Need help with Gibson GA-5T

[vortex]

Hi all,
WHAT'S UP:I just bought a Gibson GA-5T that needs a bit of TLC. It has had a "cap job" and three prong power cord added but I am not sure if it was done competently so I have a couple of questions about that. It also seems to have very low output so I hope to get a bit of direction on what may be the culprit.

THE AMP: Circa. 1965 GA-5T, 6EU7, 6CA, 2xEL84. Per factory schem this little guy puts out 12.5 watts. Right now I think it just might be putting out 1 watt or less. It doesn't sound bad, nice trem, nice tubey chimey highs and decent bass. A little bit of hum but not too bad on that either.

MY EXPERIENCE: I have been building pedals for 7 years or so and am getting into amps. I can work on an amp safely but have limited knowledge on what really makes them tick.

FACTORY SCHEM WITH VOLTAGE READINGS:
http://www.informatik.uni-bremen.de/...back_qual6.jpg


I've spent time making a LAYOUT WITH THE ACTUAL VOLTAGE READINGS to compare...

POWER SUPPLY:



TUBES AND BOARD:


POWER SUPPLY QUESTIONS:
1) The two rectifier diodes have been replaced with 1N4001's. Can they handle this much voltage? If not, what are suitable replacements?

2) There are 4 -20uF filter caps in the schematic but only three were used in the "cap job". The 20uF cap for the preamp stage (C13A)was omitted. Is this an OK way do do things or was the cap job a hack job? BTW the work looks cleanly done, but I noticed the "death cap" was still in place after the grounded power cord was installed?

3)The resistors after the B+ have been modified (1K down to 22 ohms and 3.3K down to 330 ohms) is that kosher?

TUBE AND CIRCUIT VOLTAGES:
How do these look? (See Tubes and board image) To my untrained eye the specified voltages seem to be pretty close to the factory schem. I measured what I could, where I could, with as much precision as my digital meter allowed. I am sure I must have some redundant data there but I don't know how to filter this information, so I included it all. I was surprised to see so many AC readings in the circuit but mostly everything I have built is DC. If there is something I have missed let me know and I will measure it.

WHAT I HAVE DONE: I plugged the GA-5T into a 1x12 -8ohm cab and it did provide a bit more volume and bottom end, I think that the original speaker is in OK shape. I just ordered some new tubes and a 20UF 500V cap. I tried substituting the EL84's with some spares I had kicking around and the amp sounded the same. I don't have these preamp tubes on hand to substitute... yet. I wonder if the volume problem isn't with the original OT wearing out, but I thought if an expert here spotted something in the data I have gathered I would follow that up first. What would be a good replacement OT in that event?

Well, I think that's it (probably not...). Sorry for so many questions! I feel like a "gopher", able to collect the data but not able to apply it. I really would love to know how an amp tech would approach this situation. Any insights appreciated.
Many thanks in advance!

[Chuck H]

O.K. Thats a pretty old amp. So who knows when that cap job was done. That amp could easily have been capped twice in it's lifetime. So if that's cap job number one, it may be due again. But I don't think the caps are to blame for the low volume issue. Not directly anyhow. Perhaps the guy that rewired that amp has made other mistakes as well.

Your schematic image is impossible to read and your interpetation drawings are unclear. If the amp were wired as the drawings suggest it would not be working at all. So be sure you understand the schematic, at least well enough to duplicate it, before doing any work to this amp. Also, be sure to drain the caps before you start.

"The two rectifier diodes have been replaced with 1N4001's. Can they handle this much voltage? If not, what are suitable replacements?"

No. If those diodes ARE 1n4001s they should have failed long ago. Replace them with 1n4007s

"The resistors after the B+ have been modified (1K down to 22 ohms and 3.3K down to 330 ohms) is that kosher?"

No. Some hairbrained attemt to raise the preamp voltages or some guy working with an analog meter didn't know how to read it.

"There are 4 -20uF filter caps in the schematic but only three were used in the "cap job". The 20uF cap for the preamp stage (C13A)was omitted. Is this an OK way do do things or was the cap job a hack job?"

I can't be sure what was going on when this "work" was done. Hack job doesn't begin to explain some of what your finding. You should rewire the amp to match the schematic. Check the entire amp. Anything you find that is not stock should be changed to match the schematic. Only because, 'who knows whats going on in there now?' The only exception to the schem should be the three prong power cord. If you want to mod the amp after that, you can get good advice here.

"but I noticed the "death cap" was still in place after the grounded power cord was installed?"

Clip it out. It can cause problems if it shorts.

Chuck

[vortex]

Thanks very much for the recommendations Chuck.

Your schematic image is impossible to read and your interpetation drawings are unclear.
I changed the image links, it seems that Imageshack isn't reliable, sorry about that. Things should be easier to read now.

If the amp were wired as the drawings suggest it would not be working at all. So be sure you understand the schematic, at least well enough to duplicate it, before doing any work to this amp. Also, be sure to drain the caps before you start.[/I]
I have gone through the schematic many times and made these layout files. The amp does work, and the wiring matches the schematic with the exceptions noted. Is there something you see in the layouts that looks way off? Or is there a better way I could represent my findings?. (note: i did skip the off board wiring).

Also, be sure to drain the caps before you start.
Yes, I certainly do, thanks for mentioning.

No. If those diodes ARE 1n4001s they should have failed long ago. Replace them with 1n4007s
I rechecked the rectifier diodes and indeed they are 1N4007. They were very difficult to read in circuit.

I will get the parts to return the amp to stock. I just needed to know that whoever had done the resistor changes and cap omissions wasn't operating with some "greater wisdom".

Death Cap has been clipped out.

Thanks again for the guidance.

[Chuck H]

Woah...I can read it now. Ha Ha

Looks like a nifty little amp. Unfortunately I don't think the volume issue is because of the cap job shinanigans (how do you spell that?). That should just cause noisy operation the way it was wired. So you may have a bad or loose componant/s or some other wiring mishap that you overlooked.

If you have a signal generator (or anything that can make a consistent signal) to put into the input of the amp, you can take AC voltage readings along the signal chain and hunt down the problem pretty fast.

Chuck

[vortex]

I don't have a signal generator but the my old Micromoog on "drone" fitted the bill!
Here's my readings along the signal chain. Let me know how it looks and if you need more info.
Thanks again!

[rf7]

Those readings at the speaker make me think it's a bad output transformer or speaker. Have you looked into that? Have you tried a different speaker? Go to www.geofex.com and look at the tube amp debug section for info on how to check out the OT.

The place where it goes down to 0.6mV is questionable too. It seems a bit low to me there, although you're certainly getting some signal into the OT.

[Chuck H]

Wow...That was fast.

O.K, I see what could be a couple of problem areas. There are real geeky formulas for figuring the amplification factor of a gain stage but I don't know them. With that preface, I'll continue.

You should probably have at least 5.5 volts downstream (signal wise) of c3. Could just be a weak tube, or a problem further downstream is robbing signal. I don't think the problem is here. So taking the voltage we do have as a staring point, I'll move downstream.

Your dropping alot of voltage through the bass control circuit, but thats normal. But your dropping way too much past that 470k resistor (r8). Something is bleeding signal to ground downstream of that resistor. Check r10 (250k) to be sure it's the correct value. You may aslo want to check c6 to be sure it's not shorted. less than 1millivolt is a truly tiny amount. Your guitar puts out generously more. So why the extra gain stages??? Right?

Moving downstream.

At the grid of the 6c4 you have .6mv. But at the plate you have 326.5mv. Thats a voltage gain of over 500, which is impossible. So something went wrong here. Either you had your meter set funny, read it wrong or some other AC is getting into the chain here...But, let's take that for what it is and move downstream.

Going into t3 primary you have 326.5mv. But at the secondary you only have 17.5mv. I'll admit that I've never worked with an interstage coupling transformer as a phase inverter (though I've wanted to try building with one) but this seems like alot of loss to me. I would have expected it to be just shy of unity gain. At the end of the chain so far you would normally want to be driving the grids of those el84s with about 8 to 12 volts at full tilt.

Last thing. You have indicated .1mv at the + speaker lead and 32mv at the - which is gounded. In a normal world there is no possible way that can happen. Ground as it is indicated should be 0 volts AC or DC. And .1mv of drive at the speaker should be barely audible.

You may have missed something that has been changed inside the amp. If it were mine I would start by fixing the power supply and B+ rail. When all my DC was as it should be I'd move onto the signal path.

Chuck

[vortex]

Chuck: Wow...That was fast.
You are pretty fast yourself! For me to be able to get this kind of info in a relatively remote part of Canada (Whitehorse, YT) is quite amazing. I owe you dinner at least...

rf7:
I've tried the amp with a very efficient 8 ohm speaker and the results were about what you would expect, a bit more power but not a mind blowing leap in performance. I think that I have ruled out the speaker as a problem.

I'll check out Geo for the OT testing. RG Keen deserves a medal! I can't believe how helpful he has been to me and everyone else in the pedal world.

I have an old Tweed Deluxe that I have cherished for years and the OT was wearing out through many of those years. It still is the best sounding amp I've heard even though it was only putting out about 2 watts near the end. I had the feeling that this might be the problem with this Gibson GA-5T amp, it still sounds pretty good but it doesn't have nearly the amount of poop it should. I had my Deluxe repaired by Dave Vidal in Vancouver, BC and it is still my main amp to this day. I can't say enough good things about Dave, we've never met but I can say that he is a first class act all the way.

Ok, so there's lots of things for me to check out. I have ordered all the necessary parts and they should be here for the weekend. In the meantime I will scope out what I can.
Thanks again for everything!

[Chuck H]

FWIW I don't think it's the OT. There are some funny readings there, but not the kind of thing I'd expect if the OT was going. Lots of stuff to check first. Heres a neat test technique that has worked well for me. Once you get the power supply dealt with you should try it. Use you "signal generator" (moog) set to "drone" and measure the output AC voltge with your dmm. Set it for .5 or 1 volt if you can just to make math easier. Now, starting at the grid of the 6c4, inject the signal there. Probably going to surprise you how loud it is. Then move back to the signal path input of the 6eu7. I think this will show the problem is somewhere in front of the 6eu7 plate and behind the 6c4 grid. But thats just a guess. You can inject signal anywhere that doesn't have B+ in the circuit (meaning, in front of decoupling caps). For example, you should not try to inject signal at the t3 (coupling tranny) input. There is B+ there. But you can do it at any point in the signal path that is downstream of a decoupling cap up to the grid it is feeding. This is a fasr way to locate some problems.

Chuck

[Enzo]

If your readings are correct, it looks to me like R8 is open.

[Chuck H]

OOOh. Good call. I didn't consider it because I've never seen a resistor go open without being burned to a visible crisp. But I know it happens. Could be a loose solder joint or a lead short too.

Chuck

[Enzo]

Those PI plate resistors in Fender Hot Rod Devilles and similar open often enough without appearances.

[vortex]

Hey, a small success today!

I found a splash of solder under the Volume pot (R10) lugs that was connecting pins 2 & 3. I removed the solder splash, tested the pot and put her back in. I tested the amp and it has a bit more volume but nothing near 12 watts. So that's helped a bit. I did the signal generator tests and am now getting much more AC after R8 (470K), unfortunately the readings further downstream didn't change. In fact when I tested the plate and grid of the 6C4 I couldn't get a reading. I still am getting AC readings on the speaker leads.

"You may also want to check C6 to be sure it's not shorted."
Checked it: seems OK

OT Tests:
OK, so the plot thickens:
I started going over the offboard wiring and checking the OT. The first thing that I discovered is that the OT model # doesn't match the schematic. I missed that. The schematic specifies a TF-18-01 but the OT in this amp is TF 504-0. It actually appears to be stock, but I only say that because it is riveted to the chassis and there is no evidence that it had been modified when installed.


I performed the "simple tests" for OT's outlined at Geofex.com.I pulled the tubes but I didn't pull the rectifier tube as there isn't one, so this might skew the results.

The measurements from the B+ to the primary leads (output tube plates) were encouraging (137 & 153 ohms). I checked the resistance from the chassis to the leads and things seemed weird. My digital multi meter was having a hard time getting a reading, (this has never been a problem before). The readings from the chassis to the B+ and primary leads (all around 8M) would read for a few seconds and then would not provide a reading. The speaker leads showed 1.5 and 1.8 ohms. So either the OT is toast, the tranny is a poor match or something else is creating a problem? I pulled the tubes but I didn't pull the rectifier tube as there isn't one, so this might skew the results. Should I have disconnected somewhere else?

[vortex]

If your readings are correct, it looks to me like R8 is open.

Hi Enzo, I just checked R8 "in circuit" and got a 500Kish reading.

[Enzo]

Well, the shorted R10 would have the same effect: killing the signal at the righthand end of R8.

Your transformer is probably fine. Your meter is reading the B+ caps. Resistance to ground from the OT promary leads has the B+ caps across it. The little voltage from your meter is trying to charge those caps. So what usually happens is at first you get some resistance, and then it goes higher and hiogher until it shows open. You could disconnect the center tap wire from the B+ to check for fframe shorts. If the frame were shorted to the windings, it would still measure short on the meter. I suspect it is fine. COnnect your meter to measure resistance from the transdformer leads to ground as you already did, and leave them connected. Your resistance would rise up and then no reading. Leave it that way. now with a wire, short from B+ to ground a moment. This will discharge the caps. When you pull the wire away, watch the reading ramp up from something to ever higher and then open. That is what you are seeing.

This is a Gibson amp, they usually do not match the schematics, I am surprised when they do even. I am sure the transformer is suited to the amp.


If you get no signal at the 6C4, it is pointless to look for any output, if it is not there, it will not be anywhere after there either.

YOu now have signal at R8? Fine. Is there signal now at the top of R10? Turn R10 all the way up. Is there now signal on the wiper? Better be the same on the wiper all the way up as there is on the right end of R8. And since the wiper is wired direct to the grid of C4, there should be the same signal there. R10 could be faulty.

The 6C4 lights up, but are the DC voltages OK? Is there 8 or 9 volts ( or 7 or 12) on the cathode there? If there is zero, then the tube is not working.

If the tube is indeed working, then you need to get signal to its grid. In fact, just touching its grid with a screwdriver should inject some hum into the amp and you would hear it coming out the speaker. If the power amp is working. Monitor signal at the plate and touch the grid with a screwdriver. Can you measure an increase in signal at the plate then? You should.

You show 300mv at the plate of the 6C4 - that could be signal or ripple. Measure for signal at the other end of that transformer - the end that goes to C13B. There would be no signal there, so any "signal" you see would be ripple.

Turn the intensity control to zero. Now with power off, measure resistance to ground from each power tube grid. Is there continuity? We are now checking for the phase splitter transformer secondary. I don't know what resistance to find there, but it should not be open.