Gibson Ga20t Grid Leak Bias-change It?

Discussion in 'Amps/Cabs Tech Corner: Amplifier, Cab & Speakers' started by arfy, Jun 13, 2006.

  1. arfy

    arfy Member

    Messages:
    153
    Joined:
    Oct 8, 2002
    Location:
    Milwaukee
    I recently restored a Gibson GA20T to working order and although it gets a pretty nice tone, the 12AX7 channel is very noisy and has NO sustain even when it's getting a little crunchy.
    I'm thinking of changing the input stage to common cathode bias, anyone done this or have any comments/recommendations?
    I've read in a couple of places that those yellow astrons are usually leaky so I'll probably try replacing those too.
    What I'd really like this amp to do is be like a tweed Deluxe, so any ideas or comments are welcome.
     
  2. 1guitarslinger

    1guitarslinger Member

    Messages:
    528
    Joined:
    Feb 26, 2005
    Do you mean the 12AY7? Sure, it could be set up cathode biased. The 5879 and 12AX7 are already cathode biased.

    Us a 820 ohm resistor and a 25uF 25 volt bypass cap. You will want to lose the coupling capacitors just after the input jacks and add resistors. You can decide on the values.

    Check to see if any of the Astrons are leaky and replace them. Check and replace drifted resistors as well.

    Have you replaced all of the electrolytics and installed a 3 prong grounded power cord?

    What have you done to the amp so far?
     
  3. arfy

    arfy Member

    Messages:
    153
    Joined:
    Oct 8, 2002
    Location:
    Milwaukee
    I replaced all the electrolytics and put in a 3 prong plug with the black wire going to the fuse first, and besides pot cleaning that's about it. The foot switch was missing so I wired one in, the tremolo is pretty nice sounding. I didn't have a 12AY7 so I put a 12AU7 in that first spot, I know that's a little less gain. A 12AX7 was really noisy in that spot. Maybe it needs to have the 5879 channel turned up a little to sound good since this amp has interactive volume controls.
    It was bought at a ham radio swap meet with the the head out of the cab, OT disconnected and the speaker wired to a zip cord instead of the OT, I don't know any history, there were 2 12AX7 in it as well as the 5879 and the octal oscillator tube(forgot the number), RCA 6V6 and an RCA 5Y3 which I replaced with a NOS GE.
     
  4. donnyjaguar

    donnyjaguar Member

    Messages:
    4,243
    Joined:
    Nov 4, 2005
    Location:
    Canada
    Those Hamsters are notorious for bastardizing their equipment. All the musical instrument amp's I've come across at hamfests, and I've been to many, are completely whipped. :( My friend and I picked up Dynaco ST70s once and they were only slightly modified, which was nice. I still have mine and it works fine. Adding cathode bias to your amp is the wise thing to do. Why manuf's didn't do this from the beginning I don't know.

    DJ
     
  5. 1guitarslinger

    1guitarslinger Member

    Messages:
    528
    Joined:
    Feb 26, 2005
    I suggest getting a 12AY7 for V1. Grid leak bias is so quick to distort, that a 12AX7 may be too much, and is probably putting out way too much for the PI to deal with. Keep in mind that the gain factor of a 12AY7 is 44, and a 12AX7 is 100. Using a 12AX7 is almost a 150% increase in gain over what the designers intended. And your 12AU7 is 50% less than intended.

    Adjusting the other channel's volume control can't really make up for the mismatch.

    The new EH 12AY7s sound really nice. Have you replaced the other tubes?

    Paul
     
  6. 1guitarslinger

    1guitarslinger Member

    Messages:
    528
    Joined:
    Feb 26, 2005
    Hey Arfy, I just got a GA-6 in here with a grid leak 12AY7 V1. After getting it fixed up, new electorlytics, cleaning etc. it sounds fantastic. I did notice that it is tame with the 12AY7, so I tried a 12AX7. The noise got worse, and the distorion was not very pleasent.

    With the 12AY7 being so mellow, I could see a 12AU7 being downright flaccid.
     
  7. arfy

    arfy Member

    Messages:
    153
    Joined:
    Oct 8, 2002
    Location:
    Milwaukee
    OK, so I am going to leave this amp as stock as possible, I agree that's the best way, so I've been measuring parts and replaced a couple of leaky Astrons, etc, compared values with a schematic from www.freeinfosociety.com, etc. www.freeinfosociety.com/electronics/schematics/audio/gibsonga20t.pdf And there are a couple of parts that vary from the schematic by a factor of 10. The grid leak resistors on the 12AY7 are listed as 10 meg but the ones in the amp are 1 meg, brown-black-green, and measured a little over 1 meg. Early Fender amps with similar circuits have 5 meg grid leak resistors, I believe that kind of bias is called 'contact bias.' One of the 1 meg resistors had broken at one lead, anyway, I put 10 meg resistors in and will have to see how it works when it's all put together again.
    The schematic lists a 47k resistor between a pair of 220k's but the amp has a 4.7 in that spot-yellow-violet-red.
    What gives? Anyone have similar experiences with older Gibsons?
    I'm guessing the difference might be fairly subtle with the 4.7k but the difference between 10 meg and 1 meg in must be fairly stark-the amp sounded nice when I did use it but seemed to lack sustain, maybe the grid leak resistor is the factor...
     
  8. 1guitarslinger

    1guitarslinger Member

    Messages:
    528
    Joined:
    Feb 26, 2005
    Someone may have changed the 10 Megs to 1 Megs by mistake, or in an attempt to emulate a more modern circuit, or like Fender, things may have always kind have been in flux. 10M and 5M are both so high in the whole scheme of things, that it is unlikely that there is any audible difference. I would switch the 1M to 10M as well.


    As far as the 4.7k resistor, if you'll enlarge the schematic on your link, you will see that it is listed as 4.7k. The penmanship on Gibson schematics is not nearly as nice as old Fenders.

    The difference between 4.7k and 47k, and 1M and 10M are both a factor of 10

    I hope this helps.

    Paul

    Edit: V1 grid bias is pretty weird. If you are not concerned about collector's value of the amp, changing it to cathode bias might be a good idea. Otherwise, enjoy its "funkyness".
     
  9. jhguitarlab

    jhguitarlab Member

    Messages:
    3
    Joined:
    Jan 12, 2009
     
  10. capnjuan

    capnjuan Member

    Messages:
    98
    Joined:
    Jun 21, 2007
    Hi jh; I have two GA20Ts that I'm overhauling; one matches the public domain schematic (.01 input caps, 10meg grid leak Rs) and the other has 3 6EU7s plus a 5879. The trem runs in both of them.

    If you have the 'caps under' version, I suggest you flip the board (again :mad:) and replace the three .047/400-600v trem caps. One leg of the footswitch goes to one of the legs of the depth pot; you can do it wrong, all that can happen is that the pot works in the opposite direction - if so, move the connection to the other side ... the other leg gets tied to the ground legs passing along the volume pots. Good luck! CJ
     

Share This Page