Ampeg Gemini 2 recapping problems

[Beyer160]

Heater voltage is 2.64v on one leg, 3.82v on the other- so 6.46v, correct?

Plate to cathode is 496v.

I think I'm going to tweak the bias to get about 65% plate dissipation (12.35v) and go from there. The datasheet for the JJ 7591 indicates it can take 550v on the plate, so in theory 497v shouldn't hurt it, right?

[g1]

The plate voltage should be ok, as long as you keep the dissipation within reason. I wouldn't go any higher than the 65% you mentioned. There will not be any significant cathode voltage to subtract from the plate figure.
The note I mentioned is above the box in the lower left area, marked **. If you could read that number you could figure out what kind of idle dissipation they had with 430V.

[Leo_Gnardo]

The schematic lists bias current at -22v.

Voltage, you mean. Don't sweat the bias voltage, let it be whatever it needs to control the bias current so you have correct plate dissipation and minimization of crossover distortion. I expect Vb will wind up in the high -20's.

If we set bias voltage according to schematics there would be an awful lot of output tubes overheating. Fender specified -47V for 6L6's, Marshall -37V for EL34. Took a while for me to understand those were to be taken with a grain or nineteen of salt. Plate current is what's important in the bias game.

[Beyer160]

The note I mentioned is above the box in the lower left area, marked **. If you could read that number you could figure out what kind of idle dissipation they had with 430V.

Sorry, I misunderstood-
"** Factory adjusted to give 80ma plate current for both V1 and V2"

Assuming you split that between V1 and V2, that's 40ma each, which is about what I read at the beginning (38ma and 39ma). At 430v that comes out to 17.2w dissipation. That seems high for factory bias, but I guess those old 7591s could take a licking and keep on ticking. The ones that came with the amp appear to be original from 1967.

Voltage, you mean. Don't sweat the bias voltage, let it be whatever it needs to control the bias current so you have correct plate dissipation and minimization of crossover distortion. I expect Vb will wind up in the high -20's.

If we set bias voltage according to schematics there would be an awful lot of output tubes overheating. Fender specified -47V for 6L6's, Marshall -37V for EL34. Took a while for me to understand those were to be taken with a grain or nineteen of salt. Plate current is what's important in the bias game.

Voltage, yes- thank you. It irritates the hell out of me when people misuse technical terms in my business, I appreciate the correction.

So I dialed the bias in to get plate dissipation at 12.22w, or about 64% and the B+ voltage settled at 488v. The amp sounds great.


Thanks to everyone for your help, I really appreciate your patience with my beginner level questions. I feel like I have a much better understanding of the bias procedure, as well as how the amp works in general.

Now that I seem to have the Gemini taken care of, I'll be moving on to my next project- a cap job on my '72 Ampeg VT22. I suspect that one will be a little more challenging, which is why I tackled the Gemini first.

[Leo_Gnardo]

Now that I seem to have the Gemini taken care of, I'll be moving on to my next project- a cap job on my '72 Ampeg VT22. I suspect that one will be a little more challenging, which is why I tackled the Gemini first.

Better eat your Wheaties & get in some weight-lifting practice. That VT-22 is a bear to just move from here to there, never mind pick up.

Glad to hear your Gemini's now ship-shape, hope you get lots of playing enjoyment out of it. I used to laugh at them but now think they're very respectable amps.

[Beyer160]

Better eat your Wheaties & get in some weight-lifting practice. That VT-22 is a bear to just move from here to there, never mind pick up.

My VT22 had no casters when I bought it (neither did my '70s SVT cabinet, come to think of it), that was the first thing I did to it.

I saw a friend's band once, his sister was the keyboard player. Not only was she beautiful, but she didn't take shit from anyone and could handle herself quite well in the "boys club" mentality of the rock band world. I saw her loading out after the gig, trying to heave her VT22 into the back of a Jeep while wearing a dress. I came up and asked;

"It wouldn't be sexist of me to ask if you needed help with that, would it?"

"Hell, no!" she replied, "this thing weighs a goddamn TON!"

Glad to hear your Gemini's now ship-shape, hope you get lots of playing enjoyment out of it. I used to laugh at them but now think they're very respectable amps.

I'm partial to old Ampegs. They're quirky and don't behave like Fenders, but have a cool sound all their own. I can't afford a blackface Deluxe anyway, and for the same money as a used reissue I got a handwired amp with Amperex Bugle Boys and a '60s Jensen Concert. My VT22 doesn't quite get that sparkly clean tone of a Twin, but cranking the mid will get you some throaty dirt that a Twin could never dream of doing.

[Justin Thomas]

Ampegs are a beast unto themselves. Love em. Nobt too fond of some of the construction practices in the 70s, but easily forgivable. Oddball wacky tubes? A bit less forgiveable. The sound? Juuuuuuuust about right. Everyone in the house is waiting for me to get this VT-40 head working right...

Justin

[Beyer160]

So I put the amp back together and it sounds really, really nice. I even removed the Baxendall EQ "wafer" that holds the components for the EQ circuit in channel 1, and replaced it with discrete components (the jury is still out on that one, pending further testing).

All was great until I cranked the volume up to full and let her rip. The sound was amazing, but I noticed the power tube filaments started glowing bright orange as I played, brighter as I played louder and settling down again as the chord faded out. I've seen tubes glow blue like that as you pushed them, but I've never seen the filaments do that. I didn't get any redplating, just the filaments glowing brighter. I put the amp back on my lightbulb current limiter, and watched the bulb flash as I played (the tone started getting ragged as the bulb took more current)- this kept the tube filaments from lighting up.

Is this normal, or do I need to find a way to drop that B+ voltage after all?

[Tom Phillips]

...I noticed the power tube filaments started glowing bright orange as I played, brighter as I played louder and settling down again as the chord faded out. I've seen tubes glow blue like that as you pushed them, but I've never seen the filaments do that. ...
Is this normal, or do I need to find a way to drop that B+ voltage after all?

That is absolutely not normal. I also suspect that it isn't the filaments (heaters) that are glowing brighter. Maybe the screens? There is just no mechanism to cause that except a heater to B+ short which would take out the tube completely or blow a fuse. For further evidence you could monitor the 6.3V heater supply while you play and report back.
Anyway, those are my best thoughts on the situation.
Tom

[52 Bill]

Or are the tubes red plating when being driven hard?

What part of the tube is glowing? Please describe.

[Leo_Gnardo]

That is absolutely not normal. I also suspect that it isn't the filaments (heaters) that are glowing brighter. Maybe the screens? There is just no mechanism to cause that except a heater to B+ short which would take out the tube completely or blow a fuse. For further evidence you could monitor the 6.3V heater supply while you play and report back.
Anyway, those are my best thoughts on the situation.
Tom

I'm voting for screen glow. If the bias is 38-39 mA as described in post 19, I'd dial that back to 30 or hi 20's. Also, schematic shows no screen grid resistors, and sg's tied to a point just barely below full B+. Plus a hand written note "tie B to A" iow run screens right off the B+. With low B+ amps @ 300V running 6V6 you can get away with this, Ampeg did with Jets & similar, also seen in old Gibsons and such, but with a 500V supply I'd drop the voltage to those screen grids a bit. IF there is a 470R from B+ to the screen grid branch of the power supply, I'd up that to 1K or 1K5, and put a single 1K 5W resistor in series with each screen grid. That oughta tame them. You don't want to melt those grids.

FWIW I have seen this effect in some old Marshalls equipped with EL34, driven hard.

Direct from Ampeg's 1960's one-page owner's manual: "If you hear distortion, turn the amp's volume down." Riiiiight.... They were aiming for the polite accordionist market , had little idea about what electric guitarists would be looking for. A cranked up Ampeg sounds terrific

[Beyer160]

Or are the tubes red plating when being driven hard?

What part of the tube is glowing? Please describe.

First pic is at idle, second pic was taken while playing with the volume wide open





I never saw any redplating.

Looking at the pictures, it does appear to be the screens. D'oh!


I did up the B+/screen grid resistor from 470r to 1K, and added 470r in series with the screen grids on the 7591s (originally there was just a piece of wire connecting the screen grid pins). I'll try adding a 470r in series to the ones on the tube grids, and another on the B+/screen if need be.

[Beyer160]

Direct from Ampeg's 1960's one-page owner's manual: "If you hear distortion, turn the amp's volume down." Riiiiight.... They were aiming for the polite accordionist market , had little idea about what electric guitarists would be looking for. A cranked up Ampeg sounds terrific

"Ampeg: Story Behind The Sound" relates a story of Ampeg founder Everett Hull telling his employees to get the band 1910 Fruitgum Company out of his factory because he hated rock 'n' rollers.

And yes, my cranked Gemini sounded terrific... in the few seconds before I saw I was cooking the screens!

[g1]

JJ 7591S are rated 440V max. for screens.
Also they require more negative voltage than NOS types, so for example, one report was -26V bias required for JJ's compared to -22V for NOS.
Like some of the others mentioned, as well as dropping screen voltage you should reduce idle dissipation.

[Leo_Gnardo]

"Ampeg: Story Behind The Sound" relates a story of Ampeg founder Everett Hull telling his employees to get the band 1910 Fruitgum Company out of his factory because he hated rock 'n' rollers.

Somebody should have told Everett Hull, 1910 Fruitgum Company were lightweight bubblegum pop posers, more like The Archies. Or Strawberry Alarm Clock. Wasn't long before the Stones stormed Madison Square Garden with a dozen or two SVT rigs, now that's ROCK!

Bands like that make me glad music moved on. The bubblegum period didn't much impress. I would have thrown them out too.