6386=5670=2C51W=396A?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I read on the internetZ that someone found a replacement although it has a different pinout.. but of course they don't say what it is...

BUT the clues lead to an ECC189..

EDIT: someone with a varimu that needs a 6386 should try some of the others.. and let us know how they work.
 
The search facility is littered with posts from PRR on this very issue.

In short, there is no direct replacement for the 6386.

Not only are remote-cutoff triodes relatively obscure in the grand scheme of things, but dual remote-cutoff triodes are the proverbial rocking horse's manure.

The 5670 is not a remote cutoff tube, and as such will start to distort, due to the plate becoming current starved, above about 12dB of GR.

To my knowledge, the best option - if you don't want open-wallet surgery for 6386s (or risk buying fakes - seen them!) - is an adaptation of the 6BA6 Pentode. Manley match a pair of 6BA6s in their T-Bar mod, and the Transfer Curve is virtually identical to the 6386.

If you search under 6386 at Prosoundweb, you'll see some transfer curves posted by Lab member Larry, illustrating a comparison between 6BA6 and 6386 - pretty close.

I won't name drop, but I was told a few years back about the 6BA6 conversion by a designer who'd been through the same rigmarole. If I'd known what I know now, I would've just trusted his judgement, but hindsight has proven his judgement to be correct.

IIRC, PRR suggested giving the 6ES8 a try - anyone?

The PCC189 is used in Jakob's variable-mu compressor, as well as the EAR 660. I love the EAR - very smooth.


Justin
 
BTW - I've seen 2C51 from Westinghouse reprinted with the 5-Star logo... Be very careful - the 2c51 is not an RC valve IIRC.

Original 6386 have the logo etched into the glass, not silkscreened. 6386 were also rebadged for other brands such as Raytheon, but made by RCA.

Justin
 
yes I read PRR's posts.

If the 5670 and others are not direct replacements then why do all of the old tube dudes and websites show them as direct replacements?

I'm NOT a tube guy and I have literally no experience with tubes so I'm pretty worthless at determining a replacement. I thought I could at least urge a debate over this.
 
The 5670 has a pretty linear transfer curve and is a dual-triode, which is why I suspect it's recommended. It's not a remote-cutoff type, and Manley (for example) admit that it's not a good move to push it into deep GR.

There is a frightening amount of BS around regarding variable-mu compressors - the group is very lucky to have PRR as a moderator.

BTW - if a valve isn't a remote-cutoff type, there's little debate to be had: it either has the RC-type grid spacing or it doesn't.

Justin
 
this would prob be a hot topic if prr, larrchild and nydave were still around here- i havent seen any of em in a while.

anyway word from larr and possibly others is that a PAIR of matched 6ba6 wired in triode makes a very good sub tho not pin compatible by any means!
 
Here's the thread you want: http://recforums.prosoundweb.com/index.php/m/76243/998/?srch=vari+mu#msg_76243

PRR still crops up (he's just made a post at his group) and Jakob's excused from posting as he's just had a baby.


Justin
 
Understood.

Rob, now we have tangible proof that they aren't replacements.

Anyone tried to rewire an ecc189 for a 6386 spot? I read on a forum that someone does this and it works.. ? a little adaptor..?
 
[quote author="Svart"]Understood.

Rob, now we have tangible proof that they aren't replacements.

Anyone tried to rewire an ecc189 for a 6386 spot? I read on a forum that someone does this and it works.. ? a little adaptor..?[/quote]

I was surprised that all the other information people offered wasn't tangible proof !!!

I never tried the 6es8/ecc189 out. But if you get a PCC189 they are even cheaper & the only difference is that the heaters are a different voltage (7.6v)
 
I built a 660-like compressor (near-clone), and also got a Sta-Level going with 6BA6 tubes. The filament supply for a Fairchild needs to be strong because the current consumption of the 6BA6's is rather substantial.
 
I've used 6BJ6's in a Sta-Level with good results, and it consumes half the filament current as a 6BA6. IIRC, 2 6BJ6's draws the same filament current as one 6386.

Original 6386 have the logo etched into the glass, not silkscreened. 6386 were also rebadged for other brands such as Raytheon, but made by RCA.

I heard that they were all manufactured by GE, and then re-branded as RCA, etc.(?). In any case, watch for fakes. I found 3 in my stash that apon close inspection look totally different inside than a real 6386. No wonder they sounded crappy!

Manley switched to the 5670 for the VariableMu compressor when 6386s got scarce, but there are circuit changes to accomodate it. This may be why so many people try to pass the 5670 off as a replacement.
 
The 6es8 is really good in simple varimu circuits like the Altecs. Russian 6N3P-EV is great too with good inherent matching. None of this is probably not helping you so I just butt out now... (insert disappearing butt icon here)
 
someone should invent a remote cutoff mosfet or transistor.
There are a million remotes out there, some hiding in the catalogs under pretensious names.
Make a list and try them all.
 
[quote author="CJ"]someone should invent a remote cutoff mosfet or transistor.
[/quote]

http://www.fairchildsemi.com/ds/JF/JFTJ105.pdf
A couple of selected 'big-momma' channel J105 fets may even work.
 
Not a real sub, but if you want to stick with double triodes, just for fun, check out the grid-plate curves on the 12BH7A. Not that it has bendy curves of the venerable 6386, but at -14V grid the 12BH7 still chugs 4ma and it takes -23V on the grid for cut-off, at least on paper. About the same plate resistance and amplification factor though.
It will take abuse too. You can have fido hump it for days, it will still work.
 
I am surprised that nobody has tried the 12AV7 for varimu duty.
So I got a bunch of russian remote cutoff pentodes. Good tubes. The russian EB (EV) rated tubes are always well made. For a year now I have been messing with an automated tester to plot grid/plate curves and save them to the computer. That way I will just plug them into a spreadsheet and see the closest matches. Measuring cathode current through a small resistor only though. Everything has been cheap surplus like switching power supplies except the LabJack with USB port that does the A/D conversion, the most expesive part. The grid voltage is driven by a D/A converter through a high voltage opamp from about 0 down to -50V. Trouble is when you get down to the 100uA range measurements get jittery.
 

Latest posts

Back
Top