Grid Leak Biased 6SJ7

Discussion in 'Shock Brother's DIY Amps' started by dunner84, Feb 17, 2020.

  1. dunner84

    dunner84 Tele-Meister

    Posts:
    313
    Joined:
    Jul 15, 2014
    Location:
    Victoria, bc
    Hello all,

    I am doing some tweaking on a new build which started as a stock rebuild of a 50s PA. It sounds killer with my guitar, but this is supposed to be a harmonica amp, and I am having issues getting it tweaked for harmonica. The circuit is supposed to be a grail tone for harp players, but I am getting feedback at volume 1. I have tried a few things to get some more range, and I can now get the volume control up to 2.. I am looking for some feedback to help me get this to be more harp friendly..

    I built the circuit stock, but fully removed the phono input.
    Here are things I have tried already:
    1 - Lowered R6 to approx 82K
    2 - added a 470k resistor where the phono volume control would have been
    3 - added a 2.7M and 75K resistor at the input

    All the voltages are spot on, and the bias is 80% with old tung sol 5881, and 100.5% with plain old 6l6 (19W) tubes

    I have read that switching to cathode bias may help lower the preamp gain, but I am worried it will mess with the mojo of the amp.
     

    Attached Files:

  2. muchxs

    muchxs Doctor of Teleocity

    Posts:
    13,079
    Joined:
    Aug 14, 2004
    Location:
    New England
    Wait. I thought feedback was undesirable?!

    My teachers told me I was a smart kid back when I was a smart kid. Turns out I'm a smart ass...

    :)

    You sure you're not having problems with your mic?


    355v on the plates? Swap to 6V6s. Use a 250 ohm 5 w bias resistor.

    I'd bleed some gain straight to ground after the volume pot. I'd start with a 470k resistor and go lower until I tamed the beast.
     
  3. radiocaster

    radiocaster Poster Extraordinaire

    Posts:
    6,803
    Joined:
    Aug 18, 2015
    Location:
    europe
    It is cathode biased in the schematic. edit: the output stage is, I thought you were referring to that, although I guess you were referring to the preamp??
     
    Last edited: Feb 17, 2020
  4. muchxs

    muchxs Doctor of Teleocity

    Posts:
    13,079
    Joined:
    Aug 14, 2004
    Location:
    New England
    It goes without saying so it mostly goes unsaid:

    Use a longer cord between your amp and your mic! Put some distance between you and your amp.

    If you're facing it in a small room it's going to sound like a litter of hungry piglets.


    There are guys who use tweed Bass- Men for harp amps. Don't arm wrestle those guys! They have bulging forearms from tightly cupping their mics. Occupational hazard I guess if you want to call it that.


    If it still squeals you can EQ your way out of the squeal. The bad news: You may EQ your way out of your tone.
     
  5. dunner84

    dunner84 Tele-Meister

    Posts:
    313
    Joined:
    Jul 15, 2014
    Location:
    Victoria, bc
    Sorry, that was not clear. I was referring to the 6sj7 which is currently grid leak biased
     
  6. radiocaster

    radiocaster Poster Extraordinaire

    Posts:
    6,803
    Joined:
    Aug 18, 2015
    Location:
    europe
    I was thrown off by the bias readings earlier, it makes sense now. My fault.

    6L6s are usually 30W plate dissipation though. May vary, check data sheet.
     
  7. dunner84

    dunner84 Tele-Meister

    Posts:
    313
    Joined:
    Jul 15, 2014
    Location:
    Victoria, bc
    So this will dump more signal than the 2M resistor to ground, and 75k resistor to the grid capacitor that I added? I will admit, I'm not 100% sure of the working behind common inputs. I went that direction from looking at common harp mods to guitar amps, and it looked like changing the 1M resistor to ground, (common in guitar amps), to something higher was common place.

    Also, this is with a variety of mics. I just thought that this amp would have a more useable volume sweep than some of the guitar amps I play through.
     
  8. dunner84

    dunner84 Tele-Meister

    Posts:
    313
    Joined:
    Jul 15, 2014
    Location:
    Victoria, bc
    I should clarify further. The 5881 are NOS tungsol 23W tubes, and the 6l6 are also NOS 19w variety, not modern 6l6gc types.. I referenced them only because that is was the amp was originally designed for
     
    radiocaster likes this.
  9. muscmp

    muscmp Tele-Afflicted

    Posts:
    1,623
    Joined:
    Jul 1, 2008
    Location:
    california
    hopefully you've already done this to eliminate feedback, but, as a harp player myself, first thing i do is to turn the treble way down and the bass way up. go from there.

    play music!
     
  10. dunner84

    dunner84 Tele-Meister

    Posts:
    313
    Joined:
    Jul 15, 2014
    Location:
    Victoria, bc
    Thanks for all the support on this. I think it's getting there.
    T
    grid leak bias is still a bit above my head, so for curiosity sake, how would halving the grid 15M resistor, and doubling the input capacitor shift the bias? Up? Down? And which way would I want to shift the bias to allow for a hotter input?
     
  11. David Barnett

    David Barnett Doctor of Teleocity

    Age:
    63
    Posts:
    12,447
    Joined:
    Mar 17, 2003
    Location:
    The Far-Flung Isles of Langerhans
    I've seen Kim Wilson blowing through TWO blonde Twin-Amps. In a 300-seat club.
     
  12. Junior Little

    Junior Little Tele-Meister

    Posts:
    207
    Joined:
    Jul 26, 2018
    Location:
    Rural Illinois
    I would check out a couple sources to read up on what they say about grid leak bias. One stop would be Lone Wolf and the schemos found there. Another would be Merlin, and a third would be the Megantz book on the design of tube amps. Also, check out Rob's site for drawing load lines. Made me get at least one step closer to understanding just what was going on in that neck of the circuit woods.
    Also, as @muchxs says, tight cup is a necessity. But you already knew that.
     
    Last edited: Feb 18, 2020
  13. printer2

    printer2 Poster Extraordinaire

    Posts:
    6,655
    Joined:
    May 24, 2010
    Location:
    Canada
    The cathode is grounded, the only way the grid is biased negative if it picks up some of the electrons flowing by. We do want a way to bleed off some of them so we used the grid leak resistor. I think 15M might be too high, I would go for 4.7M. If that fixes things, good. If not, the gain of the pentode is partly determined by the screen voltage. See page 5, 50V on the screen only requires 3.5V to go from cutoff to saturation. With 100V it needs 8V to go from cutoff to saturation.

    You have 23V on the screen, the maximum signal is approximately the bias voltage as on the schematic, 0.8V p-p. Think of the screen resistor and capacitor as another RC power supply node. If you lower the screen resistor it will put more voltage on the screen. Another part of the equation is the plate resistor. Reducing its value will help to reduce the gain. Playing with the screen and plate resistor will get the input more user friendly. You can swap resistors around for the two and tweak by ear. The easiest way to do it would be to put a 49k resistor in the plate with a 250k pot in series, and a 470k in the screen with a 1M resistor and a 1M pot. If need be remove the 1M resistor. Twiddle the pots, mainly changing one and touching up the sound with the other. Then do the reverse, adjust the other pot then tweak with the first. I did this with another tube and with a cathode resistor, pot actually. What would have had me pulling my hair out with individual resistors was a fun experience with the pots.

    http://www.r-type.org/pdfs/6sj7.pdf
     
  14. tikitorch

    tikitorch TDPRI Member

    Posts:
    12
    Joined:
    Jul 25, 2017
    Location:
    Montana
    Just get rid of the screen grid bypass cap. Don't mess with the plate and screen resistors. You'll have a gain of around 50 after that.
     
  15. dunner84

    dunner84 Tele-Meister

    Posts:
    313
    Joined:
    Jul 15, 2014
    Location:
    Victoria, bc
    Thanks for all the input. Lifting the screen bypass cap made a massive difference. I can now crank the volume, and tone with a decent mic. The problem now, is it shaved off too much. If I wanted to find a happy medium, should I go with a larger or smaller cap?

    Cheers,
     
  16. tikitorch

    tikitorch TDPRI Member

    Posts:
    12
    Joined:
    Jul 25, 2017
    Location:
    Montana
    Hmm... Not enough gain now? Changing cap size on the screen bypass cap (C5) will change the frequency response of the tube, a smaller cap will allow less bass to go through much like a cathode bypass resistor does. Not sure if that's what you're looking for. You could put the screen cap back in and change it to cathode bias but don't use a cathode bypass cap. Merlin Blencowe (The Valve Wizard) has a page that explains a bit about setting up pentodes you should start there. Your tube is set up for pretty high gain as it is and you might have to change everything about the stage like printer2 was describing to really dial it in.

    The easiest thing you might try is putting the G2 cap back in but put a small resistance between it and ground. Start with around 1k and experiment from there.

    Or, go cathode bias. Try dropping your plate and screen resistor values to about 1/2 their size say 220k and 1M and put in a 1k cathode resistor but leave the cathode bypass cap out. Connect the suppressor grid and screen bypass cap directly to the cathode. Don't forget to change your input circuit to cathode bias values. Take a look at Gibson 6SJ7 circuits like the GA-9.

    Anyway, those are my suggestions YMMV. Take a look at the 1951 Sylvania Tube Manual near the back of the book there are suggested values for 6SJ7 resistance coupled circuits. That should give you some ideas. http://www.tubebooks.org/sylvania_tube_manuals_online.htm
     
  17. tikitorch

    tikitorch TDPRI Member

    Posts:
    12
    Joined:
    Jul 25, 2017
    Location:
    Montana
    6SJ7 RCC Data.gif
     

    Attached Files:

  18. tikitorch

    tikitorch TDPRI Member

    Posts:
    12
    Joined:
    Jul 25, 2017
    Location:
    Montana
    sylvania_1951.gif
     
  19. tikitorch

    tikitorch TDPRI Member

    Posts:
    12
    Joined:
    Jul 25, 2017
    Location:
    Montana
    Let me know how it goes. I reread your original post and it looks like you've made some deviations from the schematics you posted. It would really help if you posted a schematic that shows exactly what you have right now along with voltages of the power supply, plates and screens.
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.


  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.