Massive grounding issues, any ideas?

[Diverted]

HI all,

I resisted asking for opinions/help as I wanted to figure this out on my own. But I’m completely stumped and frustrated.
I am building a very simple Champ 5F1 (no tone control) in an old metal box (late 40s car radio) that a friend of mine asked me to fill with a circuit. The champ is the only thing that will fit. I built the circuit in half an old dead reverb tank, running the volume control out from the chassis to the radio body itself.
I was getting some hum and some instability (volume control not working at times, at other times it did, at other times silence, at other times massive hum, volume cutting in and out etc. It was mostly dependent on whether I had the chassis mounted in the box. The circuit is correct, connection-wise, I’m positive of it. But the running wires/lead dress is what’s causing all the issues.

I took it apart, tried to reconfigure everything to make it more stable. Nothing. Still massive instability. I tried everything: Using shielded for V1A grid, V1B grid, grounding one end, or the other, or both, to a single spot in the chassis. Nothing seems to work.

Finally out of desperation I just decided to cut the volume control out completely, so I ran the coupling cap from V1 plate directly into V1B grid. I thought that would give it a steady, not adjustable signal. But nothing.Now dead silence.

For the sake of argument (and to keep my sanity) if I were to ditch the volume control altogether, how would I do it properly? It looks it would be just a matter of taking the vol control out and hooking up the coupling cap out of V1A plate (that originally went to high side of pot) directly into V1B grid (which was originally fed by wiper of pot), but that apparently is not right. I also tried running that coupling cap into a 470K resistor to simulate the volume (1M pot) half way, and running that resistor into V1B grid. Again, nothing but silence.

What am I doing wrong? For such a simple circuit this is absolutely crushing me.

Thanks for any guidance!

[Dave]

Pics would help.


The 470k wouldn't simulate a pot at half way unless there was another 470k going to ground. the junction of the two 470k's would be equivalent to the wiper of the pretend pot.


Dave

[Diverted]

Here’s a photo. It should be pretty easy to see everything going on with the circuit, there’s not that much there. But it’s killing me!

[Dave]

Do you have a center tap (or an artificial one) on those heater leads? That would cause a lot of noise.


Dave

[Diverted]

Yes, they’re the tan Dales at bottom of pic, off the 6V6 (center tube). I have them terminating at the 6V6 cathode instead of ground.

[Diverted]

In the pic, the coupling cap from V1A plate is running directly into V1B grid.

[Diverted]

At this point, I’m done trying to run wires out of the chassis and mount them on the cabinet body. I would love to be able to find a way to just bypass that volume pot and have it a constant level, and use the guitar volume to adjust amp volume.
If my solution (V1A plate coupling cap to V1B grid) is wrong, how would I do it correctly?
Schematic attached, showing how I have it wired without volume control.

[Dave]

V1B grid gets another 470k to ground. Then you have a pretend 1 meg pot somewhere around half way up on volume. Although not exactly because the pot would be tapered, but close enough.


Dave

[Diverted]

Thanks Dave. How about if I wanted to simulate the volume control being pretty much all the way up? 1M to ground of V1B grid?

[Dave]

Yeah, your signal lead would have little to no resistance and the grid to ground would be the entire meg.


Dave

[Diverted]

OK, got it... scratch that last question. I”m going to find a resistor value that works best. 1M to ground was too overdriven, so I’ve started backing down. I’m presently at about 10K to ground and that gives a nice breakup around 7 on the guitar dial, but clean otherwise, and still plenty of volume.
Thank you so much for setting me straight on the resistor fix.  I’m just glad it’s now working SMOOTHLY. The last couple of days have been frustrating, trying to chase down a million gremlins.
I’ve never seen a guitar amp with no volume or tone control before, just on/off. But I kinda like it!

[Dave]

Sweet man... I'm glad you got it worked out.

Dave

[Diverted]

Sweet man... I'm glad you got it worked out.

Dave

Thanks to you. This is a pretty odd little amp, and getting it working smooyhly meant a huge concession in getting rid of the VC. But the amp’s kind of a novelty, and I’m going to have my friend Dan over to play around with it ... once he’s here with his guitar we’re going to close in on the best resistor to use to give him the tone/volume he wants. He basically told me he just sets the volume and forgets it when he’s playing, and adjusts volume from his guitar. So no big deal once we get the right sweet spot. Here’s a pic of the amp, thanks again!

[Diverted]

OK, I want to revisit this one more time so I can learn.
Having figured out how to wire in a set resistance/volume and bypass the volume pot, I now want to understand what was causing the issues and if there’s a simple solution, see if I can get a stable volume control.

After Dave helped with the resistance issue, I removed that coupling cap between V1 pins 1 and 7, and ran it from pin 1 to a terminal strip at the bottom right of pic (not shown in this pic).
I then took a two-conductor shielded wire, and ran one lead to coupling cap on the terminal strip, the other to pin 7, and the third (the shield) to ground. I then wired the other end: pin 1 coupling cap to high side of pot, pin 7 to wiper and low side of pot to the shield, which terminates at chassis ground. The wire is about 9 inches long.

When I turned on the amp, the same problems I was having before returned: volume control did not work now and then, at other times, depending on whether I moved it, It stayed at one level, cut out mid-range on the pot, and the same level returned as I rotated pot more. At other times I got massive hum, sometimes with guitar signal, sometimes without guitar signal, getting through. It happened whether or not I let the pot just sit on my bench, or had it touching the chassis. Bottom line is no matter what configuration I tried, I had bad results.

What is going on here? How would I correctly wire a volume pot so it will behave properly? Any insights on how to do this right would be appreciated ... as in, whether to use shielded wire, how to ground, etc etc etc because obviously, what I’m doing ain’t working.

I figured I would give it one last shot before I give up and just hard-wire/bypass the volume pot.

Thank you.

[sluckey]

You must have a resistor between pin 7 and ground. Without it, the tube cannot establish bias and operation fails.

[tubeswell]

... I removed that coupling cap between V1 pins 1 and 7, and ran it from pin 1 to a terminal strip ...
I then took a two-conductor shielded wire, and ran one lead to coupling cap on the terminal strip, the other to pin 7, and the third (the shield) to ground. I then wired the other end: pin 1 coupling cap to high side of pot, pin 7 to wiper and low side of pot to the shield, which terminates at chassis ground. The wire is about 9 inches long....

So it sounds like you ran the cable shield to ground (which will shield the cable from EM fields), but you didn't ground the ground-side of the pot.


Without the pot being grounded on one side, the grid (Pin 7) of the next triode won't have a grid-leakage path to ground for any stray charge that builds up on the grid. This 'grid-charge' occurs naturally because the grid is the next-closest electrode to the cathode, which is hot (~800 degrees) and is emitting electrons in the space charge right next to the cathode. Because the grid is so close to the cathode, it too heats up quite a bit, and releases a few electrons of its own. If there is no grid leak pathway* for the surplus charge to escape to ground (and thus neutralise the grid), the build-up of charge on the grid affects the bias voltage on the grid, and this can stop (or severely weaken) the signal going through that tube.


*The resistance of the pot provides a DC-pathway for grid leak current. If there was no resistance, all the signal from the coupling cap from the plate of the previous stage would be dumped to ground and no signal would make it to the next stage. If the resistance is too high, it will effectively be the same as 'infinite resistance' , which amounts to the grid having no grid leak path. So the grid resistance has to be a nominal resistance which will allow signal to pass, but at the same time, allows the grid leak path to function. Tube datasheets specify what the maximum recommended grid leak resistance is for the optimal functioning of the tube. For your typical 12AX7, it is 1M max. But anything up to 1M will 'work' and the lower the resistance, the more the signal becomes attenuated/reduced. A pot provides a combined variable grid leak/attenuation function (but the ground side of the pot needs to be grounded)

[Diverted]

Thanks for that explanation. There is a lot of info in there I didn’t know about. Thank you.
However the ground side of the pot (left lug of pot, looking at it from the rear with lugs facing up) was grounded... the ground wire in the shielded cable was connected to the pot on one end, and the other end of that ground wire was connected to chassis ground.

[tubeswell]

Thanks for that explanation. There is a lot of info in there I didn’t know about. Thank you.
However the ground side of the pot (left lug of pot, looking at it from the rear with lugs facing up) was grounded... the ground wire in the shielded cable was connected to the pot on one end, and the other end of that ground wire was connected to chassis ground.

Can you post a hi-res picture?