DIY B15N - Noise Issue

[R.G.]

Hmmm. What happens if you temporarily un-bolt the input jack from the chassis, just dangle it in mid air? Does your ohmmeter now show the chassis open circuit from signal ground? And for a quick sound test, does the noise change at all?
With the amp off and the jack dangling mid air, does the sleeve contact on the jack show continuity to the chassis? What about to the One True Ground?

[ndawson]

Dangling in the air shows no connection to chassis. Mega hum when powered up. More of a constant hum than buzz like is the noise I'm dealing with.

With power off to the unit and the jack dangling in the air, there is continuity to the OTB and no continuity to chassis.

[R.G.]

Dangling in the air shows no connection to chassis. Mega hum when powered up. More of a constant hum than buzz like is the noise I'm dealing with.

With power off to the unit and the jack dangling in the air, there is continuity to the OTB and no continuity to chassis.

Can you run a wire from the first filter cap negative to the chassis (somewhere, anywhere) and repeat the test?

Added: For isolated input jacks, I really like using a 3/8" hole ring terminal like one of these:
https://www.mouser.com/datasheet/2/279/ ... 075478.pdf
with no insulating sleeve. The idea is to put the ring over the jack bushing, but still isolated from the jack; plastic-nosed jacks make this easy. The ring is pressed to the chassis by the jack holding nut. Then a ceramic RF-shunt capacitor gets soldered from the sleeve contact on the jack to the ring terminal. The chassis makes for a "Faraday cage", so the RF would vastly prefer to be on the outside of the chassis and not the wires inside.

[SoulFetish]

The buzz does goes away when pulling first and only preamp tube. Still a very ever so slight hum in the background but that's more than likely ground related. The main noise is gone. I've only got three 6sl7 type tubes in my possession so I've got one in the preamp stage and one as the phase inverter. I can't imagine they are all three bad. I've rotated them all around with the same noise being produced in any and all orders. I only pulled the preamp tube and left the other 6sl7 phase inverter tube in place for the test. Maybe I should order some know working preamp tubes.

Thanks for the tip/s.

Hmm.. 3 tubes should be a pretty good sample size, but I have seen those octals get noisy over time.
After you try RG’s suggestion, here’s another tip to try and isolate the source of the noise.
This involves probing around in your live amp, so please follow best safety protocols.
Make sure a probe using a .22uF/630V film capacitor:
Connect one end of the capacitor to ground and touch/connect the other end of the capacitor to the pins of your V1 6SL7, one at a time. Start with the input stage cathode, then grid, then plate. Then the second stage cathode, grid, and plate.
Usually, you will hear the noise drop out like you did with tube pulled with the cap connected to one of those. It will usually be as you probe the grids or plates, sometime both, but usually one will be quieter than the other.
So, you can practically make this probe many ways. But I like to use ez mini hooks with silicone high voltage wire connected to each lead of the capacitor. I would heat shrink any exposed lead or conductor to insulate you from a potential exposure and shock from high voltages

[SoulFetish]

Dangling in the air shows no connection to chassis. Mega hum when powered up. More of a constant hum than buzz like is the noise I'm dealing with.

With power off to the unit and the jack dangling in the air, there is continuity to the OTB and no continuity to chassis.

So you use your input jack as your sole ground connection to the chassis? Do you use a tooth washer giving you a solid connection from your input jack sleeve to the chassis?
It didn’t look like it from the picture, so if not, that could be your issue. If your aren’t going to isolate your input jack, a tight connection using an internal tooth washer is a must. But, even so, those connections can become loose overtime and even if you wanted to run your ground to chassis connection to the input, then I would use a separate connection as close as possible.

[R.G.]

Dangling in the air shows no connection to chassis. Mega hum when powered up. More of a constant hum than buzz like is the noise I'm dealing with.

With power off to the unit and the jack dangling in the air, there is continuity to the OTB and no continuity to chassis.

So you use your input jack as your sole ground connection to the chassis? Do you use a tooth washer giving you a solid connection from your input jack sleeve to the chassis?
It didn’t look like it from the picture, so if not, that could be your issue. If your aren’t going to isolate your input jack, a tight connection using an internal tooth washer is a must. But, even so, those connections can become loose overtime and even if you wanted to run your ground to chassis connection to the input, then I would use a separate connection as close as possible.

Agree. I'd prefer to see isolated jacks with ceramic shunt caps for RF and the one chassis connection be from the first filter cap negative. That's not the only way to do it, but it's predictable.

[ndawson]

Attached the first cap negative to chassis right near the OTG and no change in noise. Still present.

Here's the results from using the probe and connecting one side to ground and the other to the first 6 pins of V1.

Grid #2 (pin1) - noise lowers
Plate #2 (pin2) - noise disappears
Cathode #2 (pin3) noise increases

Grid #1 (pin4) - noise lowers
Plate #1 (pin5) - noise disappears
Cathode #1 (pin6) - noise increases

[ndawson]

So is that capacitor to ground probe I was using earlier acting as some sort of voltage divider to ground, reducing the plate voltage or eliminating it altogether? I didn't have a free hand to check the plate voltage at the preamp tube and I didn't want to fix the probe to the plate pin as to not damage my few 6sl7's that I've got.

Or it's acting as a filter of sorts to eliminate the noise, some sort of bypass cap.

I appreciate all of the tips here. I know it's a process of elimination, etc... to solve these issues. Thanks for your time and patience.

[R.G.]

So is that capacitor to ground probe I was using earlier acting as some sort of voltage divider to ground, reducing the plate voltage or eliminating it altogether? I didn't have a free hand to check the plate voltage at the preamp tube and I didn't want to fix the probe to the plate pin as to not damage my few 6sl7's that I've got.

Or it's acting as a filter of sorts to eliminate the noise, some sort of bypass cap.

I appreciate all of the tips here. I know it's a process of elimination, etc... to solve these issues. Thanks for your time and patience.

Yes, it's a low impedance to ground. Being a capacitor, it does not affect the DC conditions at all. It's very unlikely to hurt your tubes. But for AC...

A 220nF cap has an impedance of 603(ish) ohms at 120Hz; 60.3 ohms at 1200 Hz, 6.03 ohms at 12kHz, etc. Using the cap from ground to cathode "shorts" the cathode to ground for AC, increasing AC gain. The cap from plate to ground "shorts" the AC signal at the plate and dramatically lowers gain because it's effectively in parallel with a much bigger plate resistor.

It also AC shorts the grid to ground. If signal is coming into the grid, it ought to get whacked down, a lot. Actually, it's making the tube section into a common-grid amplifier. In this mode, any signal going into the cathode is transferred through at about equal current at the grid, giving a voltage gain of the plate resistor divided by the cathode resistor. This could account for the noise only decreasing, not stopping when the probe is on the grid.

That suggests that things are not right on the grounding side of things again. I really wish you could scope this thing. One explanation is the circuit managing to oscillate at low RF and "detect" the B+ ripple, de modulating it down into audio. I've seen exactly that happen.

Just so I'm not missing something, do the 0 Volt ends of the 5.6k cathode resistors in fact show lowest-ohms-you-can-measure to the first filter cap negative?
If so what happens if you connect that 0.22uF cap from the B+ to the "ground end" of a cathode resistor?
Just musing on this, is is possible that you have a high impedance wiper on the tone or volume pots? The second triode looks like it gets its 0V bias through that path, so a high impedance wiper could be fouling things up. Maybe.
What happens if you just shunt the grids, one at a time, to the ground end of the cathode resistor for that tube? DC voltage for the grid is zero, so shorting it to 0V will not change bias, but will massively affect common-cathode gain.

[ndawson]

I decided to order a oscilloscope and it will arrive today so once I get it and figure it out, I'll have a go at scoping the amp.

As this is my first rodeo using a scope, what would be a good order of troubleshooting/pinpointing this noise?

Thanks.

[ndawson]

Well my cheap little scope/generator off of the web arrived...it's really cheap but worth a shot. If it doesn't work for testing the amp, oh well.

I'm just wanting to find out if I need to make some kind of 'dummy load for the speaker output' for my particular noise issue. I'd need to go get some high watt resistors if so.

[didit]

Well my cheap little scope/generator off of the web arrived...it's really cheap but worth a shot. If it doesn't work for testing the amp, oh well.

I'm just wanting to find out if I need to make some kind of 'dummy load for the speaker output' for my particular noise issue. I'd need to go get some high watt resistors if so.

Your speaker works. Annoying to listen to but that's a choice. A scope is invaluable so work it 'til it's in scope.

..

[SoulFetish]

Just musing on this, is is possible that you have a high impedance wiper on the tone or volume pots? The second triode looks like it gets its 0V bias through that path, so a high impedance wiper could be fouling things up. Maybe.
What happens if you just shunt the grids, one at a time, to the ground end of the cathode resistor for that tube? DC voltage for the grid is zero, so shorting it to 0V will not change bias, but will massively affect common-cathode gain.

I was wondering g about those as well when I looked at the second tube stage. I’ve witnessed this kind of noise when a preamp tube looses its grid bias reference. Usually an open resistor, broken lead, or mistake/oversight by the builder.

[Stevem]

Here's another easy thing to try out that up to this moment i forgot about so I apologize.

Get a scrap of card board big enough to cover the whole bottom open side of the chassis.

Next cover that with a sheet of aluminum foil taped to it .

If you use more then one section of foil be sure they are grounded to each other.

Ground that foil to the chassis with a clip lead and then set it on the chassis.

Does doing this help any with the noise condition?

If you have a sheet of steel big enough or metal door screening you can just use that also.

[R.G.]

Well my cheap little scope/generator off of the web arrived...it's really cheap but worth a shot. If it doesn't work for testing the amp, oh well.

I'm just wanting to find out if I need to make some kind of 'dummy load for the speaker output' for my particular noise issue. I'd need to go get some high watt resistors if so.

What are the specs on your scope? You may have to use your 220nF cap in series with the scope probe, because some o'scopes can't stand the DC voltages in tube stages. I was speculating that a scope would make it easy to see if you have RF riding along with the audible signal.

As didit says, use your speakers while you test, maybe with some ear plugs or ear protectors.

Stevem's point is also good. A quick test with a metallic sheet over the open side of the chassis would also help rule out RF coming into the open side.

I should be clearer. From the back and forth in the thread, I think you've isolated the issues into the first stages, and they're just sending it along into the PI and power tubes. I'm kind of off the idea that the input stages are being driven mad by ground-wire feedback from the output tube and/or speaker ground currents, as you have those moved off the ground connections to the input stages.