JJ 6CA7 a better option for 6L6 amps w/ excessive B+?

[jamesmafyew]

Just curious if anyone has some insight on the JJ 6CA7 and how it relates to their 6L6GC and regular EL34 offerings. Looking at the tube itself, it's just a beam tetrode with the beam-forming plates externalized to pin 1 instead of connected internally to the cathode. Same dual-getter and grid cooling fin as the 6L6GC, and same beam-forming plates and general plate structure, and same bottle. See the picture down below. I get that it's basically supposed to sound like a 6L6 but bias like an EL34. I don't care so much about the 'tone' as much as its potential use in high-stress situations like the Ampeg V-2/VT40/V-4/VT22.

What interested me was that JJ's own datasheet on their 6CA7 lists its Vg2max as 500V, 50V higher than either their 6L6GC or EL34 (450V). And it has the higher screen dissipation rating (8W) of the EL34, along with its higher Vamax of 800V.

Now I know that JJ datasheets are kind of a joke, but if their 6CA7 really does sport these specs, then it seems like a good tube for Ampegs where any extra reliability margin is desirable (yes I know that they would need to be modified with a bias adjustment and have pins 1-8 shorted). And it is definitely built like a 6L6 with beam plates... so potentially would have lower screen dissipation than an EL34 when overdriven...?

Anyone have any experiences or insight on this? I have worked on a surprising number of V-4s lately that have been modified to accept EL34s (bias, pin 1 to ground, and 1K screen resistors). This tube seems like a no-brainer in that situation.

[Leo_Gnardo]

No insight here (yet) on JJ's 6CA7. FWIW I've used some EH 6CA7 on an Orange OR120 hoping they would last longer than any set of current EL34. So far they have done that, OTOH the owner of that amp isn't charmed with the tone change. I have some JJ's on the shelf now but haven't used 'em in anything.

The claim has been made that JJ's 7027 is for all intents identical to their 6L6GC, just has that internal jumper from pin 4 to pin 1. If that's the case then the cheaper 6L6GC is a candidate for use in Ampeg V2, V4, VT22, VT40 and similar amps. Good idea to put 1K screen grid resistors in those amps anyway, regardless what tubes are used. I worked on a late 70's V4 a couple weeks ago that had no screen grid R's at all - of course I put some in.

[jamesmafyew]

It's 100% confirmed that the JJ 7027 is the same as their 6L6GC. I've been using their 6L6GCs in Ampegs for a while now, as they're the most reliable I've found in that application other than the Sovtek "5881." They last fine, but I'm always looking for margin with these amps, and their 6CA7 is the same price.

[Justin Thomas]

Just make sure that's a Sovtek-brand 5881WXT, not a Reflektor-manufactured 5881 reissue or other variant. Just plain old 5881WXT. Cranked em on 10 for an hour in a VT-40, no problems at all. Can't comment on the JJs, except they do good in Fenders.

Justin

[jamesmafyew]

Well I put a set of the JJ 6CA7s in my own V-4 ('77 master volume, 535V B+, 525V screen, 1K screen resistors). Same power output as 6L6GC exactly (102W before clipping on my amp). Biases pretty much like the JJ E34L: -39V grid bias to get 25mA of idle current per tube. Seems very stable, no drift in idle current with heating. Filament winding is happy, still at about 6.2VAC.

We'll see over time how long it lasts in the amp. I have a feeling that it would even be fine with the original 470R screen resistors (at least as fine as any other current production tube). Screen current was exactly the same (measured crudely as voltage drop across the 1K screen resistors w/ true RMS meter) as the JJ 6L6GC: 3.7mA at idle, 22mA just past clipping power.

[Richard]

[ATTACH=CONFIG]38603[/ATTACH]

The paint on that EL34 looks like it got a bit toasty.

[paraph@se]

The JJ 6CA7 has a pronounced bump in the low-mids and is more rounded off in the highs compared to the 6L6GC. The 6L6GC is more scooped in the mids and the lows and highs reach further. The 6L6GC is actually rated 30 Watts as opposed to the 25 Watt rating on the 6CA7. I think the 800v rating is for full-on class B operation where the tube is in cutoff as much as it is swinging 800v. I think the 6L6GC would actually be a stouter tube in this application but the 6CA7's would probably be OK too. You may need to change the bias-feed resistors and up the screen grid resistors.

[nsubulysses]

I just put two different quads of 6CA7 in a V4. BOTH do not bias well, meaning the bias is all over, 12ma, 19ma, 26ma, 22ma. I tried two pairs of matched quads. one new, one pulled from a different amp.

Used but good 6L6s pulled from an amp bias the same

Here's the weird thing -- the tubes bias kind of uneven, but then when you crank the amp to clipping, and bring the amp back down to idle, this bias is even more off. For instance. A tube that biased at 21mA at idle, biased to 12mA and seemed to get stuck there after being brought to clipping. If you let it hang around a bit the bias will drift back up to where it was.

This does not happen with 6L6GC. has anyone seen something like this before, where the bias changes and re-settles in a different position after signal is applied?

[The Dude]

I just put two different quads of 6CA7 in a V4. BOTH do not bias well, meaning the bias is all over, 12ma, 19ma, 26ma, 22ma. I tried two pairs of matched quads. one new, one pulled from a different amp.

Used but good 6L6s pulled from an amp bias the same

Here's the weird thing -- the tubes bias kind of uneven, but then when you crank the amp to clipping, and bring the amp back down to idle, this bias is even more off. For instance. A tube that biased at 21mA at idle, biased to 12mA and seemed to get stuck there after being brought to clipping. If you let it hang around a bit the bias will drift back up to where it was.

This does not happen with 6L6GC. has anyone seen something like this before, where the bias changes and re-settles in a different position after signal is applied?

It sounds like poorly matched tubes to me. Out of curiosity, what brand of tubes are they? When you say "Used but good 6L6s pulled from an amp bias the same", do you mean the same as each other or the same as the other sets with problems? Also, are you sure it's tubes that are mismatched. In other words, if you take the one that's 22mA (or any other), does it meausure 22mA in all 4 sockets?

[Justin Thomas]

I also don't worry too much about it... In my one amp with a quad, I make sure the two sides of the push/pull are balanced. I'd just stick the 12+26 & 19+22 together.

That said, if I had paid for matching service, and I knew my amp was up to par, I'd be a bit upset...

Justin

Edit - Ulysses, that's a weird problem that I'm nowhere near equipped to figure out... But bias creep always wigs me out.

[nsubulysses]

Used but good tubes are JJ 6L6GC that I got from a SLO a few weeks ago. The guy wanted full retube because they were old even though the amp put out full power, 120W. he was going to a recording session

So these tubes in V4 bias very evenly and do about 22V into 4 ohms, 121W at clipping. I'll probably put them in my V4 when my tubes go, har har har

JJ 6CA7 new bias all over. I thought it was the quad so I pulled the 6CA7 from my own amp, and they bias equally all over, including settling in a different place after the amp is driven and then put back to a no signal condition.

I will put both quads of 6CA7 in my amp shortly and see if the new quad works fine and biases fine in mine, but for some reason not in this V4

[g1]

Have you set up the amp to accept 6CA7 as well as 6L6 ?
Or is it just the stock set-up?

[nsubulysses]

oh yeah the old tie pin 1 and 8 together scenario.....

[g1]

470R screen resistors? I doubt it would cause your issues but they are pushing pretty high voltage there. Can't really see it affecting at idle current though.

[nsubulysses]

G, you helped me big time. I've tied pin 1 and 8 together on previous V4s for EL34 configuration. Forgot this time for some reason.

I guess with the suppressor grid floating the current draw per tube can vary in strange ways. Also explains why current draw varied even when control grid, screen grid, and plate voltages seemed the same for all 4 tubes.

These tubes work normal now

They idle relatively close and also settle back down to where you would expect when you go back to no signal condition.

All screen grids have been replaced with 1K/5W. I usually change that resistor that always fails too, 470 ohm 7W. I change it to 500R 10W just so it can dissipate the heat a little easier. I often get these amps with that resistor having one lead disconnected because it gets so hot it melts the solder and it sometimes falls out. Or it is intermittent connection and sparks fly.