1. Time for some upgrades in server hardware and software to enhance security and take AK to the next level. Please contribute what you can to sales@audiokarma.org at PayPal.com - Thanks from the AK Team
    Dismiss Notice

Low (hot) negative bias voltage.

Discussion in 'Tube Audio' started by GuyNoir, Sep 9, 2018.

  1. GuyNoir

    GuyNoir VT Neophyte Subscriber

    Messages:
    1,734
    Location:
    Bismarck, ND
    Hi Folks:

    I'm troubleshooting my Sherwood 5000. At 117 variac VAC, plate voltage on the 7189 outputs are pretty much spot on...less than 406v. Bias, however is -13v.

    On 127 line volts, I'm getting right around 438 plate volts and -17.5 bias volts. Shouldn't the Russian 6n14n-EVs take the extra voltage in stride if I want to run it right off the wall socket, sans variac? The amp sounds magnificent, and while -17.5 is at the low end of normal, I'd like to get it up around 22 or so volts.

    At line voltage, I'm getting exactly -30v after the diode bridge (no dropping resistor), and exactly -24 out for the bias supply.

    My question is, should I start working backward from the output tubes, or start tracing from the bias supply?

    Thanks in advance!

    CF
     

     

    Please register to disable this ad.

  2. gadget73

    gadget73 junk junkie Subscriber

    Messages:
    36,764
    Location:
    Southern NJ
    I'd figure out how much current the output tubes are drawing and make adjustments to get it reasonable. The actual grid voltage is not as important as current flow through the tube. I think this one has a small-ish value resistor up at the front of the power supply that you can measure across to work out total amplifier current. Compare that to whatever the stock voltage drop was to see if the current draw is fairly right. The other way is to break the connection between cathode and ground on each output tube and put a 10 ohm resistor in that spot to measure across.
     
  3. kward

    kward AK Subscriber Subscriber

    Messages:
    3,459
    Location:
    Utah
    Russian 6p14p's can handle the extra voltage just fine. Pretty much the same electrically as a Sovtek EL84M. I'm not familiar enough with that amp to offer suggestions on how to change the bias circuit to get you the ballpark idle current you want, but certainly it must be possible.
     
    GuyNoir likes this.
  4. GuyNoir

    GuyNoir VT Neophyte Subscriber

    Messages:
    1,734
    Location:
    Bismarck, ND
    I'm pretty much new at troubleshooting, so pardon the newb question: Are you saying that this is an Ohm's law thing; bias is dependant on current draw? I was thinking that maybe theres's an out of spec resistor between the finals and the bias supply, but maybe that was simplistic. I do in fact have some 10 ohm resistors that I can solder in. I understand this also offers some protection to the OPTs. Thank you for the advice!
     
    Last edited: Sep 9, 2018
  5. GuyNoir

    GuyNoir VT Neophyte Subscriber

    Messages:
    1,734
    Location:
    Bismarck, ND
    Thanks Kward. I've been following with great interest your SE thread, though I see through a glass darkly.
     
  6. gadget73

    gadget73 junk junkie Subscriber

    Messages:
    36,764
    Location:
    Southern NJ
    The amount of grid voltage determines how much current will flow through the tube. The more negative it is, the less current flows. You use the cathode resistors to figure out how much current is flowing through the tube, and adjust the grid bias to get the tubes conducting the correct amount. Too much and the tubes burn out, not enough and you get increasing amounts of distortion.

    The 10 ohm resistor on the cathode will let you measure voltage drop and with a little magic you can convert it to current. 10 ohms makes the math easy. If you measure 0.4 volts, its drawing 40 ma.

    Somewhere the bias supply also probably has a voltage divider that works on Ohm's law. I don't have the schematic handy but the typical setup has output from a rectifier and eventually it goes through two resistors to ground. The point where those resistors meet is usually where the grid supply taps out, and changing the values of one or both of those resistors will alter the grid voltage.
     

     

    Please register to disable this ad.

  7. GuyNoir

    GuyNoir VT Neophyte Subscriber

    Messages:
    1,734
    Location:
    Bismarck, ND
    Okay... I think I understand, re: current flow. But bias is fixed in this unit. Bias voltage is healthy from the supply:
    s5000.gif

    My 10 ohm resistors are 1 watt. Do I need to get 1/4 watt so that they are 'fuseable'?

    Thanks for your patience with this tender foot.
     
  8. triode17

    triode17 Super Member

    Messages:
    1,308
    The whole idea with biasing is to set the tubes 'idle speed'. It all comes down to power dissipation of each power tube. We generally use a figure of 70% of max. plate dissipation. It has to be set using a nominal line voltage. Since you cited 117v and 127v, let's take the ave. of 122v.
    That should give you 422v on the plates. The 7591 is rated as 19 w plate diss. For a class A/B amp, the screen dissipation is about 5%,
    therefore, 19w x 0.7 =13.3. Subtract 5% and get 12.6 w. That's what you want to set it to. Since I=P/E, 12.6w/422v = 0.029a or 29ma.
    You should permanently install four 10 Ohm resistors. 1/4w is OK but forget the fusing idea. Resistors are bad fuses. I fuse the high voltage secondary lines, much safer. So at 122v in, set the current for 29ma or 0.29v across the 10 Ohms and that should do it.
    BTW, did anyone notice that the bias caps are drawn backwards in the previous schematic?
     
  9. GuyNoir

    GuyNoir VT Neophyte Subscriber

    Messages:
    1,734
    Location:
    Bismarck, ND
    Oh yeah...that schematic has caused more than a few people to wire the caps in backwards, only to have them explode. I got a com-positive multi-sectional ready-made from Hayseed Hamfest.

    This is a 7189 output s-5000, and am running 6P14P-EVs. I'll need to plug in your formula for those and go from there. Thanks!



    Untitled.png
     
  10. primosounds

    primosounds SE KT120 w/ 6J5G drivers. Subscriber

    Messages:
    2,717
    Location:
    Terra, 21st century CE
    If you have not done so , it would be best to completely rebuild the negative voltage bias supply with all new caps and metal film resistors. This should bring your voltages up to specs. For vintage units with fixed bias like Scott, Fisher and your Sherwood, it is good insurance and the same reasoning as rebuilding the main power supply.
     
  11. GuyNoir

    GuyNoir VT Neophyte Subscriber

    Messages:
    1,734
    Location:
    Bismarck, ND
    All can caps throughout have been replaced; the existing carbon resistors were in spec, though I'd be happy to update the resistors. That's the thing...the bias supply has more than specified voltage out, so I don't understand why I'm only seeing 17.5v at the tubes. I want to see it a little higher to maximize tube life.
     

     

    Please register to disable this ad.

  12. gadget73

    gadget73 junk junkie Subscriber

    Messages:
    36,764
    Location:
    Southern NJ
    Check the 4.7K and 2.2K resistors in the bias supply. If needed, you should be able to make the 2.2K larger to bump the bias voltage. If there is a resistor to ground at the other end of "B", that should also be tweakable to change the bias voltage. If the coupling caps leak, it will tend to drive the bias voltage high too, so beware of that. If the filter cap after the 4.7K is bad, or installed backwards, it will also cause you grief.

    I can't read the schematic clearly but it looks like there is a resistor to ground near the balance pot. Increasing the value of that will give more negative voltage. Can't begin to guess what part number or value it is though, the scan is too fuzzy.
     
    Last edited: Sep 10, 2018
  13. GuyNoir

    GuyNoir VT Neophyte Subscriber

    Messages:
    1,734
    Location:
    Bismarck, ND
    That I can do. I'm good on the 4.7k and 2.2k but haven't checked the one near the balance pot. Good tips, Mr. Gadget!
     
  14. gadget73

    gadget73 junk junkie Subscriber

    Messages:
    36,764
    Location:
    Southern NJ
    If the one near the bias pot goes, I'd expect it to drift high, which will make the voltage more negative. Usually thats how old resistors go, they rarely drift low.
     
  15. sp58

    sp58 Member

    Messages:
    69
    Location:
    Germany
    -17,5 Volt is o.K. for this 7189.
    Take a look at this 12AX7 tube - clean the pins, its heated with the Bias Voltage, if somebody changed the tube an the heatercurrent is higher then specs this will reduce Your BiasVoltage.

    The schematic for this sherwood 5000 shows 7189 or EL84 tubes, this is not right, an EL84 can´t handle over 400V in PP applications, data says 350V ( My experiences also says this )
    With Your Russian 6 - and so on -You are fine, had them in my receiver, they were the military version ,production date was stamped 1991. Built for use in MIG - Military-Jets - Very best You can get.
     
  16. GuyNoir

    GuyNoir VT Neophyte Subscriber

    Messages:
    1,734
    Location:
    Bismarck, ND
    Yeah, that schematic is in error - I don't imagine EL84 would last very long. Well, I just was just now taking measurements, and there is a wire running directly from the bias supply to the balance pots. I have -23 going into the pot and -20 volts is coming out of each pot. About -17.8 volts measured at the grid after a resistor or two. It has Sovtek 12AX7LPS in the front end, and practically new Raytheon (rca) 7199 phase splitters. I think I'm finally going to button down the lid and enjoy it. Danke!
     

     

    Please register to disable this ad.

  17. sp58

    sp58 Member

    Messages:
    69
    Location:
    Germany
    We got it. I plugged in the same Longplate Sovtek´s ( 2Pcs) into the phono preamps of my Pioneer SX 800 - some years ago - they were offered with a special anti hum filament or such ?
    As i remember , i was a little bit confused - didnt´t see tubes glowing - and did some measures of the filament current - it was very different to orig. specs.
    If You like, try it with standard 12AX7 and watch the bias voltage.

    Bitteschön !
     
  18. GuyNoir

    GuyNoir VT Neophyte Subscriber

    Messages:
    1,734
    Location:
    Bismarck, ND
    Long Plate Spiral Filament. Right - they don't glow at all, and are cool to the touch as well....I measure 50-65 c. The 7199 are a little harder to hold on to, and the output tubes are around 140 c. I do have a brand new set of 12AX7WA Sovteks, but they hum terribly. In fact, I think they are the reason I fell down this rabbit hole...looking for hum on an already refurbished amp. One or all of them must be faulty. I might try them out again. If they are bad, I've been itching to try JJ ECC803s. I've heard good things about them, and they aren't terribly expensive. Good catch, BTW!
     
  19. GuyNoir

    GuyNoir VT Neophyte Subscriber

    Messages:
    1,734
    Location:
    Bismarck, ND
    Well curiously enough, these are even quieter than the 'LPS. I had one nagging little hum left that I couldn't track down after all of my efforts. It is now GONE with the 'WA in there. Crazy. :D
     
  20. sp58

    sp58 Member

    Messages:
    69
    Location:
    Germany
    What about the Bias Voltage now ?
    Hum must come from the line voltage, because the 12AX7 gets DC -Filament
     

Share This Page