Amp hums but gets better if you push in on the tubes!?!?!?

Discussion in 'Amps/Cabs Tech Corner: Amplifier, Cab & Speakers' started by sidehatch, May 5, 2016.

  1. sidehatch

    sidehatch Member

    Messages:
    1,208
    Joined:
    Jul 24, 2008
    Location:
    zephyria
    I'm trying to figure out what is wrong on a friends amp. A cheap carvin vintage 33. It has a very decent clean tone but has a low hum that makes it unusable.

    I tried replacing all tubes with known good ones and poking around with a chopstick.

    I've checked the bias .

    It does it on both channels.

    He's had it to the tech 3 times over the past few years and he's tired of soldering on the flimsy circuit board Thats why I'm messing with it. I'm not a trained repairman but can usually poke around and figure obvious things out. I know how to discharge the amp to not kill myself with it.

    Any help will be most appreciated!
     
    Last edited: May 5, 2016
  2. DeaconBlues

    DeaconBlues Member

    Messages:
    2,988
    Joined:
    May 2, 2007
    Location:
    way down below the ocean...
    I'd look at the solder joints that he has messed with first. Has he done any soldering near the tube sockets?
     
  3. The Dirty Tube

    The Dirty Tube Member

    Messages:
    1,642
    Joined:
    Aug 14, 2012
    Location:
    A Child's Garden of Glass
    Modern tube sockets are 1) awful, and 2) PC mounted. It can take just a couple tube swaps to botch the contacts.

    PC mounted tube sockets may be the single worst concept adopted by the entire mass-market amp industry.
     
    matt1969 and pdf64 like this.
  4. sidehatch

    sidehatch Member

    Messages:
    1,208
    Joined:
    Jul 24, 2008
    Location:
    zephyria
    Should I just resolder all of the joints on the sockets or is there a method to determine whether they're bad before getting out the iron?
     
  5. dbtech

    dbtech Member

    Messages:
    33
    Joined:
    Aug 29, 2015
    Location:
    Arizona
    Resoldering is a troubleshooting technique. Re-solder all pins on the tube sockets and re-test; if fixed, it means one or more solder joints were bad.

    The tube socket pin receptacles could be dirty. Spray a little De-Oxit or equivalent on the tube pins then quickly insert-remove-insert-remove-insert the tube into the socket. This should "clean" off any dirt that may be causing the problem.

    It could be that the pin contacts in the socket are not tight enough. Depending on socket design, you may be able to slightly pinch or squeeze the metal so it better grabs the tube pins.

    Also, there is a possibility of the copper trace fracturing right at the edge of the trace ring that the socket pin goes through on the PCB. If this is discovered, you must shave off any coating on the trace (usually green in color; I use an X-Acto knife or modeling knife for this) before soldering - you must solder to bare copper. If severely damaged, just replace with a wire.


    I would see if any of these things help before getting too far out in the twilight zone on shielding and grounding problems. db
     
  6. sidehatch

    sidehatch Member

    Messages:
    1,208
    Joined:
    Jul 24, 2008
    Location:
    zephyria
    I'm going to try it now. I also noticed that there is very little gain on the drive channel. Not that the drive channel would be used anyway . I'm wondering if whatever is causing the low gain on that channel is also causing the hum?
     
  7. sidehatch

    sidehatch Member

    Messages:
    1,208
    Joined:
    Jul 24, 2008
    Location:
    zephyria
    I just resoldered the tube sockets and sprayed them and its still the same. I wonder where to check next?
     
  8. bob-i

    bob-i Member

    Messages:
    7,952
    Joined:
    Oct 16, 2005
    Location:
    Central NJ
    That's simply not true. Board mounted tube sockets have been used for years, and by high end manufactures like Bogner, Fuchs, and mid level builders like Peavey and Fender.

    As for the troublesome amp, what is needed it to trace the signal path with a scope and look for the stage where the hum starts. Once it's located it will be easier to troubleshoot. Shot gunning doesn't seem to be working.
     
  9. Tone_Terrific

    Tone_Terrific Supporting Member

    Messages:
    26,635
    Joined:
    Dec 26, 2004
    Location:
    Canada-GTA
    Better if you push on the tubes?
    Which tubes?
    If you can isolate to a specific tube use a magnifying glass (assuming the socket contacts are good) to watch the solder point on the pcb while you wiggle the tube.
    If it quiets down when pushing any tube check grounding connections, ribbon cables and any pcb area that is affected by flexing...which is likely a lot of stuff.
     
  10. UsableThought

    UsableThought Supporting Member

    Messages:
    1,620
    Joined:
    Jul 18, 2015
    That X, Y, and Z all do something doesn't automatically make it a good practice - or a bad one, either. The logical fallacy is similar to appeal to popularity & appeal to authority.

    Rhetoric & logic aside, "so and so does it" ignores the reason that all manufacturers, including luxury & high end, are forced to build less than perfect products: cost.

    Even a brief survey of this board over a few days will reveal complaints about design/build failure with products from not just Fender or Peavey and the like, but high-end or boutique amp builders as well. Peavey has an excellent rep, yes. But they can't build products that don't fail any more than anybody else. It can't be done at a price point that anyone could afford. They just try & build products that are less likely to fail and then give good service when failure does occur. The more units you build and the lower the price point, the less it's about "best practice" and the more it becomes "a practice we can get away with if we don't screw it up."

    Now, as to putting sockets on a PCB - it seems obvious this creates a structural weakness that the builder must then remediate one way or another - e.g. beef up the board, add support with bolts through nearby metal, etc. Unfortunately some builders fail to take such steps. Even then, there may be more stress than if the sockets were mounted off-PCB; we'd have to get deep into failure reports to see. Or look at PCBs themselves: they can be done well or poorly or in between; the details matter, not just "who does it." With all things we have to dig into the specifics.
     
    Last edited: May 8, 2016
    The Dirty Tube likes this.
  11. sidehatch

    sidehatch Member

    Messages:
    1,208
    Joined:
    Jul 24, 2008
    Location:
    zephyria
    Any tube pushed in will quiet the hum considerably. I've checked all the ribbon cables with a chopstick and it doesn't help.

    I can attest to the carvin's circuit board sucking. Its very thin. The tubes can just be slightly touched and will move sideways in their slots.

    Their customer service sucks too. They won't repair the amp (even for money) and the best they can offer is to send me a new chassis without tubes for $400! Funny thing is I have a 50% off agreement deal from when I used to tour heavily and I could buy the updated nomad model for $275. But they want $400 for just a chassis on a model they no longer sell. I'm not buying anymore amps that have to be thrown out in a few years.
     
  12. UsableThought

    UsableThought Supporting Member

    Messages:
    1,620
    Joined:
    Jul 18, 2015
    This makes me wonder if a track on the board has fractured - not necessarily proximal to a tube socket - such that when you push on a tube, you are pushing on the board in general & flexing it in a way that gets the trace to connect more firmly. Or it could be a bad solder joint somewhere away from the sockets, which could explain why your soldering touch-up did nothing.

    Regardless, I think @bob-i is correct - as with so many pervasive but hard-to-pin-down noise problems, the best approach is to try & find where the hum enters the signal path. A scope might be helpful, provided you have experience in seeing what different sorts of noise look like & the hum has enough amplitude to show up; but even better in my opinion is a signal tracer hooked to an audio amp (to listen, with dummy load on OT instead of a speaker) or to a spectrogram app (to listen & look). However I'm getting tired of recommending this strategy, because most DIY'ers aren't interested - apparently it seems too much work compared to guessing, re-soldering every solder joint, throwing parts at the problem, etc. If you are interested I can tell you more, or bob-i could, or pretty much anyone w/experience with a scope or tracer.

    Otherwise, there is one other simple thing you might try - which is pulling tubes to see if pulling a particular tube kills the hum. That might narrow it down to a specific area in the signal path . . . or not. A schematic (see here) would be a good thing to start looking at once you start thinking "signal path" rather than "what might be broken that I can see." Since you say it's both channels that gives you a starting point, if somewhat vague. What happens if you pull V1 (input for both channels)? If you pull V2 (drive channel only)? Etc. If you pull a tube & it kills the noise then you can do a visual inspection and also chopstick components & traces in that area, pushing on them (gently!). There seems some possibility that if pulling tubes doesn't do anything, the hum comes from a circuit that is more general in nature, e.g. heater circuit losing its ground, etc.
     
    Last edited: May 8, 2016
  13. Tone_Terrific

    Tone_Terrific Supporting Member

    Messages:
    26,635
    Joined:
    Dec 26, 2004
    Location:
    Canada-GTA
    Isolate the hum to the audio path or the power supply.
    Go from there.
     
  14. The Dirty Tube

    The Dirty Tube Member

    Messages:
    1,642
    Joined:
    Aug 14, 2012
    Location:
    A Child's Garden of Glass
    Then we'll just have to disagree. Used for years, yes, and always awful. Cold socket joints are common on even the early 1960's amplifier PC boards.

    It doesn't matter how "high-end" an amp builder is if they're using a fundamentally flawed concept.

    Sockets are being mounted on PC boards that are not designed for repeated pushing and pulling. Having the boards bolted to a chassis provides flexing assistance but it doesn't get to the root of problem, which is the strain put on the solder joints.

    EDIT: I will say that vintage PC Ampegs are the one outlier. My 1969 Gemini has MASSIVE traces, and it would be difficult to damage.
     
    Last edited: May 8, 2016
    swiveltung and DeaconBlues like this.

Share This Page

  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.
    Dismiss Notice