Ampeg Gemini 2 recapping problems

[Beyer160]

I added one 470r screen resistor in series with the one I'd already installed on each tube, and maxed out the bias trimpot which got me -56.7v bias voltage. I calculated 22ma current, which puts the dissipation at 10.9w (57%).

Screens are still lighting up when the volume is at maximum.

I'll try upping the screen resistor to 1.47K (1K and 470r in series), but I think I'm barking up the wrong tree.

[g1]

Anecdotal reports on the JJ 7591 are not very flattering. It might be worth trying the new Tung-sols, or if you have room, the big bottle EH versions.

[Beyer160]

Anecdotal reports on the JJ 7591 are not very flattering. It might be worth trying the new Tung-sols, or if you have room, the big bottle EH versions.

I re-installed the original Sylvania 7591s. They sounded flat and splatty (remember, the bias is stupidly cold), but didn't light up when I played. Could it be the JJs are just drawing too much?

It sounds as if my options are to continue adding progressively larger screen resistors in hopes the JJs will eventually behave, or start over with different tubes? I did like the tone of the JJs, but I'm not sure how far I want to go down the rabbit hole. It'd be one thing if I could source my parts locally, but I have to order and wait a week every time I discover something else I need.

[g1]

Can you increase the bias with the sylvania's back to the 12W range and see if the screens behave, and if it still sounds "flat and splatty"?

[Leo_Gnardo]

I re-installed the original Sylvania 7591s. They sounded flat and splatty (remember, the bias is stupidly cold), but didn't light up when I played. Could it be the JJs are just drawing too much?

It sounds as if my options are to continue adding progressively larger screen resistors in hopes the JJs will eventually behave, or start over with different tubes? I did like the tone of the JJs, but I'm not sure how far I want to go down the rabbit hole. It'd be one thing if I could source my parts locally, but I have to order and wait a week every time I discover something else I need.

FWIW when 7591's were nonobtainium, I would swap socket pin assignments and use 6L6 for hi B+ voltage Ampegs, 6V6 for low voltage ones. Nobody complained about the tone, and they didn't bring them back for repairs either, they just kept rockin' on.

Do you have a way to measure power going to the speaker or equivalent load resistor? Another possibility is a bad output transformer. That happened a lot in 60's Ampegs. OT-151 used to croak regularly, typically shorted speaker winding. They'd put out a couple watts at best, sounded great but very limited & wreck output tubes fast. Ohm meter measurements didn't show much: what you'd expect, a fraction of an ohm on the seconday. Checking for power delivery was the clue, if it's not 20 watts or so, then it's a sick OT.

[Beyer160]

Can you increase the bias with the sylvania's back to the 12W range and see if the screens behave, and if it still sounds "flat and splatty"?

You know, my grandmother had a sign in her kitchen that read "the hurrier I go, the behinder I get." I often repeat that bit of folksy wisdom to the people in my life. I should have heeded it today instead of trying out the Sylvanias and forgetting I was still connected through the lightbulb current limiter. Once I unplugged that thing, the amp still sounded a bit "sterile" (57% dissipation, duh) but no longer like an AM radio on its last legs. I'll try biasing back up to 12v, but I think the point has been made.

FWIW when 7591's were nonobtainium, I would swap socket pin assignments and use 6L6 for hi B+ voltage Ampegs, 6V6 for low voltage ones. Nobody complained about the tone, and they didn't bring them back for repairs either, they just kept rockin' on.

I hear you, but I'm not there yet. I know people use modern 7591s in Ampegs, so I know it can be done. I'm a little surprised I'm having so much trouble, though.

you have a way to measure power going to the speaker or equivalent load resistor?

How would that be done?

possibility is a bad output transformer. That happened a lot in 60's Ampegs. OT-151 used to croak regularly, typically shorted speaker winding. They'd put out a couple watts at best, sounded great but very limited & wreck output tubes fast. Ohm meter measurements didn't show much: what you'd expect, a fraction of an ohm on the seconday. Checking for power delivery was the clue, if it's not 20 watts or so, then it's a sick OT.

It's definitely putting out a healthy amount of power- it's nice and loud, more so than my DRRI.

I'm going to put a pair of 25K 10 ohm resistors in parallel (for 12.5K) in the screen/B+ position tomorrow and see what THAT does. Those are the only resistors I have left rated over 1/2 watt. If I can't tame the JJs with that, it may be time for a rethink.


And yeah, I thought it was pretty funny that Hull considered 1910 Fruitgum Company to be "rock 'n' rollers". I wonder what he thought of this...

[Enzo]

My old shop teacher used to say "You never have time to do it right, but you always find time to do it twice."

[Jazz P Bass]

"Measure twice.
Cut once."

[Wes]

Hey, Beyer-

I see you said you were gonna service your VT-22 next. I'd just like to add that if you are going to be rebuilding the power supply please be
aware that one of those big can caps will have a couple hundred volts on it's exterior. Save the covers from the old caps to put on the new ones
if possible.

I'm sure the other fellows here may have other solutions and/or know best. I just wanted to point that out from a safety standpoint.

Good luck!

[Beyer160]

I put the ersatz 12.5k resistor across the screen/B+ (original value had been 470r), rebiased to maximum cold bias (-56.7v), reinstalled the JJ 7591s, and cranked the volume to full. One of the tubes stopped lighting up, the other still lights up very dimly when I really punch it. Of course, the overall volume of the amp has dropped considerably, to a point I don't consider acceptable.

What the hell is going on? Could this really be the fault of the tubes? Are they drawing too much screen current? Should I try to return or exchange them? Other than the bright glowing when I really cranked the volume (which, honestly, I really like to do), I thought they sounded great.

At this point I'm thinking of returning the screen/B+ resistor to 1K, biasing them up and seeing how long I can punish the JJs before they die. I'll deal with getting a new set of Tung-Sols down the road and hope I have better luck with them.

Hey, Beyer-

I see you said you were gonna service your VT-22 next. I'd just like to add that if you are going to be rebuilding the power supply please be
aware that one of those big can caps will have a couple hundred volts on it's exterior. Save the covers from the old caps to put on the new ones
if possible.

Thanks for the heads-up, Wes. My idea was to use individual caps mounted to a stripboard in the chassis instead of cans, but I may never finish my Gemini, so who knows?

[Leo_Gnardo]

I put the ersatz 12.5k resistor across the screen/B+ (original value had been 470r), rebiased to maximum cold bias (-56.7v), reinstalled the JJ 7591s, and cranked the volume to full. One of the tubes stopped lighting up, the other still lights up very dimly when I really punch it. Of course, the overall volume of the amp has dropped considerably, to a point I don't consider acceptable.

Yes I'd expect it to compress like crazy with 12K5 plus a cold bias. Somewhere inbetween I think you'll find a setting that will keep you happy(ier). FWIW with those glowing-grid EL34s in Marshalls it was surprising they would keep working in spite of the punishment. So -

At this point I'm thinking of returning the screen/B+ resistor to 1K, biasing them up and seeing how long I can punish the JJs before they die. I'll deal with getting a new set of Tung-Sols down the road and hope I have better luck with them.

More or less, that's what I would do. And better luck with (Sovtek) Tung-Sols, recently I've had to replace a pair of TS 6L6's that rattled like maracas, and today TS 7581 that did similar, rattle-buzz along with low notes. Hope they got their act together better for 7591. Don't forget there's an escape hatch, using 6L6GC instead after swapping a couple socket pins.

[Beyer160]

Yes I'd expect it to compress like crazy with 12K5 plus a cold bias. Somewhere inbetween I think you'll find a setting that will keep you happy(ier). FWIW with those glowing-grid EL34s in Marshalls it was surprising they would keep working in spite of the punishment. So -

I'm going to order a pile of resistors tomorrow and swap them out until I get a reasonable middle ground. Or, until I get fed up and take your advice about the 6L6s.

And better luck with (Sovtek) Tung-Sols, recently I've had to replace a pair of TS 6L6's that rattled like maracas, and today TS 7581 that did similar, rattle-buzz along with low notes. Hope they got their act together better for 7591. Don't forget there's an escape hatch, using 6L6GC instead after swapping a couple socket pins.

Ugh, don't tell me that! I'd heard the Tung-Sol 7591 (from what I understand, a repinned EH7581) was the best production 7591 around.

Hmmm, I do like 6L6s, though...