Ground Loop Hum between amp and external reverb unit

[wsscott]

In 2020, with fantastic help from Sluckey and others, I built a ReVibe mod that was based on the Magnatone M10 circuit using its circuit for the Vibrato and tube reverb, but with no Output power of its own. It only has its own power transformer, and uses a solid state rectifier which is not connected to the PT's CT.  That CT is grounded at the 1st Reservoir filter cap.  The ReVibe plugs into the amp for sound and its effects.  I've been dealing with a hum all this time and it has gotten better with various changes, but it's still there.  So I decided to investigate some more.

I removed all of the tubes in the ReVibe.  There is no Input signal into the ReVibe.  The Revibe's Output Jack is connected with a guitar cable to the Amp's Input jack. The Amp and the Revibe are plugged into the same wall receptacle. The amp also has its own AUX AC receptacle and no difference if I use that for the Revibe's power source.  If the Amp is turned On, and the ReVibe is turned OFF, there is low level, but noticeable, HUM.

If the Output Cable from the Revibe's Output Jack is unplugged at the Output Jack on the ReVibe, NO HUM.

Now here is possibly a major clue to the HUM problem. If I unplug the ReVibe from the AC wall receptacle, and even if it's Power Switch is OFF, but the Amp is powered On and the Revibe is still connected to the Amp's Input jack with the guitar cable from the ReVibe's Output Jack, there is a deafening HUM and what I would call Oscillation.

So it seems there's a grounding problem that's creating what I assume is a ground loop, and the problem is worsened when the ReVibe can't get to its own ground point and goes back to the Amp's ground somehow, even if there is no power at the ReVibe's power transformer.

I feel like if I can find the fix for this, then the HUM will be cured.  But I have no idea where to go.

Anyone's help is appreciated.  Thanks.

[sluckey]

Disconnect the power cord green wire from chassis ground. If the hum decreases or is eliminated, you may wish to copy the reissue 6G15 floating ground scheme. This means isolate ***ALL*** ground connections from the chassis except the power cord green wire. Now add CR5, CR6, and R23 as shown in the attached pic. These three components will be connected between your floating circuit ground and the chassis. So, the only things that are actually connected to chassis are the power cord green wire and the three added components.

[wsscott]

Thanks, but I’ve already done that.  It did clean up the circuit induced hum a lot, but I feel like this is something else going on.

With no tubes installed, and it’s connected to the amp, and the Revibe’s power is off, I’m getting hum from it through the amp.  And if I unplug the Revibe from the receptacle, the noise is huge.

[wsscott]

I should add that the CT of the PT is grounded along with the first filter cap at a different point on the chassis from the green power ground that is soldered to its own point on the chassis.  Every other ground is floating.

[sluckey]

Well, what did I say? I even highlighted in red!

[shooter]

...
look at the RED circled ground symbols that Steve pointed out, see how they are "fingered", schematically different than the typical "upside-down" triangle ground symbol.   that tells you the difference between "circuit ground" and "earth-ground".
it's usually NOT good to mix them together.

[wsscott]

Ok, so the 2 chassis grounds are circled in red.  And if I understand correctly, you're saying that I screwed up in connecting the 1st Filter Cap Ground and the Power Transformer's CT to the Chassis also, even though at a different location on the chassis.  They need to be connected to the same Ground Buss wire that everything else is connected to and that connects to Chassis ground only through the Hum Stopper circuit that's been added at the end of the Ground Buss?

[sluckey]

Disconnect the power cord green wire from chassis ground. If the hum decreases or is eliminated, you may wish to copy the reissue 6G15 floating ground scheme. This means isolate ***ALL*** ground connections from the chassis except the power cord green wire. Now add CR5, CR6, and R23 as shown in the attached pic. These three components will be connected between your floating circuit ground and the chassis. So, the only things that are actually connected to chassis are the power cord green wire and the three added components.

***ALL*** includes all jacks too. If you used SwitchCraft type jacks you will need to isolate by using two insulated shoulder washers per jack. The mounting hole will have to be enlarged. Cliff style jacks are already isolated, so just don't connect the ground lugs to chassis. You don't need to isolate pot bodies from chassis, but don't connect any of the pot lugs to chassis.

This is worth repeating... So, the only things that are actually connected to chassis are the power cord green wire and the three added components.

[wsscott]

Sluckey-I used the hum buster circuit you attached.  I also have insolation washers on the Switchcraft Input and Output jacks, the Reverb/Tremolo foot switch jack, and the Reverb Input and Output RCA jacks.  None of the pots are isolated, but the all get their grounds by connecting their 3rd lug to the Ground Buss.

Since the PT doesn't have a 6.3VAC with a CT, I ran the filament leads from the PT to the Pilot light where they connect with the filament wiring, and then connected 100 ohm resistors to the Ground Buss.  So I think that should be correct.

I have to correct myself on the CT/1st Filter Cap ground.  They are not grounded to a separate chassis location, but are connected together and then a lead runs from there to the Ground Buss.  That Ground Buss is only connected on 1 end, and that is at the Hum Buster circuit where it meets the 2 diodes, resistor and a cap.  The other end of those components then connects to the Chassis.

I just noticed that on the Switched Output Jack, only the Tip is wired.  I don't know why I didn't wire its Shield lug to the Ground Buss. I used a shielded wire to connect the Tip on the Output Jack, to the Center lug on the Output Level pot, and connected the shield on that wire to the Ground Buss at that point.  So the Output Jack has no Ground connection.

I guess I better add that wire.

But how could I be getting an output signal from the Revibe to the Amp if the jack wasn't grounded?

[sluckey]

The output jack must be connected to the circuit ground buss.

But how could I be getting an output signal from the Revibe to the Amp if the jack wasn't grounded?

The signal flows directly through the cable center conductor. But... the signal return must travel through your ground buss then through the ground buster circuit to the chassis. Then it travels through the power cord green wire to the AC outlet, through the house wiring to the AC outlet the amp is connected to, and finally to the amp chassis/circuit ground. This is a huge ground loop and is probably the cause of your hum.

We need to see several hi-rez pics.

[wsscott]

Now that's making some sense.

I'll wire the Output jack to ground, jumping the center lug with the ground lug on the jack, and then wiring the ground lug to the Ground Buss.

I'll get some photos.  What do you want them showing?

[sluckey]

Hi rez gut shots that clearly show your board, tube sockets, control panel, interconnected wires. Particularly interested in seeing your ground buss, filter caps and PT center tap ground connections. Also want to see your ground buster.

[wsscott]

Got it

[wsscott]

So I wired the Output Jack of the Revibe so it now has a ground connection.  I wired the center "switched" lug to the far left ground lug, and then connected that lug to the Ground Buss.

This morning I plugged it in, still with no tubes, and it's DEAD QUIET!  I know there is an electrical connection between the Revibe and the Amp because if I turn up the Output Level on the Revibe almost all the way I can hear some faint noise from the Amp's speaker.  Also, if I unplug the Revibe from the wall receptacle with the Amp on and the Output cable connected to the Amp, it stays quiet and I don't hear the extremely loud hum and oscillation as before.

So later this morning I guess I'll put in the tubes and see what it's like.  Any suggestions on that process, or just put them all in and turn it on?

Thanks.

[sluckey]

Just do it. There is no risk.

[wsscott]

Well grounding that Output jack solved the crazy loud hum/oscillating problem I was having.  Thankfully.

The Revibe still has some hum that increases as the Output Level pot is increased, but it is not affected with its Volume pot or Reverb Intensity pot.  It probably would be okay playing in a band at a gig or something like that, but certainly not for recording.

So now that I cleared up the BIG Hum problem, I'll just go through the circuit and see if I can find the source.  I think it's 60 HZ that I'm dealing with.

If I find anything that works I'll let you know.

Best.

[passaloutre]

I'm in the middle of building a 6G15 and running into hum issues of my own. I followed the 60s Fender circuit and layout but I'm thinking I need to elevate the ground.

Can anyone explain the theory behind why elevating the ground a fe volts reduces the hum?

[sluckey]

Can anyone explain the theory behind why elevating the ground a fe volts reduces the hum?

I've never heard of such.

[wsscott]

I've seed a D-Lab video dealing with a ground loop on a 6G15 kit, I think a Monotone, and he used one of those 2 prong plug adapters that cleared it up.  He talks about a couple of other ways in a video that's only 2 minutes long.

The thing that would concern me is that then you don't have the safety of the earth ground--I don't think.

Anyway, that's not my problem that I've been dealing with.  Mine is somewhere in the circuit.  Why and where, I haven't figured out yet.

[wsscott]

I just used a FFT plot App, and it read the sound as 120Hz at -47.3db.

So its not 60 cycle hum, its 120 cycle.

[shooter]

that's not my problem that I've been dealing with.

don't write off the horse before the cart
try it, it's quick n easy, solved 90% of my ReVibe hummmmmmmmmmmmmmm

[wsscott]

Shooter, I just tried that plug adapter before I posted my response to see if it would now solve the problem.  I should have mentioned that.  Didn't solve my problem.  But it looks like I'm deal with 120hz and not 60hz as I had previously thought.

[wsscott]

So I started removing tubes to see the effect on the Hum.  The only time the Hum almost disappears is when I remove the last tube before the signal goes to the Output.  That is the 12AX7 that receives the Dry Signal on one half, and the Reverb Recovery on the other half of the tube.

That is the part of the circuit I have been focusing on forever.

Of course with that tube removed no signal is getting to the Output, but you can hear some amp noise at very low levels that does increase as the Output Level pot is increased.

I'll keep working on it.

[passaloutre]

Can anyone explain the theory behind why elevating the ground a fe volts reduces the hum?

I've never heard of such.

Is that not what the diodes and resistor do? Could you explain the theory behind the diodes and resistor you posted? Clearly I missed the point

[shooter]

the circuit does add a DC offset to chassis ground, fe volts anyway 

[kagliostro]

The resistor between signal ground and earth ground limits ground loop currents, reducing hum.

The two antiparallel diodes remain off during normal operation, but conduct if the voltage rises due to a fault, bypassing the resistor.

Result:

low noise in normal conditions
safety in case of fault, with a near direct connection to ground when needed

Franco

[wsscott]

In looking at the 120hz hum issue, I first thought maybe the coupling caps in the last tube position may be bad, but I realized that those caps are supposed to let AC pass through.  Since the 120hz hum is AC, it would pass through the cap. So it must be originating elsewhere.

So where could that be?  Could the bridge rectifier have a bad diode inside it? The rectifier I'm using is like a RS401L; however, since I'm using the CT on the PT the Ground lead on the bridge rectifier is cut off and is not being used for the circuit. I don't know how to test that rectifier.  Or could it be one of the filter caps on the Node serving B+ to that last tube?

I do have a scope, but I'm not very good at using it yet.

Any thoughts on where to go searching?

[shooter]

could it be one of the filter caps on the Node

I always measure BOTH...DCV AN ACV at each node
I look for VAC < 2vac at the PA node (which you don't have), <50mVac at "V1's" node

[wsscott]

Shooter-I hooked up my scope in AC mode to the first filter cap, V1, ie. the reservoir cap and it reads 5.11VPP and 4.17vRMS with a 120hz signal.

The scope is connected to the Ground lead on the cap which also connects to the CT and to Ground BUSS.  The Positive lead from the scope is connected to the Positive lead on that cap where the Dropping resistor is also connected along with the Postive lead off the "bridge" rectifier.

This doesn't sound too good does it?

[sluckey]

That sounds very typical. A good way to test would be to get a 40µF cap and a couple gator clip test leads and bridge the test cap across each of the filter caps (one at a time) and listen for a reduction in hum. Since the node E cap is only 1µF, consider replacing with a 20µF cap.

[wsscott]

I put a 47uf cap connecting Node A 40uf to its + lead and - to the chassis, and still loud hum.

I then connected it to the Node B 40uf cap the same way and it was lower.

THEN, I connected it to the Node C 30uf cap the same way, and DEAD SILENT!

Node C feeds the last plate of the last tube before going to the Output Level pot.

So that cap must be bad or something.  So I'll start tomorrow by changing out that cap.

Let me know if I'm missing something or am claiming victory too soon.

Thanks.

[shooter]

claiming victory too soon.

never claim victory, gremlins can sense the over-confident n crash the party EVERY time! 

[sluckey]

I'd like to suggest 50µF at all five B+ nodes.

Here's the original schematic. Don't blame me for the weird B+ nodes wiring. That's just the way it is in a real M10-A.

[wsscott]

So I'll install the one F&T 47uf/500V cap I have into Node C and make sure that solves the problem.  F&T doesn't make a 50uf, but I assume the 47ufs should be fine.

Would changing these cap values cause me to need to change the values of the dropping resistors between these nodes?

As far as Node E, the 1uf, is concerned, it's in a location that's really tight, and the 47uf would not fit.  Any suggestions?  Right now its a small MOD cap, I think.

[sluckey]

Would changing these cap values cause me to need to change the values of the dropping resistors between these nodes?

no

As far as Node E, the 1uf, is concerned, it's in a location that's really tight, and the 47uf would not fit.  Any suggestions?  Right now its a small MOD cap, I think.

Gee, wish I could see those hi-rez pics I asked for twice!

[wsscott]

Here are a couple of photos that hopefully will help.

#24 shows the Node E cap and the Hum Buster board I added.

#26 shows the other 4 Node caps.

[sluckey]

So, use a 22µF radial cap for node E. This cap is only 1/2"dia x 1.25" long. Just leave the 1µF in place and solder the 22µF leads to the 1µF leads.

I see a 33pF cap connected between V1 pin 7 and chassis ground. Should be connected to the ground buss.

[wsscott]

Improved, but what was dead quiet yesterday when I jumped the 47uf cap across the Node C 30uf cap, has some hum today.

I added the 22uf cap across the Node E 1uf cap.  And I also moved the ground lead on the 33pf cap from the chassis ground point and connected it to the Hum Buster board where the lead from the Ground Buss connects to that board.

I'll order some 47uf caps and replace all the 30's and 20's with them.

I checked the 30uf cap that I removed to see its specs, and it is reading 41uf.  I also checked its ESR, and it is very low at only .53. Seems awfuly low to me. The new 47uf I installed has an ESR of 2.2.  The old cap shows OL when I check its leads, and doesn't show any continuity between them.  So I don't know if that cap is bad or not.

[sluckey]

Improved, but what was dead quiet yesterday when I jumped the 47uf cap across the Node C 30uf cap, has some hum today.

So use an 80µF or 100µF.

[wsscott]

I jumped the old 30uf that I just took out, but no difference.

So I'll get a 100uf and see if that changes anything.

Yesterday when I jumped the cap with the 47uf that I just installed, I connected the ground lead of that 47uf to the chassis rather than the ground lead off the 30uf cap.  Would that make any difference, or give an indication of what's going on?

[shooter]

Would that make any difference

try it, won't hurt anything

[sluckey]

Yesterday when I jumped the cap with the 47uf that I just installed, I connected the ground lead of that 47uf to the chassis rather than the ground lead off the 30uf cap.

Didn't we just go through a long process to ensure that all circuit grounds were isolated from chassis ground?   

[wsscott]

Yes we did.  It didn't make any difference.  I has hum, at a little lower level with the new cap, but it's the same whether the ground jumper is attached at the ground lead on the cap or on the chassis.  So no difference.

[shooter]

it's the same whether the ground jumper is attached at the ground lead on the cap or on the chassis.

now you now, now you can sleep soundly 

[HotBluePlates]

So I started removing tubes to see the effect on the Hum.  The only time the Hum almost disappears is when I remove the last tube before the signal goes to the Output.  That is the 12AX7 that receives the Dry Signal on one half, and the Reverb Recovery on the other half of the tube.

Skimming, I think I saw you have some capacitors available.

EDIT:  recommended test was for a Fender 6G15, not the actual circuit used.

Go to that last 12AX7/7025 and focus on the section with the Dry signal, which has the 100kΩ resistor from cathode to ground.
   - Tack a filter cap (30µF? 47µF?) across the 100kΩ cathode load resistor; you need a relatively high voltage rating here.
   - Turn the Mix control up for full-Reverb.

   - Do you get Reverb in the output?  Of course the "Dry" signal is gone... but is "Hum" also gone?


I'll try to remember to come back and check for the result of this test.

[wsscott]

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?

[wsscott]

I jumped a 30uf cap across the 2.7k resistor with the - lead of that added cap being connected to the lead of the resistor going to ground, and got full reverb all the way turning the reverb intensity all the way up, but also got a really loud HUM.  Louder than without the added cap.

[wsscott]

So, I've removed that cap jumpered to the 2.7k resistor, so the amp has the new Node C 47 uf cap installed.  This is the V7, 12AX7 tube which has Nodes C and E filter caps connected to it.  Node E is for the Reverb Recovery and Node C is for the Dry path from the reverb circuit.

With all tubes installed, there is hum that increases when the Reverb is turned on with the foot switch, and the hum also increases when the Output level pot increases.

With the V7 tube removed, obviously no guitar signal goes to the output, but there is a much lower level of hum at the amp's speaker but the hum increases as the Output Level pot increases.

[shooter]

With the V7 tube removed, obviously no guitar signal goes to the output, but there is a much lower level of hum at the amp's speaker but the hum increases as the Output Level pot increases.

isolate the output jack from the chassis, just unbolt n let it "float" by disconnecting the ground wire also, so the jack is ONLY passing the signal at the tip n NO ground reference

[wsscott]

Shooter-That jack is already isolated from the chassis with isolation washers.  Also I previously did not have its ground connected and it made a horrible loud hum and oscillation when it was unplugged from AC but still connected to the amp which was turned on.  So I wired the ground back onto the jack and that problem was solved.

So I think this has already been tested.