Hi, all. After years of procrastinating, I'm finally working on my Gemini. I've had this amp since the late '70s. It's the 7591 / 7199 version.
The amp has always disappointed me in terms of sound. It always sounded dull, but since there were some obvious issues, I never got rid of it. The first issue is the reverb--the tank was missing when I bought it.
Yes, the power tubes are old. And yeah, they need to be replaced. It should be said that there was some arcing, loss of volume and a high pitch squeal when the treble controls were turned up (more about the treble controls later.)
Youtube of the 7591s when the treble is cranked (you cannot hear the hi-pitched squeal on the camera's mic, and it's fairly muted anyway--not an out-of-control oscillation.)
I bought a new tank a year or so back (yep, it's an appropriate impedance for the ampeg reverb schema), and replaced the C6G7 reverb driver tube, but no dice.
So three days ago I pulled the chassis. The resistor for the screen supply has some heat damage, I'm sure it has something to do with the arcing (you can see it between tags 62 and 69 on the board.) It needs replacing, but is still near it's speced value.
And the bias supply cap looks bad (Gold medium size cap in the chassis pic), so that's on the to-do list, but the supply still reads -22V.
RE: the reverb-- first thing I noticed was a missing ground on one of the reverb lines. OK, this is a ground-loop thing--connect the grounds inside the reverb, not at the chassis. Success! The reverb is functional for the first time in 30 years.
Here's where the story gets weird.... Both reverb and tremolo are available on both channels. And that's not supposed to be the case. There's a strange interaction between the controls--the effect on each channel can be eliminated by turning down the tone on the other channel. And the volume controls on the other channel act more like a tone mixer.
Once the reverb was working, the power tubes stopped squealing and arcing--treble settings now had no detrimental effect.
So there are two slotted POTs on the back of the chassis. One is a 100 ohm rheostat to balance the 7591 cathodes, the other one I thought was the bias voltage (would make sense, right?) But no, it seems to connect the signal path from the tag board to the Echo (reverb depth) POT.
You can see the routing for the pot in this photo (prominent grey shielded wire to and from the pot at the bottom.) You can also see the real bias voltage adjustment--the little blue trim pot on the tag board.
Did some evil genius modify this years ago? Should I leave it? I sort of like a lot of the sounds that result from knob twiddling...despite their non-intuitive interactions.
Edit: correct schematic below
Last edited by gmoon; 02-12-2009 at 04:31 PM. Reason: Remove incorrect schematic drawing.
Is it safe to say that the slotted stubby-shaft POT was the original Bias Voltage adjustment? To match the schematic...
I don't have anything to add, but will be watching this carefully, as I, too have a Gemini II that has been sitting unused for years because it sounds dull--not like the one I played some time back that was crisp, punchy and wonderful. I suspect a cap job and a JBL would go a long way, but for now I will watch your progress, which will hopefully inspire me!
Crisp, punchy and wonderful were not terms I'd have used to describe mine... Maybe with the bias cap replaced; and now that 7591's are readily available...
Re: the speaker--I'm in the same spot, it's not the original. Fliptops recommends some replacements, if vintage can't be found.
You didnt mention it,but if you havent changed the filter caps,I would do so.I havent met a Gemini that didnt benefit from new filters.
There's a 100 ohm bias "balancing" rheostat on the 7581's cathodes. That's still intact.
The bias voltage POT is 15K. That's the one I was speculating re: original purpose. Sorry about the confusion.
Sounds like you are on the right track. I've done a ground-up rebuild on a Gemini I (all I started with was the chassis) and just did a major tune-up on a VT-22. There are a couple things that you might consider:
1) That screen resistor is working really hard. I replaced it with a Vishay/Dale 3 watter (very small, Mouser #71-CW2B-1.0K). I also put 470R 3 watt screen resistors on the tube sockets. With household volts going from 110 to 120 since then and that you might have to retube with eastern European 7591's, I don't think it will hurt.
2) I agree on the filter cap idea. Fortunately, exact fit and value canister caps are being manufactured again. Fliptops and Antique Electronics have them.
3) Have a look at this site:
One recommendation I got there was replacing the tone stack module with discreet components. The entire Baxandall circuit is baked into what looks like a cracker with 7 wire leads sticking out the bottom. It is located behind the front board. They aren't famously the best parts. You can use a better grade of resistors and caps. It makes a difference. Search around for the pictures. If you don't find them, let me know and I'll send you the ones I saved.
4) Cool amp! With so many little dirt machines around it is nice to plug into something clean and sweet at the same time. And the reverb ...
Thanks for your thoughts, Skip.
1) The HV is definitely higher than the schematic value. The screen voltage is about 5V lower than the (new, 120VAC) plate voltage, which is close to 500V. Yep, I think an additional screen resistor for each 7591 is a good idea.
2) Thanks--using new can caps is a good idea (vs. the alternative..)
3) Hadn't seen that one. It's a little thin on the older Ampegs, but looks like it's just getting started. And thanks for the pics offer.
The reverb is really nice.
I'll have more to say about this amp as time goes on. The older mod definitely "chains" channel two into channel one, before the chan one tone stack and tremolo. There's a definite gain (and noise) increase, along with a more "scooped" sound. I'll trace the wiring modification and post it.
Those volts are way too high. I suspect it makes the amp a little harsh. You might think about redoing the standby with a big zener on the center-tap. It will mess up the bias voltage too so have a couple resistors nearby.
One of the schematics I've got shows 720VCT on the secondary but only 430 at the first node. At that draw, 500 volts would seem closer than 430. I just think 7591s seem happiest between 420 and 450 fixed bias.
Yep, the plate voltage is about 490V.
If I use several 12-15V zeners in series on the CT to drop the voltage, are 5-watt zeners sufficient?
I've always used the big bolt-on units. If you want to see schematics that use smaller ones, have a look at the Electar Trainwreck designs on Blueguitar.org. or Schematicheaven. Those are 5 watters IIRC. sh
-- Definitely needs a new speaker. Been playing it through a vintage 2x12 cab (nothing special) and it sounds much brighter, alive and looses the "thud." Went back to the (non-original) 15 in. in the Ampeg, and the difference is confirmed.
-- Noticed there was still some redplating at higher volumes, even though the weird "treble knob" behavior had stopped. Changed out the toasted screen resistor with a 1.5K 5-watt, that fixed it. I'll take your suggestion about separate resistors into serious consideration...
Cool stuff guys! I have been working on three different Ampegs (Jet, Reverbrocket, B-12XT) lately and like to stay up on what other guys are doing with their amps...
keep going on this, and I do hope you post more pictures
Just a little update about the Gemini II:
-- Cord replaced, death-cap removed.
-- Removed the modified channel-routing.
-- Most of the filter caps replaced, the can cap is on order. The amp is, and always has been, very quiet. But it's just routine maintenance for an older amp..
-- Replaced the 7591's.
-- Found a vintage Heppner 15" Alnico speaker. I know opinions vary on these, but, wow, what an improvement... (well, that and the power tubes.)
There are still quirks to case down. There is still some reverb on the second channel. Maybe 1/4 of the amount on the first chan. Maybe that's normal, I don't know. The channels don't sound identical, even with the reverb an trem off...
But the amp has been transformed from sounding "dead thud" to bright and very much alive. With the treble past 3 oclock and the volume above 2/3rds, it odrives nicely. Sure, it's not a Bassman, but anyone who says the G-15 is "clean only" needs to change the 7591s...
Finally chased down one remaining issue with the Gemini II today--one of my own making.
I'd done some mods of the incoming mains wiring, including the usual three-prong plug, ditching the death-cap, etc. In the process I added four 5W 12V zeners on the PT secondary centertap to drop the voltage a bit closer to 1965-66 levels...
Being too clever for my own good, I modified the standby/polarity switch to for a choice of either the low or the full voltage (no zeners.) The mod words well, but the lower voltage is always noiser...especially on channel one, which also needs to be zeroed when using channel two.
However, the Gemini has an unusual (to me) way of wiring the power lamp that works in conjunction with the standby. I just discovered that the lamp, etc., was wired such that in the low voltage setting it was on the PT side of the zeners, and therefore elevated at that setting... sigh.
Sorted it out by using the unused side (formally death-cap) of the standby/polarity switch for the pilot lamp. It's now super quiet--I can even jumper the two channels together without noise for the first time...
There are currently 1 users browsing this thread. (0 members and 1 guests)