Ampeg Gemini II 7591 biasing

[Apexelectric]

Just picked up an all original 1966 Ampeg G-15 Gemini II recently and am doing the typical cleanup and age related maintainence as well as a few other items. The amp was in pretty good shape overall and doesn’t look like it’s had much of anything done to it over the years. What I have done was the filter caps (4 section cap can + two axial), power cord, death cap, output tubes (one was good and one weak), one preamp tube, changed the 7199 driver/pi and rewired the tube socket to take a 6GH8 instead. The reverb and tremolo are both functioning which was a plus, glad I didn’t have to swap out the trem module.

The dilemma I’m having is with the plate voltage and plate current. The schematic on the back of the amp is showing 430 volts at the plates of the output tubes and mine is delivering 500 at the 123 wall volts. The schematic shows the bias being set at 80ma for the pair of 7591’s at the 430 plate volts.  Even with the original voltage numbers they have this amp running the tubes near their maximum 19W pdis. Let alone at 70% of max for Class AB. Is this just the way they ran these amps or am I missing something here.

The way I look at it now is that I need to change the bias resistor to get a more useable range and possibly run this amp through a bucking transformer to tame the voltage a bit. I’d like a little guidance on how to best address this situation from someone who might have had a similar experience with these amps. I’m not too familiar with Ampegs in general so some insight into how these are typically setup would be helpful. Any other advice on these amps would be welcome too.

Thanks

[sluckey]

The schematic shows the bias being set at 80ma for the pair of 7591’s at the 430 plate volts.  Even with the original voltage numbers they have this amp running the tubes past their maximum 19W pdis. Let alone at 70% of max for Class AB. Is this just the way they ran these amps or am I missing something here.

80mA for the pair means 40mA each. 40mA * 430V = 17.2Watts. That's about right for a 35 watt amp. But... 430V is not the plate voltage. 430V is the B+ at node A. The real plate voltage will be less than 430V. In fact, the real plate voltage will be the voltage measured between the plate and cathode.

This is how I would set the bias today... Disconnect the wiper of the balance pot from ground and connect an ammeter between ground and the floating wiper of the balance pot. Set the bias pot for 80mA (or your desired current). Use a second meter to measure the voltage between V1-5 and V2-5 (cathode to cathode). Adjust the balance pot for zero volts. May have to repeat this procedure a few times until you have both pots set correctly.

Alternately, permanently install a 1Ω 1% 1W resistor between the balance pot wiper and ground. Now set the bias pot for 80mV (or your desired voltage) across this resistor. Now adjust the balance pot for zero volts between V1-5 and V2-5. May have to repeat this procedure a few times until you have both pots set correctly.

If you find the bias pot will not lower the current enough for your liking, then lower the value of that 220K that's between the PT winding and the bias diode.

This was the first real amp I ever owned. Bought it new in '68. The only time I ever messed with the bias I just set the bias pot for -22V on the grids and balance pot for zero volts between the cathodes. Sold the amp 7 years later. Wish I had it back! 

[Apexelectric]

Hey Sluckey,

I had a feeling you would chime in on this one.

This one has just the bias pot and no balance adjustment.

Measuring plate to cathode on the tube socket, with my wall voltage at 123VAC, it is at 500V which is also about the same measurement I get at node A at the cap can.
The bias voltage is set at -26V, which is as low as I can go. This has my current more like 45ma for each tube.
I did install some 1 ohm resistors off the tube sockets for measuring current.

I tried the amp plugged into my bucking transformer and with a 12V drop in the wall voltage it runs dead quiet and closer to the numbers shown on the schematic. Filaments are right around 6V. At a 6V drop in the wall voltage filament voltage is right at 6.3 but a slight hum returns and plate volts comes in at around 465 and current at 32ma each. Better, but still a little high if we’re shooting for 70% of max.

[sluckey]

If you find the bias pot will not lower the current enough for your liking, then lower the value of that 220K that's between the PT winding and the bias diode.

[Apexelectric]

Got it. Thanks.

Any thoughts on how hot i should run the tubes before I’m pushing my luck? I’m going to swap the bias resistor regardless but given the fact that my B+ is at 500V as opposed to 430V my inclination would be to permanently install the bucking transformer, drop the 12V, and bias after that.

Make sense to you or should I go about it differently?

[shooter]

an alt idea to the bucking T;

I'm looking at this schematic;

https://el34world.com/charts/Schematics/files/Ampeg/Ampeg_g15.pdf

The "B" tap is ||'d off the main PS rail, Do the same kinda thing to make a "New" A tap for the PA, see sketch
tweak the 2.2K as desired
anyway, fwiw

[sluckey]

Why bother? Those tubes can handle 500V.

[shooter]

Why bother?

I agree, just throwing spaghetti at the wall 

I would however ensure the caps can

[Apexelectric]

I’m Italian so I like spaghetti. 



 

[shooter]

I’m Italian

My 1st Wife was Italian, we had a Basset hound I named Banjo, when he got the leftovers, he'd spit the noodles on the wall, sauce, meat, swallowed, was a great thing to watch!

[Apexelectric]

 

I used to have Bassett hounds with my first wife as well. Ricochet and Mudflap.
They didn’t do the spaghetti trick though.

[Apexelectric]

Back to the topic though, it seems to me that the amp is happier with the proper voltage. There is a fair amount of noise that gets introduced at the higher voltage. Maybe the balance pot could help there if I added one. From what I’ve figured, this thing runs right at 110V supply which is not an option in this century, and still low for the time this thing was manufactured, without dropping the entire power supply. The filaments run at almost 7 volts now, more like 6 volts when I get the B+ down around 430. Amp runs nice and quiet then.

[sluckey]

I would use a bucking transformer and change the bias range resistor.

[Sonny ReVerb]

The plates can handle 550V but the limit for the screens is 440V. I would bet they are well over that at 500V on the plates.

[Apexelectric]

The plates can handle 550V but the limit for the screens is 440V. I would bet they are well over that at 500V on the plates.

Noticed that too Screens are only a few volts lower.