Ground Loop Hum between amp and external reverb unit

[wsscott]

I was just thinking, what if I disconnected the guitar cable's shield cable at the plug on the end that plugs into the amp?  In other words, just the tip of the guitar cable's plug going into the amp's Input jack is connected.

[sluckey]

That will cause a terrible buzz/hum.

[wsscott]

Ok, won’t do that.

[HotBluePlates]

HBP-That last tube has a 2.7k resistor, not a 100k resistor, going from the cathode to ground.  There is the Output Level pot which is a 100K pot that comes off the last plate and its wiper goes to the Output jack.

Are you talking about the 2.7k resistor that I should jump?

I skipped over the fact this build was based on the Magnatone M10 reverb, and I gave advice based on the Fender 6G15 circuit.

My thinking was you had a preamp tube prone to heater-to-cathode leakage, and when such a tube is used with a 100kΩ plate load (as with a cathode follower or split-load inverter) they will HUUUUUMMMMMMMMMMM. I was looking for a way to get you to short-out the leakage hum to verify a bum tube.

The challenge is you may not have any good way to pre-test preamp tubes to verify a new tube doesn't have leakage (and hum).  There is a simple way to do such a test with a switchable bypass cap for a high-ish-resistance cathode resistor, but not everyone already has that feature on one of their amps.

[wsscott]

I was reading an article by Jeff Gehring on the Revibe grounding issue, and his 3 methods of solving the hum problem.  One of which is using a ground buss with every ground connected to the buss and then the buss connected through the 15 ohm resistor/diode hum buster connection.  This is what I'm using, except I used a 10 ohm resistor because I didn't have 15r.

He comments that after its been installed to test to see if everything is grounded at the one ground point on the hum buster, check it with a meter by setting it to ohms, and then connecting 1 lead to a point on the chassis steel, and then check every connection going to the buss.  There should be a reading of 15 ohms at each connection.

So I get 10 ohm readings at every point EXCEPT at one of the ground connections on one of the two 100 ohm resistors for the artificial center tap for the 6.3VAC Filament supply connected at the Pilot Light.

I also have noted that neither the Pilot Light frame/mounting bracket nor the On/Off switch are isolated from the chassis.

So maybe that's the source of the problem.  I don't have an isolation washer to fit either of those components.

Thoughts?

[sluckey]

So I get 10 ohm readings at every point EXCEPT at one of the ground connections on one of the two 100 ohm resistors for the artificial center tap for the 6.3VAC Filament supply connected at the Pilot Light.

So, what resistance reading do you have? Now fix it!

I also have noted that neither the Pilot Light frame/mounting bracket nor the On/Off switch are isolated from the chassis.

Good. Neither of those need to be isolated. They are not causing any issues.

[wsscott]

The reading on one of the resistors to ground is jumping all around.  So I'll replace both of them.

I now have some 15 ohm resistors, so I'll also replace the 10 ohm that I installed in the hum buster circuit.

Hopefully this will help.

[wsscott]

Its better, but still there.

There is very little difference in the amplitude of the hum regardless of whether the reverb is engaged or not.  So that's probably a good thing.

It's affected only by the Output Level pot's setting.  If I turn it down low enough, ie. almost all the way CCW, it's totally gone, but then I'm losing the output getting into the amp.  So bringing it up to a proper output amplitude to feed into the amp is where the hum comes into play.

Thanks for everyone's help.

[sluckey]

I've been studying the couple pics you posted, and I'm convinced that your soldering skills play a BIG role in the hum issues you are having. I don't think you'll ever solve this problem.

[wsscott]

Fair comment.  This was my first build that wasn't a kit, so it definitely was a learning experience.

[acheld]

I agree with Stuckey's comment about the soldering, backed up by the "wiggling" of one of your grounds when tested.  But, I think you can make this work.  It will take a lot of work and time.

a)you have a lot of time invested already, b) there are a lot of good (eg, expensive) components in place,   c) you have some of the skills already (your buss looks great in the picture) and d) really, there's only one way to learn -- do it, make mistakes, do it again, and again . . ..

Some tips based on your photos:  1.   Don't use the cloth covered wire (yet).  It can be messy (and is in your case) and when you pull back the cloth to make a connection, the cloth slides back over your solder connection and obscures it.  Best to use standard PVC single stranded wire.  It will be neater, and you will see your mistakes.  Cloth wire has that vintage look, but remember that when Lupe or Lydia were wiring up Leo's amps, they were making maybe thousands of connections per day -- experience counts.
2.   Clean the area around your solder joints.  I see a number of solder drops/splashes in your two pictures.   They are not causing problems now, but later they certainly can when they move around your build.
3.   Your turret solder joints appear sketchy.   What I do, and this may be controversial, is to heat the turret itself just below the top lip with your components in place, and allow solder to flow into the turret.  Not too much at this point, but just enough to solder the component(s) and create a barrier inside the turret.  When that has cooled, you can come back later and by heating the top of the turret and adding solder, you can build a nice rounded blob at the top.   The point is that trying to do that all at once often heats the turret enough to cause your solder to flow right through the turret.
4.  Use an electric drill to twist your heater wires.   It is for sure more work than just soldering one end to a socket, then twisting by hand, but it is much neater.  (the twist doesn't need to be super tight).   PVC insulator is neater IMO; when I've tried this with cloth insulators, I've ended up kinking the wire.
5.   I worry when I see long runs of unsupported and uninsulated wire from resistors and capacitors.   It's one thing if you're running a resistor from one turret to the next -- that should be very safe -- but not so safe when a lot of components are close together and you have wires running here and there (as in your photo with those pretty Atom Caps).  18 or 20 gauge PTFE insulation is your friend, and not expensive.  Good thing about this type of insulator is that it does not cover up mistakes.

I don't think these tips alone will solve your hum problem.  BUT, I bet if you go back over your build and reflow every connection, make it pretty, use wire that makes it easier to spot mistakes, insulate where needed and clean up everywhere, you will be a lot closer that you might think.   If you haven't looked at Stuckey's amp pages https://sluckeyamps.com/index.htm, it is worth it.   By emulating his attention to detail, you will be a better builder. 

[wsscott]

Can an effects unit like this ReVibe mod still produce 120hz filament hum if no tubes are plugged into the amp?

[wsscott]

Sorry, I meant to say 60hz filament hum.

[acheld]

Yes, it is possible.    Even though there will be no current flowing, there will be an electrical field surrounding the heater wires.  Because there is no current, no magnetic field will be present.

[wsscott]

I thought maybe I should use a shielded cable to connect the 1M resistor to the Input lug of the Output Level pot to see if that helped.

So I disconnected the plain insulated wire between those 2 points, and replaced it temporarily with a shielded cable, that has its insulated lead connected to the Input lug on the pot, and its ground shield wire connected to the Ground Buss only.  It is not grounded at the end where the cable's insulated lead connects to the 1M resistor.

Results, no sound or signal at all at the Output jack.  If I disconnected the ground shield wire from the Buss, then the hum and signal return.  Why is that?

Should the shielded cable be grounded at both ends?

[wsscott]

Ok, solved that problem.  Part of the ground shield in the cable was touching the 1M resistor.

So the shielded cable doesn't make any difference as it's now connected, with just one end of it grounded as you would normally do.

[acheld]

Shielded cable is grounded on one end only. 

If you ground both ends, ground loops become possible.

Theoretically, it doesn't much which end is grounded.  But, since most of these are small signal wires attaching to the front panel (with signal either to or from the panel/pots/switches) it is easiest to connect to your ground bus.

[wsscott]

Thanks for your help.

I’m going to put this project away for awhile, and maybe revisit it at a later date.