Sag phenomena?

[Danelectron]

I built an amp a while back that is loosely based on a gibsonette, but with 2 6L6s instead of 6V6s. It has a 5y3 rectifier tube and only 1-12ax7 preamp stage. I had it cranked wide open the other night and noticed a curious trait. When I would hit a note and let it ring it would seem to climb in volumn in a few little steps (like waves) as if it was going thermal runaway, but would then fade out gradually. Is this a result of sag? Depending on the note, the sustain would hang in there for quite some time.

At this volumn the amp has an awsome distortion. Not suitable for general playing, but fun for what it is. A bit muddy, but very gritty. Leads get a bit lost(though not impossible if the speed is reasonable), but for power chords or rhythm it really makes you want to elbow somethin'.

[Chuck H]

That's sag alright. It sounds like there's too much though. What is the mA rating on your PT HV secondry?

[Mandopicker]

I was always under the impression that 5y3's work better in lower wattage amps. (under 20 watts) With 6L6's you might try a 5ar4/GZ34 or even a solid state rectifier. That might help reduce the sag issue. That is...assuming the PT isn't too small for the current draw you are getting as a result of tube choice.

[guitarmike2107]

I was always under the impression that 5y3's work better in lower wattage amps. (under 20 watts) With 6L6's you might try a 5ar4/GZ34 or even a solid state rectifier. That might help reduce the sag issue. That is...assuming the PT isn't too small for the current draw you are getting as a result of tube choice.

That was my thought, the 5Y3 is not up for supplying two 6l6 tubes

[Steve Conner]

The overall description of the tone sounds like a bad case of blocking distortion though. The Gibsonette might well be one of these designs that was never intended to be pushed with humbuckers, and sounds like crap when you do.

[kg]

also, strange volume changes can be the result of parasitic oscillations somewhere else in the amp... though with only 1 ax7 it's not likely in your case.

[guitarmike2107]

Besides the 5y3 issue that amp also doesn't have huge filter caps, which won't be helping

Prowess Amplifiers - Gibson - Schematics - Gibsonette

Have you used the correct grid leak resistors too,, if they are too big you may have run away bias.....

just guessing.....

[pdf64]

What's the value of that 1st dropper resistor in the B+ supply, between the reservoir cap and plate node?
Parasitic oscillation with a single 12AX7? If you cascade the stages with no attenuation (eg vol up full) or grid stoppers, and get the lead dress bad enough, I can testify that oscillation will occur! The unbypassed 2nd cathode will tend to mitigate for that, though. The key thing is to keep sufficient physical seperation between the input circuit wiring / components and the 2nd plate circuit wiring / components
The root cause may be a combination of several effects mentioned. Blocking distortion can be mitigated by reducing the value of the 1st 0.022uF coupling cap to the vol control; 4n7 is plenty. Also consider adding a 470k to 220k resistor between vol control wiper and ground, so that at high vol settings, a little bass gets rolled off. Pete.
To reduce sag, the plate and screen node caps could be doubled in value

[Steve Conner]

Yes, I have got single 12AX7 amps to oscillate!

[Howlin' Mad Mac]

Yes, I have got single 12AX7 amps to oscillate!

Lol... Me too.

[Danelectron]

Sorry, I will get more info. It was redesigned by a friend with bigger power supply caps. When he first built it he was frying 6V6s on contact so used KT66s (I should have stated that instead of 6L6s) and redid some associated component values. It originally didn't have screen resistors and had some bad oscillation at full volumn. It was fine up to about 9/10ths, but full open produced a squeal. Screen resistors have been added. 470 0r 500 comes to mind. I don't remember all the values so I will open it up, and check the components, voltages and transformer specs and address all your comments as they apply. I had never heard this phenomena so had to ask. It is apparently greatly exaggerated from what I guess a good sag quality should be. It actually sounds pretty cool if the tempo is slow enough to let it happen. It sounds fine at lower volumns with the humbuckers.

Thanks for your comments!

[Chuck H]

Sounds like you have some high plate voltage.?. If yes then the amount of power the amp is trying to make is probably too much for the 5Y3 (as already noted) but I still wonder what the mA rating of the PT HV secondary is.?.

[Danelectron]

Can anyone explain the design of the power section of the Gibsonette (GA-8)?

Prowess Amplifiers - Gibson - Schematics - Gibsonette

Thanks for the schematic guitarmike2107.

Specifically the bias components. That is definately different than the redesign. It appears to be cathode biased, but what would be the purpose of the 470 and 220 grid resistors? Or why the different values? One thing I can add. My friend is a Gibsonette fanatic and said it is one of maybe a couple of Gibson amps that sound great. Specifically the GA-8. Not the tremolo version. He set about redesigning it for some channel options, but this one is just one channel of the final result. He did use a bigger PT than the Gibsonette. Details to follow.

[d95err]

Can anyone explain the design of the power section of the Gibsonette (GA-8)?

It does look a bit wierd yes. It is a single-ended amp with two parallell output tubes. It has cathode bias (the 200 ohm resistor). The 220k resistor is the ground reference for the output tube grids. Then the wierd thing that one output tube has a grid stopper (470 ohms) and the other does not.

I'm guessing:

1) Schematic error (both tubes are supposed to have the 470 ohm grid stopper)

2) They decided to "Muntz" the circuit design (skipping non-vital components to save costs) and removed one of the grid stoppers. The other one has remained in the circuit, because they needed something to connect the output tube grids.

If you intend to push the amp into distortion, I'd add the missing grid stopper resistor.

[Danelectron]

O.K I have a bunch of info. The only real commonality to the Gibsonette is the preamp stage which has an added tone control. Very simple, 1 meg pot after the first stage. The caps in the power supply are 47uf, 30uf, 20uf and 20uf. The resistors for the voltage divider are 470 5 watt, 10k 2 watt, 22K 1/2 watt providing (tested) 279V B+ to the power tubes, 247V to screens (pin4), power tube bias voltage is -17v at pin5. Preamp b+ is 145v. The power tube section grid resistors are 470K and the screen resistors are 500 5watt.
The power transformer is a Hammond 272JX 600v centertapped 250mA - 5v / 4A - 6.3v/8A. The output transformer is a Hammond 125FSE 20 watt 90mA. So that is undersized for one issue. Is it in imminent danger? It doesn't seem to get hot and I've been playing it pretty regularly.

I have a JTM 45 output transformer that I can use for the output but may look for something different as this one is slated for a Marshall build. The 5Y3 will be upsized. Any recommendations? I want to stay with the tube rectifier if the power supply is adequate.

d95err, Thanks for the response on the Gibsonette. That IS weird in my mind also.
Thanks for all your input, Dan