1955 Gibson GA-30

[boroman]

Hi guys. I recently got Gibson GA30 from 1955, quite unusal design not covered in any schematics over the web (well - I can't find it). Around 300 were made in erly 55 and then they switched to different preamp layout.
It makes very loud cracks, and almost any noise you can dream of... I managed to play some notes while it didn't cracked for a while, and sounded sweet, but now, just after it's started it's going wild.

Tube layout
12AY7
12AY7
12AX7
2 x 6V6
5Y3

1/ It's been worked on, since some supply caps are from 1980s - they had bubbles so I replaced them before I turned the amp on.
2/ Checked all tubes (replaced them for a different set)
3/ Cleaned pots, tube sockets
4/ Removed all three preamp tubes and it's still the same, so the prpoblem is in the power amp section
5/ That old 10uf cap across 6V6 sockets and big 250R resistor below are also good (replaced them for new ones but problem is still there, so I put back originals)

Does anone have drawn connections LAYOUT or detailed schermatics of those mid 50's GA30s? I found only very basic schematics and I am not so pro to "imagine" the rest. I'd better look at untouched one (or a drawn layout) and try to make all the connections the same as they should be.

[Jon Snell]

That looks like the GA-20 not 30.
The GA-30 has 6SJ7 pre amps and 6SC7 phase splitter.
The GA-20 has three 12AX7 with two 6V6 and 5Y3 rec.
GA20.pdf
ga30.pdf

[catalin gramada]

Lot of versions over here : https://youtu.be/sVyctsTpIK8

[boroman]

Thanks for the GA20 schematics.
MY amp is GA-30, with unique preamp - more like your GA20 schematic, more about the amp:


Before 1955 they were using pentodes in the pre-amp stages. 1955 was the first year they used triodes, and the design can only be described as bizarre. Dual 12AY7's in the pre-amp, with each of the four inputs (3 instrument and 1 mic) getting it's own triode pre-amp. One 12AX7 split as separate drivers for two 6V6's. A "tone expander" switch that bypasses a 3000pF cap in a degenerative feedback loop that, as far as I can tell, does absolutely nothing.

Some other oddities:

Usually, when you want to control the signal level going from one stage to the next you dump the signal into the top of a pot, and pick your signal for the next stage off the wiper - a traditional voltage divider that gives you a constant load on the output of the prior stage. This circuit actually dumps the output into the wiper, and picks the signal for the next stage off the top of the pot. This means the load on the preamp stage changes depending on where the volume is set.

The three triode sections used for the three instrument inputs have their plates all tied together to one plate resistor. Weird stuff happens when you try to use more than one instrument input at the same time - when the current through one triode increases then the voltage drops on all of the plates.

Also, the quiescent (no signal) plate voltage is lower on the instrument triodes than it is on the one mic triode because both sides use the same size plate resistor, but the instrument side has three tubes sucking current through the same resistor. Because of this, there is far more headroom through the mic input than there is through any of the instrument inputs. I've read comments from two other people who own this particular year and model that confirm that this is the case with their amps, as well, so I know mine isn't unique.

I drafted a schematic of this thing years ago, but I couldn't believe what I was seeing. I was sure somebody must have severely modded it until I found pics on the internet of another 1955 model that appeared to be identical inside, and comments from other owners who described the same anomalies I was hearing. I sent an email to Gibson asking if they had a schematic for this model. I sent them my schematic for comparison. They said they no longer had any records whatever from that year, but thanked me for providing a schematic because they'd been trying to find one for years.

It seems that 1955 was the only year they used this particular circuit design, and only about 365 were made that year.

[The Dude]

I'd be particularly wary of those white wax/paper caps. They are notorious for not aging well.

[boroman]

I'd be particularly wary of those white wax/paper caps. They are notorious for not aging well.

Are talking about everything, incuding those in the tone stack?
I already changed that bigger one on the lft side of the pic (between 6V6) but its the same so not this one...

[bmccowan]

Is this the one?

[bmccowan]

And a few things - Gibson played it pretty loose with circuits and schematics in those days - if it was in the parts bin it was fair game. I would not trust that the circuit in your amp is as intended - a lot of "repairs/alterations" have been made. And I would correct that AC input - Green to chassis bolt - white directly to one of the PT leads - Black to fuse > switch > the other PT lead. Tone expander just changes the frequency of the signal sent back into the circuit from the feedback loop - works like a bright cap. The change is subtle - only dogs can hear it.

[g1]

Is this the one?

Nothing showing here, can others see this image?

[boroman]

And a few things - Gibson played it pretty loose with circuits and schematics in those days - if it was in the parts bin it was fair game. I would not trust that the circuit in your amp is as intended - a lot of "repairs/alterations" have been made. And I would correct that AC input - Green to chassis bolt - white directly to one of the PT leads - Black to fuse > switch > the other PT lead. Tone expander just changes the frequency of the signal sent back into the circuit from the feedback loop - works like a bright cap. The change is subtle - only dogs can hear it.

Yes, I figured it out it doesn't change anything... But tone stack wasn't touched here. Only changes were made in the power amp / filter cap area and I would like to find the actual layout of those GA20/GA30s from fifties what goes where VISUALLY to see what's going on there now

[bmccowan]

I doubt you will find a layout for that amp. Lots of Gibson schematics out there, but very few layouts. BTW that tone expander does do something, or should if the cap is good. It allows high frequencies to pass around the feedback resistor. Since its a negative feedback loop, it will cancel out some of the highs. Some folks call it the icepick filter. Does your amp have a choke? As far as I have seen, there are only 2 different power supplies for the 2x6V6 amps or that era - one with a choke and one with only dropping resistors. And the layout for the power tubes is also repeated from model to model. The filter capacitors they used did not hold up well. Best to replace them with good quality caps of similar value - F&T is a great brand and they make some 2 section axial caps that fit well in a chassis of that type. Are you able to sketch out what you have?

[Helmholtz]

A series cap in the feedback path causes a lower frequency boost.
E. g. a 1nF in series with a 470k resistor means a corner frequency around 300Hz
Bypassing the cap equalizes the frequency response.

The boost effect will more or less disappear with power stage clipping.

[bmccowan]

You're right Helm. That cap is in series, not bypassing the resistor. I should have looked more carefully.

[boroman]

Okay - last year I created this topic, I tried to think about it and brought it back to life today.
Sounds great BUT

Plate voltage across caps on the power supply is approx 288V with 250 ohm cathode resistor and 10uf cap (changed for 15uf, havent got anything nearer). So it's like 65% dissipation. I should be close to 100% here, so question is - change the resistor or that 5uf difference in the bias cap across cathode resistor is the problem?

[Helmholtz]

Cap value doesn't matter as a cap doesn't pass DC (except if it's leaky).

What does the cathode resistor measure?
What is cathode voltage?