JTM45 build - bias too high

Discussion in 'The Workbench' started by badoogie, Aug 30, 2020.

  1. badoogie

    badoogie Member

    Joined:
    Feb 3, 2014
    Messages:
    32
    Likes Received:
    0
    I’ve built a JTM45 based on the MetroAmp instructions, and it was sounding really good (aside from some squealing with the presence above 8). All of a sudden it started making a really loud continuous machine gun-like sound. I checked the bias and it had risen on each output valve to about 70mV. I adjusted it back to about 42.5mV and the noise went away for a couple of minutes. Then it came back. I checked the bias and it had risen to about 70mV again. Unfortunately this time adjusting the pot did not lower the reading, it would only increase it to about 90mV. Does anyone have any idea what might have caused this?


    I have to say, the noise produced was impressive, like the opening of Audioslave’s ‘Cochise’, without the musicality!
     
  2. Chris-in-LA

    Chris-in-LA Well-Known Member

    Joined:
    Jul 19, 2018
    Messages:
    1,677
    Likes Received:
    1,378
    Location:
    Los Angeles
    Bad bias pot? Bad caps? Can you post some pics of the insides?

    Side question; you built this from parts you spec’d yourself?
     
  3. StingRay85

    StingRay85 Well-Known Member

    Joined:
    Aug 18, 2019
    Messages:
    589
    Likes Received:
    311
    Redo the negative bias supply I would say. You can use a voltage doubler scheme if you would be limited in the supply. Its just one more cap and diode
     
    thetragichero likes this.
  4. Guitar-Rocker

    Guitar-Rocker Well-Known Member

    Joined:
    Mar 5, 2012
    Messages:
    2,832
    Likes Received:
    2,475
    Location:
    Indianapolis, Indiana
    Start by pulling all power tubes, with speaker load connected, fire the amp back up and monitor the bias voltage. With no power tubes in, the amp's bias should stay consistent at whatever reading it is set on, and you should be able to get a reasonable range of adjustment by twisting the bias pot. If the bias varies while idleing, then troubleshoot that section of the amp.
     
    pedecamp likes this.
  5. South Park

    South Park Well-Known Member

    Joined:
    Sep 14, 2019
    Messages:
    2,967
    Likes Received:
    2,486
    Pull the power tubes and stand by fuse . You should have a dead power rail . Now check bias voltage is it positive or negative and check filter caps . If any voltage in in filter caps your wire lay out is wrong . I’m have a 2204 with this problem building a new circuit board
     
  6. badoogie

    badoogie Member

    Joined:
    Feb 3, 2014
    Messages:
    32
    Likes Received:
    0
    Thanks for the replies guys. :)

    So far I’ve checked the bias pot and adjacent 68k resistor, and they both measure correctly.

    I removed the output valves but then found no bias voltage at either socket (pin 8).

    I started working through the tests on page 20 of the Metro instructions. The heater voltages measured correctly, as did the voltage between pins 2 and 8 of the GZ34. Then I think I found something. I’m getting a good voltage at the wall (234V AC) but at the standby switch I measured 434V DC! This should be about 325V, as it was when I first built the amp. :shrug:
     
  7. neikeel

    neikeel Well-Known Member

    Joined:
    Feb 12, 2014
    Messages:
    4,722
    Likes Received:
    3,132
    434vdc is a more normal voltage at the standby switch on a JTM45, 300vdc is not.
    Strictly speaking you are not measuring voltage across the 1&8 pin tied resistor that goes to ground. You are measuring current by the voltage drop across the resistor when the tube draws current so you will get no measurement unless the tube is in and switched on.

    Without out put tubes in (and all the others in) what are the negative volts on pin 5 of the output tubes and what do you get at extremes of the bias pot? What voltage do you get on pin 3?

    Get those numbers and we can help.
     
  8. badoogie

    badoogie Member

    Joined:
    Feb 3, 2014
    Messages:
    32
    Likes Received:
    0
    Thanks for the reply neikeel. I put the KT66s back in and at first the bias readings were okay. Then I measured pin 5 of V5 first, and it was in the -40s VDC. I went over to pin 5 of V4 .... and there was a spark. :facepalm: I don’t think I touched the probe on anything other than pin 5, so I’m not sure what happened. Unfortunately the spark was followed by a continuous fast ticking sound! :ohno: This was coming from somewhere in the amp itself, not the speaker (which was muted through a Powerbrake).

    Any ideas what’s going on?
     
  9. badoogie

    badoogie Member

    Joined:
    Feb 3, 2014
    Messages:
    32
    Likes Received:
    0
    I decided to take the KT66s out and run through some of the tests on page 20 of the Metro instructions again. Pin 5 of V4 and V5 was about -53VDC. There was about 5VAC between pins 2 and 8 of V6.

    Some of the preamp VDC readings were a bit higher than the manual suggested. These are the readings I took:

    V1
    Pin 1 256
    Pin 3 1.6
    Pin 6 250
    Pin 8 1.6

    V2
    Pin 1 214
    Pin 3 1
    Pin 6 352
    Pin 8 214

    V3
    Pin 1 276
    Pin 3 42.5
    Pin 6 267
    Pin 8 42.5

    Without the KT66s in the continuous ticking sound has gone, so I’m guessing the sound was coming from one of them.

    Is there anything else I should test before installing new KT66s?
     
  10. neikeel

    neikeel Well-Known Member

    Joined:
    Feb 12, 2014
    Messages:
    4,722
    Likes Received:
    3,132
    Check your output load is correct and matches your selector.
    Check both fuses are of correct type and value.
    -50v on pin 5 is a safe (ie cold) bias to start with. Btw when you wind the pot in opposite direction how high does it go (if it does not vary you may have an issue).
    Temporarily desolder one limb of your NFB resistor.
    Install the tubes and recheck the voltages. Do you get hiss from the speaker?
    If it is stable and working then with an insulated screw driver push the resistor into contact with the turret. If it ticks, motorboats or screams you have reversed primaries.
     
  11. badoogie

    badoogie Member

    Joined:
    Feb 3, 2014
    Messages:
    32
    Likes Received:
    0
    The speaker load and fuses are all good. With the bias pot fully counterclockwise pin 5 of V4 and V5 are -40VDC.

    If the primaries were reversed would the amp still have been able to operate correctly when I first built it? :confused:
     
  12. neikeel

    neikeel Well-Known Member

    Joined:
    Feb 12, 2014
    Messages:
    4,722
    Likes Received:
    3,132
    I guess I read the squealing presence and then put putting noises which can both be a symptom. How long had you played it for before it went off?
     
  13. badoogie

    badoogie Member

    Joined:
    Feb 3, 2014
    Messages:
    32
    Likes Received:
    0
    I’d say it had been running well for a total of about twenty hours. It sounded really good!
     
  14. junk notes

    junk notes Well-Known Member

    Joined:
    Dec 10, 2018
    Messages:
    1,624
    Likes Received:
    1,656
    Location:
    house arrest
    edit: For future maintenance perhaps getting yourself cheap (power) tubes to have on hand as to do quick troubleshooting with one, or two, or the quad at once. Then you are able to put your good set in after you have narrowed it down.
     
    Last edited: Sep 5, 2020
  15. badoogie

    badoogie Member

    Joined:
    Feb 3, 2014
    Messages:
    32
    Likes Received:
    0
    I tried another pair of KT66s, but the problem is still there. :( To summarise:

    - The amp sounded great for about twenty hours, and the bias readings were good, so I’m pretty sure everything is wired correctly.

    - When the output valves are installed and the amp is turned on there is a continual fast ticking noise. This comes from the amp itself, not the speaker.

    - With the bias pot fully clockwise the bias readings at pin 8 are 280mV on V4 and 60mV on V5.

    :confused:
     
  16. junk notes

    junk notes Well-Known Member

    Joined:
    Dec 10, 2018
    Messages:
    1,624
    Likes Received:
    1,656
    Location:
    house arrest
    Check all electrolytic caps to see if they dissipate similar to each other.
    Turn on amp, then turn off the amp.
    Check each of the electrolytic capacitors, and watch how your meter reads the cap to see how fast or slow it drops the charge.
    good luck.
     
  17. neikeel

    neikeel Well-Known Member

    Joined:
    Feb 12, 2014
    Messages:
    4,722
    Likes Received:
    3,132
    that bias is an issue. You need to check the connections to the grids. If your tubes out readings are steady then you have a duff tube. Does the really high reading follow the tube?
     
  18. badoogie

    badoogie Member

    Joined:
    Feb 3, 2014
    Messages:
    32
    Likes Received:
    0
    Thanks for the suggestions guys. :wave: I swapped the output valves, and now the bias readings at pin 8 are 132mV on V4 and 28mV on V5. Both readings are lower with the valves swapped, but either way around V4 is much higher than V5.
     
  19. badoogie

    badoogie Member

    Joined:
    Feb 3, 2014
    Messages:
    32
    Likes Received:
    0
    I just measured some of the resistors around V4 and V5, and I found something with the 1 ohm bias resistors. The one on V5 measured just over 1 ohm, but the one on V4 measured just over 4 ohms. Those measurements were taken with the resistors in place, so I’m not certain something else in the circuit wouldn’t affect the readings. Should I replace the bias resistor on V4?
     
  20. Chris-in-LA

    Chris-in-LA Well-Known Member

    Joined:
    Jul 19, 2018
    Messages:
    1,677
    Likes Received:
    1,378
    Location:
    Los Angeles
    It’s not uncommon to get a 3 or 4 ohm reading on a 1 ohm resistor. When I check the resistance in the leads to my meter, they can read 2 or 3 ohms of resistance. It’s hard to read exactly 1 ohm of resistance, it’s probably fine.
     

Share This Page