Gibson GA-35 RVT Refurb/Mod

[jaymz168]

UPDATE: I'm putting this at the top for others who are looking to repair these amps.  THERE ARE TWO VERSIONS OF THIS AMP.  And the most common schematic you'll see around is for the least common revision.  There are schematics for a "12-65" and a "12-66" and the 12-65 was made in larger numbers but most schematics you'll find are for the more rare 12-66 version.  They have different tubes in V4, different tone stacks, etc.  If your Lancer is super shrill and you have a 12AX7 in V4 then you may just have a 12-65 that is supposed to have a 12AU7 there.  I figured this out because once I got it actually making sound again I went through it and thought the last guy must have changed a lot because the tone stack was completely different, etc. There were new caps, earthed line cord, etc. so not a stretch.  In prep for sorting it out I found a pic of someone who fully refurbed one of these but it was for a studio with a site that didn't exist anymore but the images were still cached by google.  So I went to the Wayback Machine and they had cached the page: https://web.archive.org/web/20180824034803/Infernalracketstudios.com/Lancer.htmThere you will find a bunch of info on the 12-65 Lancer including a SPICE sim and theory of the bridged T bass control. 
So anyway, if you have one of these it's probably the 12-65 version and you probably have the wrong schematic, I've attached the correct one.

Hello all and welcome to my first thread!  I help mod another community and I'll try my best to be respectful of the forum here.
Luck has recently smiled upon me and I've found a Gibson GA-35 RVT in my home, likely left by a previous tenant (it's a place full of musicians, etc).  Since I had heard it didn't power up I checked the fuses and lo and behold, one of them was blown.  After replacing the fuse with a suitable working unit the amp powers up and howls like a cat with it's tail in an outlet.  I've looked around a bit and discovered that this thing uses some odd power tubes (7591) and is pretty shrill and piercing when working and in a stock config.  I've found some examples of mods, such as removing the tone stacks and biasing the tubes hotter, which seem to be worthwhile.

The twist is that I've never worked on a tube amp before.  I've done small signal stuff like solid state mic pres and compressors but never a high voltage tube circuit like this (though I've got an LA-2A board I'm trying to get around to).

Here are some pictures of the interior with notes and here's a link to a larger album I uploaded on Imgur where you can view or download larger pics : http://imgur.com/a/uaJlv

Overview



A quick overview shot of the amp head

The Main Board



Here's the main board.  Some of the caps sorta look newer to me but I don't know enough about vintage gear to be sure.

Filter Caps



Here are the big old twist lock caps, one wrapped in electrical tape for some reason.  They're rated 4 x 10uF @450V but the schematic (if I'm reading it correctly) shows 20uF.

The Power Supply



And here's the power supply section.  This has had a grounded plug added.  I'm not sure how that affects the use of the four position switch (two on positions for mains polarity).


Here's a link to a larger album I uploaded on Imgur : http://imgur.com/a/uaJlv

So, advice on how to proceed?  My plan right now is to test what tubes I can and check the filter caps and transformers.  If that gets her running then I'll move on to modding, but if it's still not working what should my next plan be?

[tubenit]

To help with the discussion:

link to schematic also:  http://www.gibson.com/Files/schematics/GA-35RVT.jpg

[jaymz168]

Thanks tubenit, I meant to link that schematic but forgot :D

[sluckey]

That's about the cleanest Gibson layout I've ever seen! You may be interested in another current thread on this forum about this same amp.

[eleventeen]

From the looks of things, there's been extensive work done inside that amp. Normally we would not expect a manufacturer to use a mixture of orange drop caps and yellow ones. The parts board looks odd, in that we can see purple residue on a chassis hole further down the line, with kep-nut marks on it, yet with nothing in it. And obviously those cap cans aren't stock.

So you have a dual conundrum, 1 part being, you haven't worked on tube stuff. As long as you can use your soldering iron and remember to keep one hand in your pocket at all times when working on it, that's not a big deal. (but it CAN kill you, LOL) The second conundrum is whether you want to put in the work to discern what has been modded and whether restoring the thing to stock is what you want to do. That could be a lot of work. EVEN IF you determine that the thing is wired stock and all the parts are good, there is some possibility that if this is a different parts board, that the thing cannot work very quietly because the parts are jammed together. Someone here, I think it was Platefire, just had a GA-35RVT and I hope they chime in as to whether the innards of theirs, specifically the parts board, matches what you've shown. Get a pix.

OTOH, looks like you have virtually every part you need to make a 1-channel Fender Vibrolux Reverb, a very useful amp, IMO. You could buy a blank AB763 board from Doug here, populate it with parts snipped from your existing parts board with very, very few additional parts needed, and have a pretty cool and known-stable amp. You'd have to invest about $100-$150 for the tubes (change 6EU7 to 12AX7s, retain your 7591s) and new electrolytics (odds are you should replace those anyway) and less than $5 worth of R's and C's. You have all the switches and jacks and sockets (those can really add up)

That's how I would lean. OTOH, if that parts board is stock and you find out that the mods are/were minor it may not be that big a deal to get it working. Further eval needed.

[jaymz168]

Thanks for the interest guys!  The hilarious part is that I moved in and this thing was sitting by the front door with it's front to the wall.  At first glance at the back all you really see is "Lancer" and we thought it was some no name combo amp.  Imagine my surprise when I decided to take a look at her!

Looking at this thread (http://www.el34world.com/Forum/index.php?topic=9230.0) it looks like user Plexi50 had one of these.  The board is similar, though some components are clearly different.  This thread (http://www.el34world.com/Forum/index.php?topic=13356.0) also shows a similar board and a better close up of the small part, though someone mentions that they think that one had Fender mods done to it... it seems like it's going to be interesting working this out.

BTW, got some pics of the tubes that were in there





The power tubes don't look terribly happy to me, what do you guys think?

[jaymz168]

That's about the cleanest Gibson layout I've ever seen! You may be interested in another current thread on this forum about this same amp.

Thanks, I just found it!  I did a few searches but it didn't show up, but there it is on the front page of the forum.

[HotBluePlates]

BTW, got some pics of the tubes that were in there
...
The power tubes don't look terribly happy to me, what do you guys think?

The 12AX7 marked "The Fisher" and "Made in Gt Britain" is almost certainly a Mullard ECC83; the 2nd line of the code (even with "Made in") looks like it might start with "B" which would indicate Mullard's Blackburn plant.

The power tubes look fine to me. The getter flashes have some browning around the edges, but that may not be a reliable indicator of the tube's condition (at most, it means the getter has been doing its job of combining with any free gas molecules in the envelope). That the getter flash looks mostly black is also not necessarily bad... Some Philips small-signal tubes had a specially-developed getter that appeared mostly-black instead of silvery. My best info is that much of the Japanese tube industry was set up by Philips-owned companies, explaining some features these tubes have in common with Philips-made tubes (Philips being the parent of Mullard, Siemens, Amperex and numerous other companies).

From the looks of things, there's been extensive work done inside that amp. Normally we would not expect a manufacturer to use a mixture of orange drop caps and yellow ones. ...

I see one green cap that looks stock; every other cap in the amp appears to have been replaced. Resistors could be stock, as they are the right type and age.

Agreed that the first challenge should not be powered tests, but a close inspection to see that the existing amp conforms to the original schematic (i.e., the previous "repairman" didn't do something goofy).

I wonder if a novice tech tried to "restore/upgrade" this amp and wound up digging himself into a bigger hole.

[jaymz168]

I wonder if a novice tech tried to "restore/upgrade" this amp and wound up digging himself into a bigger hole.
Posted on: December 29, 2013, 01:59:15 pm Posted by: jaymz168

Yes, this is what I'm trying to avoid by coming here!  By looking at Platefire's thread it looks like quite a bit of the main board is stock.  My main concern after taking a closer look (haven't metered anything yet) is with the added grounded mains plug.  The ground lead seems to go to a lead on the power supply instead of to the chassis.  I'm not going crazy right?  The safety ground should go to the chassis!

Anyway, my next move is to take everything out of the chassis and start sorting things out.

[sluckey]

Looks like it's going to chassis to me.

[PRR]

> The ground lead seems to go to a lead on the power supply instead of to the chassis.

It goes to a grounded leg-lug on a terminal strip. See pic.

It is sorta well grounded. Plenty good enough for the old days.

Personally (and in modern electric practice), I would not trust a single rivet to a steel chassis after 50 years of oxidation.

Also modern code likes a *dedicated* ground point ONLY for the power-cord ground wire.

Get a toothed lug, and a good screw and nut. (Our host sells such stuff, excellent quality.) Scrub an area to loosen tarnish. (Since it hasn't outright rusted, it is probably plated, and you probably don't want to scrub through the plating.) Drill, deburr, best-quality solder-join green wire to tooth lug, then screw until happy.

[jaymz168]

Thanks, I see it now.  I was too focused on the nut/washer near there that I didn't notice the ends lugs of the turret board are grounded.  I'll check back once I've gotten her sorted.

[jaymz168]

Hi everyone, I'm resurrecting this six and a half year old thread.  A lot of life has happened but I've managed to get some projects done in that which have brought me up to speed to be able to tackle this.  I have a much better bench situation now and a lot more knowledge of safety and circuits.

Currently it fires up without smoke or destruction but makes no sound until the input is turned up about halfway and then it comes up very quickly and distorted.  What I've gone through so far:

• Brought it up slowly on a variac and limiter in standby with no tubes and measure the heater lines which are ~6.4V with no load
• Put it into 'on' and all voltages check out, filter caps are good
• Inserted tubes and brought to standby.  I can only directly see filaments lit on the last 12AU7.  I can sort of see filaments lit on the first 6EU7 and the 7591s, but only through reflections in the glass.  Other tubes do not seem to be lit.  Bias voltage on power tubes seems good, haven't checked current.  Filament voltage with tubes inserted seems to be down to ~4.4V.  B+ is present on OT.
• Fiddled with knobs and there's a little scratchiness but nothing out of the ordinary.  It is audible from a throwaway speaker.  Also some chopstick poking with no effect.
• Fed tone from a Qbox to input 1 and followed signal through the preamp stages.  I see DC plate voltage on V1A, and gain to the AC signal which continues fine all the way to V4A (the next stage) and then I've got no gain at that tube.

So is the next step to just buy some new tubes and see if that sorts it out?  I have some spare 12AX7 but no 6EU7 unfortunately.
Bonus picture of part of the "bench" I'm working on now:

[sluckey]

Brought it up slowly on a variac and limiter

These have served their purpose. Now plug the amp straight into a convenience outlet, repeat voltage checks, and proceed with troubleshooting.

[jaymz168]

Brought it up slowly on a variac and limiter

These have served their purpose. Now plug the amp straight into a convenience outlet, repeat voltage checks, and proceed with troubleshooting.

Oh man, I totally missed that I still have it plugged into the limiter.  Sometimes you just need someone to look with new eyes, thanks!  I'll see how it goes without the limiter and report back.

[Joel in Texas]

CRITICAL INFO: Now that you're putting tubes into the amp to do testing, you will want to connect a load (speaker or other appropriate load) to the output.  You started this thread as a self-described beginner with tube amp repair.  Some time has elapsed, and maybe you know more now.  But I just wanted to make sure you knew this critical point.  You really don't want a signal hitting the amp, when no speaker (or other appropriate load) is connected - especially when the power tubes are in the circuit.  That can seriously damage the amp. A good rule: if tubes are in, speaker must be connected.

[jaymz168]

Thanks for the reminder!  I have a couple 100W 8R resistors from Parts Express I can use as dummies and I also have a couple old 8R woofers laying around so I don't inadvertently kill the CTS that's in the cabinet.  I've been running it into a Radio Shack woofer I don't mind killing.

Here's a link to an album of the hifi amp I built in the meantime: https://imgur.com/a/5aAQL That was my first shot at SMD, 2x150W@8R solid state with a microcontroller running Arduino that monitors temperature, DC offset, current, etc.  Drag soldering that little puppy was a lot easier than I expected!  Not my project, someone else designed it, but I had to decide on an output level and design the supply accordingly.  Also had to drill and tap those heatsinks!  There was no guidance on how to handle ground so I tried a couple schemes and now it's so quiet you can't even tell it's on with no signal through it.

Anyway, it was getting late last night and since the amp was screaming feedback the last time it was turned on I decided to put it to rest for the night.  I'll be following up later today.

[jaymz168]

OK I fired it up the other night without a limiter or anything in line and it seems to come up fine and I don't see any red plating or anything visually awry.  Again running signal from a Qbox into the first channel and turning up the volume resulted in no output, just a little rustle near the mid point.  I have some time to work on it this weekend so I'll go through a bunch of points and do some measurements starting at the diodes.  Also the output jack wiring actually looks wrong to me.  I understand how the shorting jack works but it seems like it's shorting even with a plug inserted.  Who knows, maybe the last guy in here did that and just resigned to using the External Speaker jack.  I'll have to double check that as well.

Re: output tube bias, I see a lot of people talking about bias probes, soldering in 1R resistors and measuring voltage drop and calculating, etc. but is there any reason I can't just put my Fluke 117 across the plate feed in current mode?

[Dave]

Just food for thought... I came upon one of these one time. It was a basket case. After several hours of fruitless troubleshooting, I ripped it out completely and rebuilt it as a Deluxe Reverb. It required all new transformers, sure, but it turned out to be a really nice amp after all was done. Wound up selling the original transformers for more than I paid for the amp.


If you look at the layout of controls and what the amp was trying to do, it seems obvious, to me, that this was a direct attempt to compete with the Fender amps of the time.


The amp, ever since, has been one of my go to amps.


Dave

[sluckey]

Also the output jack wiring actually looks wrong to me.

It is! Cut the jumper as shown in the attached pic.

Re: output tube bias ..... is there any reason I can't just put my Fluke 117 across the plate feed in current mode?

Yes. To use a meter in current mode you must break the circuit and connect the meter in SERIES. In order to measure the current for each output tube you would disconnect the OT plate lead from the tube socket and connect the black meter probe to the socket and the red probe to the dangling PT lead. Repeat for the other tube. I suggest you don't do this.

There is a method that doesn't involve disconnecting the OT leads. First, with power off, measure the resistance between tube socket plate lug and the primary center tap of the OT. Do this for both tubes. Now turn power on and measure the voltage drop between tube socket plate lug and OT primary center tap. Do this for each tube. Now calculate plate current for each tube by dividing the voltage measured by the resistance measured. Once you have calculated the actual plate current, you can measure the plate voltage and finally calculate the actual plate dissipation.

If this seems a bit dicey to you, well, it is. I highly recommend you reconsider the 1Ω cathode resistor method. Or, since this amp has no bias pot, just insure there is approx. -20.5V on the grids of the output tube as shown on the schematic ***AND*** the tubes are not red plating. And call it good. Setting bias is way over-rated IMO.

[jaymz168]

Success!  Sort of!  Whoever was in here before ended up with shortened OT leads and so they just soldered it to the EXT jack first and then jumpered to the main one which is further away and then wired up the main one wrong.  Lovely!  As a quick test I just clipped that jack out and used the EXT jack and it worked.  The first input of the first channel seems weak like it's leakage, which is also a switching jack and the second jack works fine.  I haven't dug into it yet but I'm guessing they wired up the other switching jacks wrong.  But I've got sound coming out so that's a good start.

Next I think I need to:

• Double check all of the coupling caps for DC leakage onto next stage.  It still seems a little weak for the wattage but I'm also currently running it into a Radio Shack woofer, not a high sensitivity guitar speaker
• Double check the wiring on that input stage
• Remove that damn death cap soldered into the power switch
• Start testing on the RVT section
• Redo the output jack situation
• Maybe redo some of the poor wiring that seems to be the work of the last guy
• Pots are a little scratchy so either clean or replace

If you have any more suggestions in there let me know please!  Thanks so much for your help so far!

[jaymz168]

Well taking care of the output wiring sorted it all out.  Everything works including reverb and tremolo, pretty neat watching that 0A2 cycle with the tremolo rate.  I'm not seeing any DC leakage from stage to stage and no DC on the pots so I think they just need to be cleaned.  The only thing I'm worried about right now is the output tubes.  They get super hot super fast like they're biased into Class A or something and the thing turns into a space heater, moreso than I expect from a tube amp.  I *think* one tube is getting it worse than the other and the output transformer has some discoloration on the side near that tube. So I've gone ahead and ordered some Tung-Sol 7591A from AES along with some parts to mess with the tone stack and put a bias pot inside per the Uncle Doug videos.

As far as biasing goes, I see a lot of info out there for cathode biased output stages but these seem to be grid biased if I'm following this correctly?  Is the procedure any different for measuring and biasing this topology or do I just substitute grid for cathode in all of the measurements and calculations?  I've been playing with the calculator at robrobinette.com but it's giving me wacky numbers, I suppose since cathode is at ground on this amp.

*Oh also I got a 0.5W metal film resistor for that bias circuit, that should be a high enough rating for that, right?

[sluckey]

or do I just substitute grid for cathode in all of the measurements and calculations?

Although the grid voltage is what actually controls the bias, the grid voltage is NOT used in any dissipation calculations.
read this...   http://www.duncanamps.com/technical/lvbias.html

*Oh also I got a 0.5W metal film resistor for that bias circuit, that should be a high enough rating for that, right?

There are 3 resistors in that bias circuit. Which one are you talking about?

BTW, for comparison, I have a fixed bias 7591 amp with 470V on the plates and -25.2V on the grid. It's safely biased.

[jaymz168]

or do I just substitute grid for cathode in all of the measurements and calculations?

Although the grid voltage is what actually controls the bias, the grid voltage is NOT used in any dissipation calculations.
read this...   http://www.duncanamps.com/technical/lvbias.html

OK so I still just put a 1ohm resistor in series with the cathode, etc.  I managed to forget to order 1R resistors for this, and I have some 8ohm non-inductive dummy loads, any reason I can't use one of those and just change the resistance variable?

*Oh also I got a 0.5W metal film resistor for that bias circuit, that should be a high enough rating for that, right?

There are 3 resistors in that bias circuit. Which one are you talking about?

BTW, for comparison, I have a fixed bias 7591 amp with 470V on the plates and -25.2V on the grid. It's safely biased.

Following the Doug mod, he swaps out R40 for a 39K and a 10K pot.  It looks like the high power resistors are marked as 1W, etc. on the schematic and these (R40, R41, R43)  are not so I assume an 0.5W should be fine.  I think once I clean the pots and get new output tubes in there it's going to run great but I still might give the variable fixed-bias a shot.

[sluckey]

OK so I still just put a 1ohm resistor in series with the cathode, etc.  I managed to forget to order 1R resistors for this, and I have some 8ohm non-inductive dummy loads, any reason I can't use one of those and just change the resistance variable?

8Ω resistors would be fine but you must do the extra step to calculate cathode current. The advantage of using 1Ω resistors is doing the math in your head, ie 25mV across 1Ω is 25mA. Very easy.

I would not waste an 8Ω dummy load resistor for this. A 1/2W resistor is sufficient.

[jaymz168]

Ok so just to be certain before I do this:  I solder a 1R resistor to pin 5 of the 7591s to chassis (or just in line with their existing return?) and measure voltage drop across that.  Your link says a lot about pin 8 being cathode but my cathode is pin 5.


James

[sluckey]

I solder a 1R resistor to pin 5 of the 7591s to chassis (or just in line with their existing return?)

Your existing return is chassis. Right?

Your link says a lot about pin 8 being cathode but my cathode is pin 5.

OK

[jaymz168]

I solder a 1R resistor to pin 5 of the 7591s to chassis (or just in line with their existing return?)

Your existing return is chassis. Right?

It looks like a wire comes from the secondary center taps at the terminal strip, heads to Pin 5 on the first tube, heads to pin 5 on the second tube then goes to the PCB.  From that point the capacitor cans are tied and then they also go back to the same point on the terminal strip with the two center taps.  But it's never actually bonded to the chassis except maybe incidentally through the capacitor cans being in contact with the chassis.

[jaymz168]

I went ahead and did the 1R resistor method and ended with quite different current results for the two tubes which isn't unexpected given how they've been acting.  I swapped the power tubes to different sockets and the high current draw followed the tube, ~30ma for the better looking tube and ~40ma and rising for the not-so-hot looking one.  New tubes should be in tomorrow, I'm going to take the time to detail pots, etc. and update after checking the new tubes.

[jaymz168]

New tubes did the job as far as the bias imbalance and it sounds much better now.  After cleaning the pots they're still quite scratchy, I'm guessing the tracks are just shot since I don't see any DC on them causing it so new pots are on order.  I'm not happy with the (lack of) color coding of the wiring from the fuses to the switch so I'm going to redo all of that, maybe add X+Y caps, redo the heater lines as twisted pairs raised up from the chassis and other wiring, and replace some of the burnt up wires that the last person tore up with their iron.

Onwards and upwards!  After this I've got a Knight KN928 that I'm going to try to turn one of the inputs into a something proper for guitar.  Runs 12ax7s and 6BQ5/EL84s and has a built in 'mono' switch which simplifies things as well as 4+8+16 taps.

By the way if anyone here has a Google Pixel phone check out the night sight feature, this was taken in a very dark room!

[Dave]

Alright, this may sound a little unorthodox and some may yell at me about this, but... with regard to the pots... i use WD-40 and I believe in it. WD-40 is non-conductive and non-corrosive. It cleans the dirt off the resistive material (if you squirt it liberally) and lubricates the same. I have been using it for years and have never had any issues with using it. (Other than cleaning up after it). It, in many cases, will rejuvenate a noisy cruddy pot.

Dave

[jaymz168]

Admittedly I only blew them out with some canned air.  I think I'll spray them out tonight, I have the Deoxit "AV Survival Kit" which has a little bit of each formulation including the F5 and F100.  The big catch on this amp is that there are some odd potentiometers: a 2M audio taper that isn't showing up at the usual suspects and a 1.5M reverse audio which I've yet to find anywhere.  So I think I'm going to end up getting some practice rebuilding pots.  I've had this link in my bookmarks for a while, I guess it will probably come in handy!  http://www.geofex.com/article_folders/potsecrets/potscret.htm

[jaymz168]

OK so the volume pot didn't like the F5 or D5, I'm guessing the solvent just wiped out the track because it's all over the place.  I probably should have started with the F100 or D100 since they are solvent-free, but you live you learn.  250K audio taper are easy to replace which is why I started with that one and more are coming today.  It's the 2M A and 1.5M RA that are the tricky ones and currently they don't seem bad /fingerscrossed.

Anyway, is there any downside to leaving the 1R current sense resistors on the cathodes?  I was thinking of maybe putting them on terminal strips or something inside along with the bias pot mod for to make for quick and easy biasing in the future.

[Dave]

I don't see any reason why you couldn't move them anywhere you wanted them. Most people put them on the pot and then run a wire from the cathode to an outside-the-chassis bias port. But, as long as one end connects to the cathode and the other to ground, you are the creator.


Dave

[jaymz168]

Well it seems to work beautifully with a new input pot.  I'm going to live with it for a bit and decide if I want to change any values in the tone stack which looked to be modified already and maybe do a bias pot.  Also reinstall that mis-wired speaker output I removed because it's running off the EXT jack right now.  It IS a little noisy, I've only changed the power tubes so far but I do have some new 12AX7 and 12AU7 I could try out.  I also took out that death cap so there's no line filtering other than the choke I guess so maybe a Schurter filter, who knows.

Anyway, thanks for all of your help here, I'm very glad to have this up and running!