My thinking is there is a ground loop somewhere with the volume pot. The way I wired the volume pot was by tying the "bottom" terminals (the ones that go to ground) of the pot together and then grounding that whole node altogether. Should I separate the grounds out? Didn't think this would be a problem but as I'm unable to track the buzzing down I'm kind of just guessing now.
Ok. I just tried separating the grounds as I was thinking in my above post and that didn't seem to help with anything. Now I'm drawing a blank.
If the buzz is in the middle of the volume pot thats where the grid input is at its highest impedance. If there's a lot of treble in the buzz then is pickup is capacitive onto the grid of the the first stage. Check heater wiring etc. Are you screened from the centre of the pot? Is the pot body earthed, is the chassis earth and you signal ground linked? It could also be your 1M is earthed to the wrong place (ie ground loop but less likely). Other options are using a 22k pot. Usually ground loops will generate hum and lower order harmonics so it sounds different from a buzz.
Last edited:
The buzz does sound like it has some treble in it. It's definitely what I would call buzzing and not hum. I can hear some 60Hz hum itself but that doesn't bother me and it's a good amount deeper than the buzz.
I have the input jacks grounded in 2 places. One is to the chassis ground, this is the ONLY place where the circuit touches the chassis. The chassis is then connected to earth ground (3rd wire on IEC inlet). The second input jack ground is for the signal going to the main circuit ground. And the main circuit ground is the bottom of the rectifier. The 1M grid resistor and bottom side of the pot is connected to this ground as well.
Not sure what you mean exactly by screened from the center of the pot... If you're asking if I've got the grid connected to the center (wiper) terminal of the pot then yes.
I thought it might be heater wiring related but then wouldn't I get buzz in both channels? The buzz is only in one channel that's why I wasn't thinking it was power supply related. Would it be worthwhile to break any DC connection the input jacks have to the 6SN7 grids with an input capacitor? If so what type / value would be good as an audio coupling capacitor in this part of the circuit? I want to say a high capacitance since my instinct is that would make sure the signal can flow without much impedance in the cap, but in many circuits I've seen the input capacitor usually seems like a low value (or it could just by my luck with which circuits I've actually seen).
EDIT: Technically speaking in terms of exactly how I wired it, the main circuit ground is connected to the negative terminal of the 3000uF filter cap.
Last edited:
Sorry run screened cable from the centre of the pot to the grids. Could you take some photos. The input cap will just block any DC not affect the buzz.
Last edited:
I'll post some photos later in the day, I gotta get to sleep it's quite late here for me hehe! Also I just remembered that the resistive divider for setting the voltage regulator is grounded to the chassis as well, but I don't think this would matter?
One thing to note though in case it matters is the buzzing stops immediately when cutting the power to the amp even though the tubes are still conducting as the caps drain and filaments cool. For example if I have no audio playing and I run the volume high enough that I can quite audibly here some RF being picked up but not so high that the buzz is gone, when I kill the power the buzzing stops immediately but I can still here the RF noise for however long it takes for the caps to drain / filaments to cool.
The ground current from C1 and C2 go back to the rectifier and transformer. That is where the hum currents flow. So you should on have on point on this node to take the rest of the ground from. Try and use this as your star ground point. The RF noise may need a small cap (22p) across the 100k pot. The other possibility is you have some HF oscillation. That can sometimes produce buzzes too and demodulate AM radio. The grid stoppers on the first stage are very low maybe change 33R for 2K2. You don't have any at all into the pentode (again maybe 2K2).
Definitely you need about 100R in series with C5 and C6. All these will stop oscillation. Maybe your buzzy side is oscillating.
Definitely you need about 100R in series with C5 and C6. All these will stop oscillation. Maybe your buzzy side is oscillating.
Last edited:
Well I'm really stumped now. I ended up not get any pictures as I spent my time tinkering with the amp instead.
Here's what I tried:
Installed 2.2k grid stop resistors on each one of the 6SN7 grids. If it had any effect, it was very subtle.
Tried temporarily converting the filament supply to the 6SN7 to DC. This consisted of a bridge rectifier into a 5000uF filter cap, then into an adjusted high current linear regulator that regulated the voltage down to about 6.1VDC. This had little to no effect on the buzz. It did however make the deeper 60Hz hum seem quieter if I had the volume pot all the way up (where there is no buzz anyways, buzz only in the middle (ish) of the pot sweep).
I flipped the 6SN7 outputs to the opposite 6CA7 inputs to essentially switch which power tube the 6SN7 plates were sending the signal to. The buzz switched ears so this tells me it's source isn't in the power amp.
Tried moving the grounds around as well. No effect.
Tried rolling preamp tubes around. This seemed to affect the buzz a little bit but it wasn't night and day difference. For reference the 6SN7 I'm using is an NOS (well it was when I got it, has some hours of use on it now) GE 6SN7GTB.
I thought it could be because some of the output caps are very close (just right under) the 6SN7 socket. So I pulled it out of the amp completely then ran external wires to remotely connect it. Had no effect.
The RF noise doesn't bother me as the volume pot has to be almost all the way up for it to be audible and even then it's quiet. My main focus is removing this buzz right now.
EDIT: I don't have any grid stoppers between the 6SN7 plates and power tube control grids as I'm more or less following the design of the DV336SE / Bottlehead Crack design with some alterations (mainly using the different output tubes) and neither of those designs use anything but the wire between the 6SN7 plates and power tube grids.
Here's what I tried:
Installed 2.2k grid stop resistors on each one of the 6SN7 grids. If it had any effect, it was very subtle.
Tried temporarily converting the filament supply to the 6SN7 to DC. This consisted of a bridge rectifier into a 5000uF filter cap, then into an adjusted high current linear regulator that regulated the voltage down to about 6.1VDC. This had little to no effect on the buzz. It did however make the deeper 60Hz hum seem quieter if I had the volume pot all the way up (where there is no buzz anyways, buzz only in the middle (ish) of the pot sweep).
I flipped the 6SN7 outputs to the opposite 6CA7 inputs to essentially switch which power tube the 6SN7 plates were sending the signal to. The buzz switched ears so this tells me it's source isn't in the power amp.
Tried moving the grounds around as well. No effect.
Tried rolling preamp tubes around. This seemed to affect the buzz a little bit but it wasn't night and day difference. For reference the 6SN7 I'm using is an NOS (well it was when I got it, has some hours of use on it now) GE 6SN7GTB.
I thought it could be because some of the output caps are very close (just right under) the 6SN7 socket. So I pulled it out of the amp completely then ran external wires to remotely connect it. Had no effect.
The RF noise doesn't bother me as the volume pot has to be almost all the way up for it to be audible and even then it's quiet. My main focus is removing this buzz right now.
EDIT: I don't have any grid stoppers between the 6SN7 plates and power tube control grids as I'm more or less following the design of the DV336SE / Bottlehead Crack design with some alterations (mainly using the different output tubes) and neither of those designs use anything but the wire between the 6SN7 plates and power tube grids.
Could a possible cause be having the input jacks grounds tied together? I have 2 1/4 inch mono inputs (one for each channel) and I tied the ground of those together then I'm just running three wires (2 signal 1 ground) from the jacks braided together to the volume pot. Volume pot is dual pot design and is an Alps Blue pot with audio taper.
I think you will need to post some photos. Could you send a scope picture of the buzz or failing that a .wav file.
I have an old (like 1970s old) scope I can set up and see what it shows me.
I can confirm that shorting out the 1M resistor on the buzzy channel kills the buzz. What still perplexes me is how that one channel is quite buzzy while the other one is pretty quiet.
Me too!
Maybe some photos might point to the issue. Does point to capacitive pick up.
Maybe if the 6v3 is floating add two 220R on each side to ground (or 50V as you also have the output devices). Try the heater on a DC supply. The heater supply if floating could pick up some signals through the mains transformer.
I notice one side heater is next to grid (pin 1 and 8) but not the other.
Maybe some photos might point to the issue. Does point to capacitive pick up.
Maybe if the 6v3 is floating add two 220R on each side to ground (or 50V as you also have the output devices). Try the heater on a DC supply. The heater supply if floating could pick up some signals through the mains transformer.
I notice one side heater is next to grid (pin 1 and 8) but not the other.
Last edited:
THIS!!! THANK YOU!!!! 🙂🙂🙂
Telling me this got me thinking... The buzzy channel grid is the pin next to the heater pin! I already tried using DC on the 6SN7 heater but it didn't make a difference. But I got to thinking maybe my little 6.3VAC to 6VDC supply had too much ripple and I could have tried using the pin next to the grid pin as the positive DC supply. So I added another 5000uF filter cap to my 6.3VAC to 6VDC supply and made absolutely sure that I connected the positive to the pin furthest from that grid pin....
And the amp is, quite literally, dead silent. No hiss, no buzzing, nothing. Even with the volume wide open.
Now I just need to build out a permanent AC to DC filament supply. I'll leave the power tubes on AC as with this test they are on AC still and it's silent. Plus it'll be way harder to keep my little voltage regulator cool flowing that much current. Right now my AC to DC 6SN7 filament supply is just mocked up on my breadboard. But the results are quite nice 😀
Thanks again for your help. Without mentioning that heater / grid pin proximity I'm not sure I would have thought of that.