totally lost the thread and don't seem to care about fixing the problems. If you put them in a real Fairchild 670, they barely compress and you have to hit them really hard for even that pitiful result.