Ampeg G-15 Gemini ll

[g1]

BIAS: Yes, I see the bias is marked as -50v but doesn't that sound hot considering everything is running hotter on wall voltage?
After adjustment it actually tested at -46v on wall voltage?

Higher value (ie. more negative number) bias voltage gives cooler tube current. So -46 runs tubes hotter than -50. As wall voltage increases, so should bias voltage become more negative. So if you can only get -46V with higher wall voltage, I'd be looking for a problem in the bias circuit.
You will be able to see the bias voltage vs idle current effect when you have the cathode resistors installed.

[Helmholtz]

BIAS: Yes, I see the bias is marked as -50v but doesn't that sound hot considering everything is running hotter on wall voltage?
After adjustment it actually tested at -46v on wall voltage.

It's the other way round: More negative bias voltage makes the tubes run cooler.

Actually I doubt that the higher idle currents cause the mains fuse to blow.

When adding 1R cathode resistors make sure to measure voltage across the resistor and not wrt ground.

[Glenn63]

[Glenn63]

Ah, my error on negative bias voltage. Been a while.
Ramped up to 115vac I now see a sawtooth on filter A. Don't remember if I didn't look there before.
With tubes out, disconnected OT from A... sawtooth still there. Seems very suspect. Other filter taps are very smooth.

I adjusted the variac voltage until bias sat at -45v
Found tube 1 conducting at 71ma while tube 2 reads 0ma. (made sure the cathode resistor was good).
All voltages to sockets are identicle.
Bad new tube? Gonna test that theory.

Negative on tube swap. Perhaps I have a problem in tube 2 socket?

[Helmholtz]

Ah, my error on negative bias voltage. Been a while.
Ramped up to 115vac I now see a sawtooth on filter A. Don't remember if I didn't look there before.
With tubes out, disconnected OT from A... sawtooth still there. Seems very suspect. Other filter taps are very smooth.

Seems your first filter cap is bad - or not well connected.

Please explain your scope picture above.
Scopes are not only imaging/viewing devices but more importantly they are wonderful measuring instruments which allow for AC and and DC voltage, frequency, phase and time difference measurements (and some more). You just need to know how to read it.

So please always post scope settings with good scope pictures.

[Glenn63]

Thank you Helmholtz.
Scope set on 5v/Div. DC.
Nearly 20v!
Filter connection at board should be excellent.
Ground connection is at 0ohm.
I temporarily placed the original 30mf/600 back in and found the saw tooth slightly larger.
Could there possibly be a bad rectifier diode?

[Jazz P Bass]

It's not unusual to have Vac ripple on Node A.

[Helmholtz]

I read around 17V peak-to-peak of ripple. If this is with OT CT disconnected, it is certainly too high.
If you also post the time base setting I could tell if it is 60Hz or 120Hz.

If the zero idle current stays with the socket, it may be bad. Try to retension socket contacts and check connections. I guess you also checked screen voltages?

[Glenn63]

"If you also post the time base setting I could tell if it is 60Hz or 120Hz."

Time Base is at 2ms.
= 120hz.
Is there a problem with the rectification?
Yes, all voltages are identical. I will flip the chassis to examine the sockets after dinner.

JPB "It's not unusual to have Vac ripple on Node A" Now I hear Tom Jones.
I agree but this ripple is excessive and not smooth. Got to be something there.

Thanks guys.

[Helmholtz]

Power supply ripple increases with current (without any load ripple must be zero).
With OT CT disconnected, there should be only around 10mA for the preamp and PI left and ripple should be low.

[Glenn63]

I reported 'incorrectly' that T2 was not conducting as read at cathode R... I found that I still needed to cut a trace to ground.

Alright, after a week of going back and forth between 120vac and 115va to check every change I make.
I'm pretty convinced to buy a new PT from Fiptops and solve the mains voltage problems.
This power supply is running almost 80v hotter than it should be!
It's putting 532v on the B' filter of 47mf/500, and on the OT to plate connections, where plate voltage is around -25v from A', which doesn't seem right?
This is resulting in the screen being +5v above the plate.

I've replaced filter caps as follows:
A' 2x22mf/500 parallel
B' 47mf/500
F' 47mf/500
C' 47mf/500
E' 40mf/500 in can

Main, Plate, Bias and Current findings:
@120vac Asup= 538v
Plate= 513v
Screen= 518v
68k Bias= -63.8v and cathode current was 22ma. (Too cold) I changed the 68k to 56k
56k Bias= -53.8v and cathode current was 54.8ma.(Too hot)

Goal values I am trying to achieve would be -50v and 30-35ma.

@115vac Asup= 457v
Plate= 449v which is -8v from A'?
Screen= 454v
Bias= -45.6v and cathode current was 41ma. Which puts this amp close to goal values.


NO CHANGE TO LOUD HUM present from PI stage to output.
Pros... Refreshing on a lot of amplifier basics. Lol...

[Helmholtz]

I'm pretty convinced to buy a new PT from Fiptops and solve the mains voltage problems.

I don't think that will solve your problem. Your DC voltages are still within the 20% tolerance, often specified with Fender amps.

This power supply is running almost 80v hotter than it should be!

I don't see that from your data.

This is resulting in the screen being +5v above the plate.

No problem at idle.

68k Bias= -63.8v and cathode current was 22ma. (Too cold) I changed the 68k to 56k

Tacking a 680k across the 68k will result in 62k. 470k will give 59k.

Plate= 449v which is -8v from A'?

That's caused by the voltage drop across the DCR of the OT.

BTW, what is your heater voltage at actual mains voltage?

[Glenn63]

Thank you Helmhotz for explanations of my findings!
Heater voltage is 7.3.
GOOD NEWS!
I was thinking if bias is correct...
tube output is correct...
no dc voltage on coupling caps...
what about OT and NFG?
I found 3.5v on nfb and dummy load...
Upon further examination I found the OT ground wire had snapped at a board pad!
Problem solved!
I will try further bias adjustment with your recommendations.
Thank you for your time and patience.
Now it feels good to be back at the bench.

[Helmholtz]

Heater voltage is 7.3.

That's indeed huge (assuming all tubes were in).
It's around 16% above nominal. Anything above +10% is likely to reduce tube life. Are you sure your meter is accurate (battery fresh)?

I found 3.5v on nfb and dummy load..

Do you mean 3.5Vrms (hum) at speaker output?

Problem solved!

What exact problem?

[Glenn63]

Yes, by "problem solved" I meant the loud hum at speaker. Where 3.5v dc was found.
High voltages are another issue.
I would think voltage that high would be hard on many components. Definitely exceeding tube data specs..
Could it be possible to reduce the heater voltage via resistors to a center ground?