1. Win a Broadcaster or one of 3 Teles! The annual Supporting Member Giveaway is on. To enter Click Here. To see all the prizes and full details Click Here. To view the thread about the giveaway Click Here.

120hz hummm w/a buzzzzz

Discussion in 'Shock Brother's DIY Amps' started by Jewellworks, Jan 22, 2021.

  1. Jewellworks

    Jewellworks Tele-Meister

    Posts:
    321
    Joined:
    Mar 2, 2020
    Location:
    Orlando
    i finally got this Vintage style amp built, and im hitting all my voltages as designed. however, it has a really bad humm and a buzz throughout.

    if i turn the volume /gain knob all the way down, there is a steady 120hz humm. i took my meter, set it to HZ and measured at the end of the speaker jack cable. 120hz. to me, that could be a tube?

    then there is a buzz that comes on top if it when i turn the volume up. its a lot harder to read on my Hz meter, but it steadys out around 1k. but that could be anything since its not a single frequency.

    when building this, i got some of my cathode resistors badly wrong, and sent the plate dissipation WAY HIGH on 2 of the tubes. i could have easily damaged 1, if not both of them.

    the filament wires are grounded at the center tap, along with the negative leads of the filter caps. i know i was getting this same buzzing when i built another amp that didnt have a center tap, and i had to install a virtual c/t and lift if by about 200 ohms. not sure if this applies here.

    any help would be most appreciated
     
  2. Nickfl

    Nickfl Friend of Leo's

    Posts:
    2,490
    Joined:
    May 24, 2016
    Location:
    Florida
    Move the heater center tap to the 6L6 cathode resistor. If the hum is heater related, that should help.

    That being said, it really would be good to have extra tubes to swap in. You don't want to spend days chasing down a hum only to find out it's just a noisy tube, trust me...
     
    Jewellworks and King Fan like this.
  3. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    6,154
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    For sure tube swaps. But that would be too simple, right?

    One simple test to see if it's heater hum is to disconnect the fils from the PT and run on a 6V battery.

    upload_2021-1-22_9-22-50.jpeg
     
  4. Jewellworks

    Jewellworks Tele-Meister

    Posts:
    321
    Joined:
    Mar 2, 2020
    Location:
    Orlando
    all great ideas ill try tonight. i have replacement tubes on the way.
    one other thing i should have mentioned, all my tubes (except the 6L6) are metal canister tubes. i have the canister (pin 1) tied to the cathodes. should that go directly to ground? i thought i read somewhere that by tying to the cathode, the canister acts as an antenna, and not a shield??
     
  5. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    6,154
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    Is the bias cathode or fixed?
     
  6. Lowerleftcoast

    Lowerleftcoast Friend of Leo's

    Posts:
    2,918
    Joined:
    Dec 30, 2019
    Location:
    california
    The metal canister tubes would use the canister as a shield. It should be tied to ground, not a higher potential.
     
    King Fan and Jewellworks like this.
  7. Jewellworks

    Jewellworks Tele-Meister

    Posts:
    321
    Joined:
    Mar 2, 2020
    Location:
    Orlando
    V1 is Grid leak biased, and therefore, no cathode resistor. so canister shield is tied to cathode, which is direct to ground.
    V2 is cathode biased. currently, canister shield is tied to the cathode w a 560ohm resistor. i can move that directly to ground
     
  8. Lowerleftcoast

    Lowerleftcoast Friend of Leo's

    Posts:
    2,918
    Joined:
    Dec 30, 2019
    Location:
    california
    Why are there no pictures of the progress of this vintage style amp?

    The ground scheme is questionable if it is as shown on the layout on post #105 of the ["Vintage build???" Shock Bros.] tread.

    How close is the reservoir cap to the rectifier tube?

    Why are the 15uF filter caps in parallel?

    Is the power cord ground on the same bolt with the center taps?
     
  9. Jewellworks

    Jewellworks Tele-Meister

    Posts:
    321
    Joined:
    Mar 2, 2020
    Location:
    Orlando
    its taken me months to gitterdun. once i finally had the chassis ready, the rest came together pretty quick

    ???? heres the latest, although, not much changed... cept i got the resistor values correct and the grid leak bias corrected, as mentioned above.

    what concerns me is the running of the filaments, not shown on this dwg... they hit the pilot light first, near the top of the dwg, behind the Vol pot, then wrap around the left side to the 6L6, and jump from there. i figured it was better than running it under my tone components (dead center of dwg). and my input jack is in the center, and not far far away from the PT and rectifier.

    basically, ive got the cathodes on the left, tone circuit in the middle, and power on the right

    Deco Amp Layout Final.jpg

    its really hard to draw it in the layout and see everything else, but its right on the pin. along with the 620 ohm resistor and lead feeding the OT. in this drawing, it has a lead to the tag strip, but i didnt do it that way. i put it right on the pin, elevated about a half inch or more, so nothing touches.

    dont i need a filter cap for each plate? if i can delete one, maybe change the 1, 15uf
    to a 22 instead?

    yes.

    right on pin 8? that would mean id have to stretch the C/T all away across the chassis. its resistor is 540 ohms. i get a 19 Watt P/D with that. is that too high for an elevated C/T?
     
    Last edited: Jan 22, 2021
  10. Jewellworks

    Jewellworks Tele-Meister

    Posts:
    321
    Joined:
    Mar 2, 2020
    Location:
    Orlando
    Correction, the plate and screen of the 6SJ7 pentode is on one cap while the 6J5 plate is on the other. If this isn't necessary I'll gladly delete it.
     
  11. Lowerleftcoast

    Lowerleftcoast Friend of Leo's

    Posts:
    2,918
    Joined:
    Dec 30, 2019
    Location:
    california
    The two 15uF are set up in parallel. It is the equivalent of a 30uF cap for just one node. It services both the 6J5 and the 6SJ7 plates. It would be better to place a resistor between the two caps for an additional RC filter. Each tube would then have a filter node for each plate. The resistor should be 4.7k to 10k.
     
    Nickfl likes this.
  12. Jewellworks

    Jewellworks Tele-Meister

    Posts:
    321
    Joined:
    Mar 2, 2020
    Location:
    Orlando
    @Lowerleftcoast what does that do to the voltage I need at the plate and screen of V1?
    What about all the other questions you asked?
    Making adjustments now
     
  13. Lowerleftcoast

    Lowerleftcoast Friend of Leo's

    Posts:
    2,918
    Joined:
    Dec 30, 2019
    Location:
    california
    Having a look at the filtering for the circuit, the screen filter node (620R / 22uF C) is only providing about 20db attenuation. The next node (4k R / 30uF C) provides about 38db attenuation. This could be the reason for some of the 120Hz hum.
    It will lower the voltage.
    The reservoir cap may be too close to the tube. iirc there should be a min of about 3cm.

    Sometimes the power cord ground (wall noise) will interfere when the amp grounds are tied with it.

    You have the filter nodes ground on the power amp side of the amp.
     
    Last edited: Jan 22, 2021
  14. Lowerleftcoast

    Lowerleftcoast Friend of Leo's

    Posts:
    2,918
    Joined:
    Dec 30, 2019
    Location:
    california
    By contrast the 5F1 attenuates 36db and 42db at those filter nodes and that is with 8uF caps.
     
  15. Nickfl

    Nickfl Friend of Leo's

    Posts:
    2,490
    Joined:
    May 24, 2016
    Location:
    Florida
    I wouldn't worry about stretching that center tap across the chassis, I would guess that elevating your heater voltage will outweigh the downside from the logger wire run. The size of the resistor and the dissipation don't matter, it's the cathode voltage that is providing the elevation.

    Your layout isn't ideal, you've got some big ground loops in you power section and you're using the chassis for current return from your cathodes to the filter caps and as @Lowerleftcoast said you need a dropping resistor between the preamp filter caps. I'd also probably up that 620 ohm resistor between the 6l6 plate and screen to 1k.
     
  16. Jewellworks

    Jewellworks Tele-Meister

    Posts:
    321
    Joined:
    Mar 2, 2020
    Location:
    Orlando
    Well.... ... I'm baffled by this. What is being attenuated? And how are you coming up with this?

    ... ground loops?? You can't use the chassis as a return? Would it be better if I didn't have 2 buss bars, and just 1 for both the cathodes and filter caps?

    I can raise the 1 filter cap further away from the rectifier. But it sounds like I've got a lot more to do overall, and no clue what my layout problems are, or how to address it
     
  17. Lowerleftcoast

    Lowerleftcoast Friend of Leo's

    Posts:
    2,918
    Joined:
    Dec 30, 2019
    Location:
    california
    The power supply smoothing filters in this amp are RC filters. The RC filters attenuate the 120Hz which is created by the rectifier. The R-C filter rolls off the frequency response at 6 dB per octave above the cutoff frequency.

    To figure the attenuation:
    Use an RC filter calculator to determine the cutoff frequency. (Do the math if you prefer.)
    Find how many octaves are between the cutoff frequency and 120Hz.
    Multiply the number of octaves by 6db.
     
    Jewellworks likes this.
  18. Nickfl

    Nickfl Friend of Leo's

    Posts:
    2,490
    Joined:
    May 24, 2016
    Location:
    Florida
    You can use the chassis as a return it's just not ideal. Most vintage amps did it to some degree or another but they're not known for having a low noise floor.

    The way I think of it is that the current provided from the positive side of each filter cap has to make its way back to the negative side of that same cap and the smaller you can make that loop the better it will be noise-wise. By using the chassis as part of that return path you are greatly increasing the area of that loop.

    Ideally you should connect those ground buses into one and only reference them to the chassis in one place, such that the chassis is not used as part of the grounding system but is only referenced to it. The input jack is often chosen as that reference point for convenience. That is the ideal situation though in a lot of builds all of the jacks will be grounded to the chassis and that's usually not an issue because the chassis is not being used as a major current return and any ground loops caused between for instance the input jacks are usually very small because they are physically close together.

    All of that being said... I don't think you need to fully rebuild your grounding scheme quite yet. This is a pretty low gain old fashioned design and as mentioned many of those traditionally did similar things and got away with it. I suspect you problem is more likely a bad tube and/or the power supply design issues (the lack of dropping resistors/insufficient filtering) mentioned by @Lowerleftcoast . Address that and try new tubes first since that is much easier to do, then worry about your grounding if you still have noise issues.
     
    dan40, Jewellworks and Lowerleftcoast like this.
  19. Jewellworks

    Jewellworks Tele-Meister

    Posts:
    321
    Joined:
    Mar 2, 2020
    Location:
    Orlando
    You've given me a lot to re-think, and a lot to re-do. Thanks for your help. I'll keep you posted
     
    Lowerleftcoast likes this.
  20. Jewellworks

    Jewellworks Tele-Meister

    Posts:
    321
    Joined:
    Mar 2, 2020
    Location:
    Orlando
    ive redesigned the layout, with several changes.

    the input jack, and output jack have been swapped. input jack is now on the far left, as far away from the PT and Rectifier/power section as possible.
    removed the Cathode ground buss bar, and made the Main Buss Bar long enough to catch everything, with a single point of contact at the chassis, using the PT bolt. its a little wavy to get around all the components, but its a piece of 12 gauge solid copper, so its pretty stiff and should hold itself in place well over time. if it does sag, itll sag away from the chassis to a wooden cabinet floor.
    moved the 1st 22uf filter cap to the tag strip in the middle. the distance is about 2 1/2", and not right on top of the rectifier anymore. moved the Node 1 resistors there as well.

    while i was at it, i removed the jumper connecting the parallel 15uf caps, and put a 4K in between. i also bumped up the 620 ohm to 1.5K. moved the filament C/T to pin 8 of the 6L6, using the cathode resistor to raise the ground.
    and reversed the leads on the volume pot.
    everything else stayed the same, but im going to have to go back and recheck my numbers and estimated voltages, so many of the component values may change. but this should be a better, basic layout.

    that said, i have 2 places im grabbing ground off the tag strip, using the chassis and not the main buss bar (tone circuit in the middle) as well as the input and output jacks. good or bad idea? if i can, ill try and get the Buss bar close enough to the tone components so i can use the buss bar, and not the tag strip/chassis as a ground. but, just asking... good or bad if i left it on the tag strip/chassis?

    Deco Layout V5.jpg

    i found an RC calculator on line, and
    1.5K/22uf is about -30db at 120hz (compared to 620/22uf = -22)
    4k/15uf is about -34db at 120hz

    im assuming these roll off's are additive? not just the 1st node?
    my OT resistance is 240, so that comes to -12 or so at 120hz

    if it IS additive, then i should get
    -12 + -30 + -34 + -34 = -110db at 120hz. that sound like an awful lot to me... is it?

    please take a look and tell me if this is better.

    thanks
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.