Reducing noise in DIY tube amp project.

Discussion in 'The Projects Forum' started by planeguy67, Jan 29, 2015.

  1. planeguy67

    Thread Starter Member

    Jan 29, 2015
    59
    16
    Hey everybody, new guy here.

    I recently constructed a modkitsdiy.com 102 tube amp. It works just fine, however, it is a bit noisy with nothing connected and the input jack grounded. I imagine there's excessive ripple in the power supply and I ordered a few parts to quiet it down.

    Attached is a section of the schematic. As you can see, perhaps to keep costs down, the filters caps are rather small. I intend to parallel a 100 uF capacitor across the 47 uF and the 22 uF caps. I also want to replace the 100 ohm resistor with a 1.5 H choke and form an LC Pi filter. According to some Butterworth calculators I found online, this should give me some pretty good rejection above 20 Hz (basically anything audible).

    Does this modification make sound sense? My concern is mainly the choke having a 52 ohm resistance and replacing that 100 ohm resistor. Should I place a 47 ohm resistor in series to keep the resistance around 100 or enjoy the extra current available with the choke by itself? Would this change cause any appreciable concerns with the tube's biasing? The circuit itself is fairly conservative.

    Thanks everybody.

    P.S. - Sorry, doesn't appear I can upload an image from my phone. I'll try when I get home. In the meanwhile, just imagine a Pi filter with a 47 uF cap connected to ground, a 100 ohm resistor on top, followed by a 22 uF cap to ground.

    --------- 100 Ohm -------------
    ....|................................|
    ....|................................|
    47 uF........................22 uF
    ....|................................|
    ....|................................|
    ...V...............................V
     
  2. #12

    Expert

    Nov 30, 2010
    16,261
    6,770
    Here's the schematic.

    You can do the modifications as you imagined and not hurt anything. Just plug the choke in where the 100 ohm resistor was, but I would point out that the power supply noise is at 120 Hz and the usual method for using an amplifier that is noisy when nothing is plugged into it and the input jack grounded is to not play it with nothing plugged into it and the input jack grounded.
     
  3. planeguy67

    Thread Starter Member

    Jan 29, 2015
    59
    16
    Thanks #12. That's the schematic. Plugging a guitar in does not reduce the noise one iota. The reason I mentioned the grounded input was to rule out any hum from the guitar.

    The noise is generated within the amp itself. It's either the AC ripple in the DC supply, the power transformer too close to the output x-former (not much I can do there), inadequate shielding around the preamp tube (hard to imagine since it's housed inside an aluminum sleeve), any combination of the three, or something else I haven't thought of.

    Thanks again.
     
  4. Rory Starkweather

    Member

    Jan 27, 2015
    41
    4
    Sounds like you really know what you are doing. I will watch this thread closely, but there isn't much I can add to what you have already done.

    One question: Are you working with British ripple or US ripple? That frequency difference still irks me.

    I grew up during the days before transistors were commercially viable, and worked mainly in the UHF/VHF range, where noise caused by interelectrode capacitance was a problem, but I wouldn't think that would be a problem in the audio range.

    I suppose another remote possibility is switching noise from the power supply maybe? Do you have an Oscilloscope by any chance?
     
    planeguy67 likes this.
  5. planeguy67

    Thread Starter Member

    Jan 29, 2015
    59
    16
    US ripple. I live in the states. I'm suspecting inadequate filtering from the full-wave rectifier. I have some other mods in mind as well, but first is combatting the noise.

    Yes, I have a scope. Never used it for such high voltages before, however.
     
  6. Rory Starkweather

    Member

    Jan 27, 2015
    41
    4
    I'm glad you mentioned the voltage. Not usually a concern any more . . . unless you work with tubes.

    PS noise is a constant issue for me. Some formula suggested that 3300 uF was really the best for US ripple. Finally found an EBay seller who has hundreds of them and would rather have cash. But they are electrolytics and and only rated for 45 volts.
     
  7. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,778
    932
    check for flourescent lights or small AC motors like fans which may share the same breaker circuit or be within 10 to 20 feet of the amp. This is a fault found mostly in radio receivers but I have come across it in audio amps as well.
     
    planeguy67 likes this.
  8. Rory Starkweather

    Member

    Jan 27, 2015
    41
    4
    Well said.
     
  9. MrChips

    Moderator

    Oct 2, 2009
    12,432
    3,360
    I assume you bought this as a kit and you followed the assembly instructions.
    The reason I ask is because for a DIY construction, things like the placement of the output transformer and its orientation can have an effect.

    As a test, you can remove the output transformer, extend the connections and move it away from the rest of the circuit and see if that makes a difference.

    If that fails to eliminate the hum, the next test is to do the same with the power transformer.

    I would also go over the wiring and see if you have missed any connections.
     
  10. #12

    Expert

    Nov 30, 2010
    16,261
    6,770
    I need to shoot that formula in the head.

    Every power supply is designed for its load. Every design needs different sizes. I believe planeguy is going way overboard, but he didn't ask my opinion about that.

    As for the real problem, finding hum sources can take all day. I rebuilt a bass guitar once, and it could pick up a fluorescent light at 50 feet. The cure was probably to get the owner to buy humbuckers, but he didn't.
     
  11. MrChips

    Moderator

    Oct 2, 2009
    12,432
    3,360
    Is the hum still there when the volume control is turned all the way to minimum?

    If it is, unsolder one end of the 0.022μF capacitor that connects between the anode of V1 and the input of V2 and see if the hum goes away.
     
    #12 likes this.
  12. #12

    Expert

    Nov 30, 2010
    16,261
    6,770
    I didn't comment on the filament connection to the cathode of V2 because I don't understand why anybody would design an amplifier like that.
     
  13. MrChips

    Moderator

    Oct 2, 2009
    12,432
    3,360
    I see what you mean if you are referring to the way they connect the filament to the cathode on V2.

    The other thing the OP can try is to swap the GREEN wires from the power transformer that supplies the filaments. Just a wild idea.

    Edit: We cross posted.
     
  14. Rory Starkweather

    Member

    Jan 27, 2015
    41
    4
    That is something that a lot of people say, but, as is true with most things that 'a lot of people' say, it has nothing to do with reality.

    The majority of power supplies are built on the 'black box' principle, then sold as part of a 'lowest bidder' package. The only person who suffers is the consumer. It's the American way. Who cares about about the American consumer?

    If you are not willing to look at what you get in the way of a power supply, you deserve the results. Your piece of equipment sounds bad? Whose fault is that? Look in the mirror before you start pointing fingers.

    Another option is the 'homebuilt'. Unfortunately, most home builts are made by people who know nothing about about power supplies, and don't wan't, or have the time, to learn.

    S0, if a piece of electronics gear costs less than $1000, and you didn't build it yourself, you're kind of stuck with something designed by someone who is not as smart as you are.

    The parameters for PS filtering are well known and promulgated. Check them out. I build reference quality equipment. I don't buy power supplies. I make them.
     
  15. planeguy67

    Thread Starter Member

    Jan 29, 2015
    59
    16
    I looked at some very large filter caps (> 1000 uF), but at 350 volts or more, they were the size of a soda can. Too much for this project.

    Good point. There is only one fluorescent in our home (a small utility light in the garage), but nearly everything else in the house is incandescent and on dimmers. Still, the noise is there with everything off. It's not a rare complaint with this kit.

    Thanks, but after triple-checking everything before the initial "smoke", everything looked good and the amp does work, it's just a bit noisy. There have been other complaints about the hum in this kit. I'm not the first. The power and output transformer are mounted fairly close to each other because the chassis is small, but they are oriented at 90 degrees from each other. Another idea would be rotate one down another 90 degrees (on its face) such as done on my Blackstar amp, but I'm not sure how much difference that will make.

    I'm known for using a sledgehammer on a thumb tack. My reasons for building this kit in the first place was to get more familiar with tubes as I didn't grow up with them. I have other amps for gigging and this amp is really more of a toy for education and experimentation. I'm very curious to see if adding the LC Pi filter will quiet it down. If not, ok, I'm just out $25 for some parts.

    The pickups aren't an issue. As I originally stated, the hum is present with the input disconnected, which immediately grounds the grid on V1. And the only fluorescent we have is a small tube in the garage. Still, I'll shut it off the next time I fire up this amp and see what effect, if any, that has.

    No. The hum disappears when the volume gets turned to zero. It's just bearable at around 2 (which is fairly loud) and truly annoying above that.

    I've been wondering about that as well. It's the only part of the amp I really don't understand. I thought the filament circuit was completely independent of the signal chain. Perhaps the designers are trying to strap one side of the filament power to a reference voltage (why not ground it?) and forcing the other end to swing? I would disconnect it, but I'm sure it's there for a reason.

    If it's any help, here are photos of my kit. I know I promised the schematic as well, but #12 took care of that. Thanks again.
    IMG_1657.JPG
    View attachment 79668
     
  16. #12

    Expert

    Nov 30, 2010
    16,261
    6,770
    That looks like a very good layout with the transformers at right angles to each other and a shield on the preamp.

    and I wasn't talking about YOUR guitar. I was talking about how difficult it can be to find a hum problem.
    This might come down to how you routed the wires. Can you turn it belly side up and make a photo?
     
  17. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,780
    1,103
    Have you used twisted-pair wiring to the tube heaters and kept that wiring spaced away from amp input connections?
     
  18. planeguy67

    Thread Starter Member

    Jan 29, 2015
    59
    16
    Check the other photo I uploaded. For some reason it wouldn't display full-size.
     
  19. #12

    Expert

    Nov 30, 2010
    16,261
    6,770
    Oops. My bad. I missed that one.

    Red/yellow...current induction loop to the tone control.
    Tuck that wire away from the signal path.

    At first look, that is about the most beautiful job of wiring.
    This might not be easy. :(

    Note: You said the hum goes up with the volume control. This seems to narrow it down to stage 1 and the tone control. I say, "seems to" because I've been fooled so many times that I have to act cowardly.
     
    Last edited: Jan 30, 2015
  20. planeguy67

    Thread Starter Member

    Jan 29, 2015
    59
    16
    Good eye, #12. That red/yellow wire is the power transformer's center tap. I left it long because I intend to install a "Standby" switch on the front panel. I could try moving it away from the tone circuit, but I don't think it'll do much good. That wire is connected directly to ground.

    As far as the origin of the noise, yes, the evidence points to the first stage. The modkits web site points to a faulty/less than ideal pre-amp tube as the cause of filament current amplification. I may try a tube swap. I'm also going to test the noise and see if it's 60 or 120 Hz hum. That might help pin-point it, maybe?
     
Loading...