Reverb Hum-How to get rid of it

[wsscott]

Several years ago I built a mod of a Mag10A's pre-amp circuit to be used as a "ReVibe" type effects unit.  Sluckey, with others, helped greatly with the schematic's mods and the build.  This was my first build from scratch so there was a lot of learning for me to do.

Anyway, it works great except it's always has had a Hum when the Reverb is turned on that I can't get rid of.  When the Reverb is Off, it's perfectly quiet.  I feel like it's a ground loop, but I can't figure out how to fix it.

Attached is a copy of the schematic for this part of the circuit.

It is wired as follows:

The RCA Input and Output jacks are attached to the chassis and isolation washers are used.  I hoped that might help, but apparently not.

The Neg. side of the 1uf cap is wired directly to the Tip tab on the RCA Input Jack, and the Shield tab on that Jack is wired directly to the GND Chassis BUSS.  The Input on the tank's RCA jack connects with a RCA cable to this input jack.

Then the Output Tip tab on the RCA Output Jack connects to the Grid on the Recovery Tube through shielded cable whose shield is grounded only on 1 end at the Shield tab of the RCA Output jack. There is a 47K ohm resistor connecting the Tip tab with the Shield tab on that RCA Output jack. The Shield tab on that jack is then wired directly to the GND Chassis BUSS. The Output on the tank connects with a RCA cable to this Output jack.  The Shields on both of these RCA jacks are in continuity with the Chassis.  So its pretty simple.

Any thoughts on what's causing my HUM?  Thanks.

[tubeswell]

Which way around is your reverb pan mounted? (If the output transducer coil is sitting next to your amp’s PT, it will hum more)

[wsscott]

The Output is at the opposite end from the transformer.  So its not close.

[shooter]

does it hum as a "stand-alone unit"


my ReVibe would hum "in circuit" but standalone it was fine, I had to "float ground" on either the amp or the revibe

[wsscott]

It has no Output tubes, so its output just goes from it's output jack and then through a guitar cable to the input on my amp.  It's perfectly quiet just sitting there as long as the amp isn't turned on, or if the amp is on and the signal is bypassing the Reverb part of the circuit.

[SEL49]

Disconnect the power cord green wire from chassis. Any better?

Show us the entire schematic.

[wsscott]

No difference if mains ground is disconnected.

Attached is entire schematic.

[shooter]

move V1's "E" tap to "D" for testing

[SEL49]

You have several threads that deal with this same issue. In one thread you said you had solved the hum. Read this thread and see if there are any clues.

     https://el34world.com/Forum/index.php?topic=25746.msg279595#msg279595

[wsscott]

Shooter-Do you mean swap D & E and V1's Pins 1 and 6?

SEL 49 I got the hum much lower than it had been and thought I'd just live with it.  I can lower the Output Level and keep the Reverb Intensity low, and it's certainly acceptable at those levels, but I'd just like to solve this itch that I've been scratching for a long time.

Don't you guys spend a lot of time on this.  As SEL 49 said, its been discussed a lot.

[shooter]

Use Tap D for V1, NOT tap E

[wsscott]

No difference with D node on both plates of V1.

Shooter, was your ReVibe tube driven or transformer driven?

When you say "in circuit" and "standalone", what do you mean exactly?

[shooter]

mine was the one Doug has posted with the 6V6
by stand-alone, I used it like a pedal, guitar>Revibe>amp


EDIT: so I had 2 separate PS's, one in the Revibe, one in the amp.  Once I "lifted" ground from one unit the hum went to an "acceptable" level
I my world that means I can hear it with no signal, but once I roll the guitar volume past ~1 it's no longer an "issue"

[wsscott]

Shooter--Did you lift the green Mains safety ground wire, or some other ground?  Wouldn't that be dangerous not to have the PS grounded to the chassis?

[shooter]

used a 3 prone to 2 prone adaptor on the ReVibe unit.
when you work on 120,000 VDC for a living, "Dangerous" is sorta relative   

[Willabe]

Did you use 5 B+ filter caps in the power supply, like in the schematic?

I think, if you haven't done this, you need to try isolated RCA jacks for the reverb tank cables.

And make sure the verb tank is only at 1 point/1 RCA jack in the tanks chassis/housing, not at both RCA jacks.

So looking at the schematic;

V6 1M grid R and 270 K R and the RCA jack ground for the verb tanks input, get grounded with the B node filter cap.

Then for V7-A, the 47K grid R, 2K7 K R and the RCA jack ground for the verb tanks output, get grounded with the E node filter cap.

And the verb ft sw ground, V7-B's 1M verb pot on the grid, 2K7 K R, 100K output pot and the output jack, get grounded with the C node filter cap.

I'd try that and see if it helps. 

And you still probably have to lift the chassis ground from the chassis in the preamp/verb/vibrato unit.

It will still have a earth/chassis ground the through the guitar cable and chassis mounted 1/4" jacks.

[Willabe]

used a 3 prone to 2 prone adaptor on the ReVibe unit.

Try it and see if it kills the humm?

[Willabe]

The problem might be that with the dry signal path going from V6 A/B's K's to V7 B's K there's only 15K between them separating/isolating them from each others ground.   

Merlin might have a way to fix this humm problem?

[wsscott]

I did use a 3 prong to 2 prong converter with no impact.

I'm using isolation washers on the RCA jacks so the jacks themselves are isolated from the chassis.

I used all 5 B+ caps as shown in the schematic.

Not sure what you mean by this phrase:  "And make sure the verb tank is only at 1 point/1 RCA jack in the tanks chassis/housing, not at both RCA jacks.

I feel like the Output Level pot is where the hum appears.  Adjusting the reverb Intensity pot minimally affects the hum, so that's why I think it's coming after that point in the circuit.  Maybe the Output Level needs to be set very low.  If the Input test signal is 1kHz at 150mVrms, should the Output level of that signal that goes from the Output Jack to the amp also be at that level? 

[shooter]

If the Input test signal is 1kHz at 150mVrms, should the Output level of that signal that goes from the Output Jack to the amp also be at that level?

while signal amplitude may cause distortion by OD'ing the amp, it typically doesn't "induce hum"
maybe it's scope time to figure out hum Frequency, where it "shows itself" along the signal path...

[wsscott]

I don't want to beat this horse too much longer, but here's what I've determined.

The guitar is plugged into the reverb/vibrato headunit's Input Jack, and the headunit's Output Jack is plugged into the Input Jack of the amp.  Standard guitar cables are being used.

The Output Level pot causes the hum to increase or decrease when the Output Level pot is increased or decreased regardless of whether the Reverb is On or Off.

The Reverb Intensity pot only has a very minor affect on hum, but only when the Reverb is ON.
If the Reverb is Off, increasing or decreasing the Reverb Intensity level have no effect on the hum.

My conclusion is that the Reverb circuit is fine.  The hum is in the Output Level adjustment.  My "guess" is that it is increasing the gain of the Output signal going to the Input Jack on the Amp, and maybe its just "line-level" noise that's being magnified.

[Willabe]

I'm using isolation washers on the RCA jacks so the jacks themselves are isolated from the chassis.

I used all 5 B+ caps as shown in the schematic.

Ok, these are both good.

But how did you ground the circuit grounds to the filter caps grounds?

In the verb circuit, did you ground the grounds like I wrote out?

Not sure what you mean by this phrase:  "And make sure the verb tank is only at 1 point/1 RCA jack in the tanks chassis/housing, not at both RCA jacks.

Take the tank out and look inside at the RCA jacks. Only 1 of those tank jacks should be grounded to the tanks chassis/housing/frame. If both are grounded to the chassis/housing/frame it's a ground loop. Disconnect 1 of the tanks chassis/housing/frame grounds if that's the case. That will break the ground loop.

I feel like the Output Level pot is where the hum appears.  Adjusting the reverb Intensity pot minimally affects the hum, so that's why I think it's coming after that point in the circuit.  Maybe the Output Level needs to be set very low.

 

The output level needs to be set loud enough to make up for any losses in the signal from the vibrato and reverb. So unity gain or a maybe little more, if it can supply that?   

If the Input test signal is 1kHz at 150mVrms, should the Output level of that signal that goes from the Output Jack to the amp also be at that level?

That would be unity gain, not less gain/loss, not more gain/amplified/boost.

[Willabe]

I'd also try this.

If you reverb ft sw is not isolated from the chassis, could be a ground loop.

Try leaving it wired up, but remove it from the chassis. Wrap it up with some tape so you don't short something out.

The ground for that verb ft sw should be grounded with the output pot's ground, V7 B's 2K7 K R, 1M reverb pot, the output jack and C node filter cap. If you were using just that jacks connection to the chassis as ground, you'll need to isolate it from the chassis and then run a wire from that jacks ground lug over to the C node's filter caps ground wire.

And the output jack should be isolated from the chassis too.

These 2 jacks and the output circuits grounding might be why the humm gets louder turning up the output pot.

[Willabe]

And can you please post some pics so we can see how it's laid out and how you wired it up?

[wsscott]

Willabe-The footswitch jacks are also using isolation washer to isolate the Reverb and Vibrato from the chassis.

As to the tank, I had already unsoldered the connection on its Input jack to isolate it and that made no difference.

I'm attaching my Layout.  It's ugly, but it is the way the amp's built.  I separated the Grounds for the filter caps from the Grounds for the preamp that are located on the circuit board, and used the Ground BUSS to ground everything.  So I think its grounded as you suggest, but I may be wrong.

[Willabe]

As to the tank, I had already unsoldered the connection on its Input jack to isolate it and that made no difference.

So it had both RCA jacks grounded to the housing?

And now only 1 RCA jack is grounded to the housing?

Go back and read what I added to my last post, about the output jack grounding. 

[Willabe]

You said you used 5 B+ filter caps like the schematic. I only see 4 filter caps.

And that is not grounded anything like I wrote out.

You ground all the grounds from a tube section with filter cap's ground lead that feeds that tube section.   

Are all the green wires ground wires? You should only use 1 color for all the ground wires and only that color for the ground wires. Then you can trace the ground scheme.

[wsscott]

The E node 1uf cap is on the upper right hand side of the board, to the left of the large 470 ohm resistor on the far right end of the board.

You are correct on how I did the grounding.

The Layout is not consistent with colors using DIY Layout Creator.  Remember this is the first time I ever did anything like this, so it's not pretty, but the unit seems to work pretty well.  I created the Layout from the Schematic with no one assisting on how to create the Layout.  Sluckey was great with the Schematic, but he said I needed to learn how to do a Layout by myself.  And that was a good way for me to learn.

I did some adjustments to the Output Level pot, and with a guitar signal input I can get proper output volume with extremely low hum right at about Level 1, or 7 o'clock.  The pot is just barely turned on.  The more Output, the louder the hum.  I also inserted a 150mVrms signal and with the Output Level in this position, the Output jack reads about 27mVrms.  A much lower amplitude but works.

So maybe I should just live with it.  It's not like I've got to push the amp's volume way up.  But there's still that itch!!

[SEL49]

I did some adjustments to the Output Level pot, and with a guitar signal input I can get proper output volume with extremely low hum right at about Level 1, or 7 o'clock.  The pot is just barely turned on.

There is an onboard 1M resistor connected to the output pot. Change that to 2.2M or 3.3M and you'll be able to turn the knob up higher. This won't change your hum issue but it may just feel good being able to turn the knob higher.

Alternately, replace the output pot with a 25K-A pot for the same results.

[SEL49]

...with a guitar signal input I can get proper output volume with extremely low hum right at about Level 1, or 7 o'clock.

This indicates that you don't really have a hum issue.

[Willabe]

I also inserted a 150mVrms signal and with the Output Level in this position, the Output jack reads about 27mVrms.  A much lower amplitude but works.

Lower than unity gain.

[wsscott]

Thanks fellas.

SEL 49-are you saying there's no issue of hum?  Sorry but I hear hum, so I don't understand your comment.

Willabe-yes lower than unity gain.  But is that normal, or is it just a meaningless fact?

[SEL49]

SEL 49-are you saying there's no issue of hum?

I didn't say that. I said, "This indicates that you don't really have a hum issue." That was in response to your remark, "proper output volume with extremely low hum". Extremely low hum is a non-issue in my mind. Don't you agree?

[wsscott]

Sure, I agree with that.

Its not that the hum isn't there, it does exist, its just that its controllable.

So I guess the "salve for my itch" is the conclusion that the circuit is fine, the build is fine, and that's just the way it is.

Like some of the mechanics at car dealers say about some MB's and BMW's, "They all do that"!! HAHa!

[tubeswell]

Noting that you report that the hum occurs when the reverb is activated (but isn't present on the dry signal) it seems that the problem is grounding of the signal returns for the (highly noise-sensitive) reverb recovery stage is getting muddled/overwhelmed by higher current returns. As this is a stand alone effects unit that is intended to be plugged in in front of a guitar amp, you might want to consider regrouping the ground returns and installing a signal ground hum-loop blocker between the ground buss and the chassis of this stand alone unit. This should significantly quieten/dampen loop hum.


Here's an example of how I wire 6G15 ground returns to eliminate loop hum - I can vouch that this yields a quiet noise floor. The same principle of galactic grounding around filter cap nodes with a daisy-chained buss to a hum-loop blocker can be applied to your unit - but it may require a bit of re-work.


Or maybe you're happy to live with it

[wsscott]

Tubeswell-thanks for the lead.  I've seen that sort of wiring before.  A couple of years ago I got a Pyle PHE 400, but it really didn't make much difference.  Originally I built this to use with a Fender Champ kit that I built, and then later I build a mod of the Magnatone Twilighter.  When I built the Maggie I install a AUX power socket for plugging this headunit into to hopefully solve the grounding/hum problem.  And that's how I use it now, but still the issue. I've certainly improved the hum issue since I built it, but it's just this nagging bit that hangs on.  I'm not one to give up on stuff, so that's why it's more of a matter of principle now.  I'm not a perfectionist, but I do think logically and analytically, so that's why this dog won't leave this bone alone.

I've wondered if that last section of the circuit where the resistor and pot come off the plate and go to the Output jack is where it all lies.

[tubeswell]

Due to a need to have 3-prong safety earth mains cords on all electrical gear of this type, stand alone units are prone to hum loop if you don't get very fastidious with the ground return scheme.

The way you group the ground returns is important. The place you choose to attach the hum-loop blocker to the chassis is important. Isolation of all jack ground returns from the chassis (and having separate return wires to the relevant daisy chain location on the ground buss) is important.

[Willabe]

Willabe-yes lower than unity gain.  But is that normal, or is it just a meaningless fact?

No, it's not normal.

Your knocking down, way down, the output from your guitar PUP's.

Lots of guitar pedals have an output volume control so you can adjust the output with the FX turned off to match the output of the FX on, so their both the same.

[Willabe]

I did some adjustments to the Output Level pot, and with a guitar signal input I can get proper output volume with extremely low hum right at about Level 1, or 7 o'clock.  The pot is just barely turned on.  The more Output, the louder the hum.  I also inserted a 150mVrms signal and with the Output Level in this position, the Output jack reads about 27mVrms.

That's just  1/6  of the input.

What ever your PUP's put out the output of the pre/vib/verb unit will be only 1/6 of what the PUP's put out at that setting.

A much lower amplitude but works.

Just barely.

You should not be hand cuffed like that because you have to set the output volume so low to keep the humm down. That is not normal for any build.

[wsscott]

Willabe-In your Reply #26 you said "You ground all the grounds from a tube section with filter cap's ground lead that feeds that tube section.

Are you saying that "You should ground ..." or that I in fact "grounded all the grounds...".  Are you saying the way I should ground the grounds from a tube section is to ground it directly with that tube's Filter Node cap?

In my case, all the filter caps' Neg/GND leads are wired together and then 1 wire connects them to the GND BUSS.  Also, for what its worth, the two 100 ohm resistors which connect to the Pilot Light and the 6.3V filament leads for the artificial center tap, connect to the Node A Filter cap, which along with the other filter caps attach to that same GND BUSS.

If I recall correctly, the rectifier I purchased was a full wave bridge rectifier, and I was told not to wire its GND lead because it would produce too much voltage and that I should cut that leg of the rectifier off.  So I did.

Would there have been any need to use the 100 ohm resistors for an artificial center tap?

I used a Hammond 269JX that of course has a center tap for the Secondary power supply.  I connected that wire with the Primary's Green Mains ground lead directly to the chassis.  Its 6.3V filament leads do not have a center tap.

Hope this is helpful.  I do have an entry level digital table oscilloscope if that would be useful.

[wsscott]

So I did some more tests.

The Hum is there whether there is any input signal or not, and is lower when the reverb is turned off.

If I disconnect the Reverb TANK completely from the circuit by disconnecting the RCA cables from the RCA Jacks on the chassis, the result is the same as above.

So the Tank has no impact on the HUM.

[wsscott]

Also, with the tank removed, and the guitar plugged in, guitar sound (as well as the HUM) is heard at amp regardless of whether the Reverb is switch on or off.  Just no reverb effect.

[Willabe]

Willabe-In your Reply #26 you said "You ground all the grounds from a tube section with filter cap's ground lead that feeds that tube section.

Are you saying the way I should ground the grounds from a tube section is to ground it directly with that tube's Filter Node cap?

Yes, this ^^^^^^^^^^^^

Each B+ filter cap node feeds (supplies B+ to) a tube/tubes in a section of the build. The grounds for that tube/tubes get grounded with that B+ filter caps ground lead.

This forms a ground star. Then you daisy chain all the B+ filter cap ground stars together. Then you ground the daisy chain to the chassis at only 1 point.

Read this from Merlin on grounding, go down to multiple star grounding;

https://www.valvewizard.co.uk/Grounding.pdf

In my case, all the filter caps' Neg/GND leads are wired together and then 1 wire connects them to the GND BUSS.

That's not the way you should run your grounds. You said earlier that you "separated all the filter caps from the preamp".  You don't do that. You probably were thinking of when the main B+ filter cap that feeds the OT CT (feeds the power tubes) is grounded separate from the preamp chassis ground. That's because the power tubes draw a lot of current and the ground on that filter cap if grounded with the low current draw preamp tubes will modulate the preamp tubes, causing humm. But this build has no power tubes.

And you wired up 1 of the ground busses (you have 2 busses) in the worst place. The most current in that build is at the 1st B+ filter caps ground lead with the PT's CT. So it's the noisiest ground in any build.

You did the best thing to run the PT's B+ CT directly to that 1st filter caps ground lead. But then you shouldn't grounded 1 of the ground busses with it. You should have run the buss and filter caps in separate ground stars. But at least connect that buss to 1 of the other nodes 

Also, for what its worth, the two 100 ohm resistors which connect to the Pilot Light and the 6.3V filament leads for the artificial center tap, connect to the Node A Filter cap, which along with the other filter caps attach to that same GND BUSS.

The 2 xs 100ohm R's grounded there is probably ok. 

Would there have been any need to use the 100 ohm resistors for an artificial center tap?

Yes, that PT's secondary heater supply has no CT. 

I used a Hammond 269JX that of course has a center tap for the Secondary power supply.  I connected that wire with the Primary's Green Mains ground lead directly to the chassis.  Its 6.3V filament leads do not have a center tap.

I looked at the spec shet for that PT, I see no "Primary's Green Mains ground lead."

Look at how Merlin shows how to wire up a ground buss, bottom example b.

2nd Merlin drawing shows how to wire up a tubes grounds with the B+ filter cap that feeds it B+ dcv. 

[Willabe]

Your PT spec sheet link;

EDB269JX.dcd - 269JX.pdf

[wsscott]

The CT for the PT's Secondary is grounded to the chassis at the same point as the PT's Green GND wire.

The 100 ohm resistors creating an artificial CT for the 6.3VAC filament leads is connected to the Node A filter cap, which is connected to Nodes B and C, which are connected to the GND BUSS.   Node D is connected directly to the GND BUSS, but not to the other caps.  Node E is connected to the GND connections on the Circuit Board which is only connected to the GND BUSS at 1 point at the far right end of the amp where the BUSS connects to the chassis.

Looking at the pots from the rear, with their tabs L to R being 1, 2, and 3, Tab 1 for the Volume, Bass, Output Level, Vibrato Intensity and Speed are all connected to the GND BUSS.

The Switching Input and Output jacks are also grounded to the GND BUSS.

I hope I'm not making this too confusing.

[tubeswell]

The CT for the PT's Secondary is grounded to the chassis at the same point as the PT's Green GND wire.

The 100 ohm resistors creating an artificial CT for the 6.3VAC filament leads is connected to the Node A filter cap, which is connected to Nodes B and C, which are connected to the GND BUSS.   Node D is connected directly to the GND BUSS, but not to the other caps.  Node E is connected to the GND connections on the Circuit Board which is only connected to the GND BUSS at 1 point at the far right end of the amp where the BUSS connects to the chassis.

Looking at the pots from the rear, with their tabs L to R being 1, 2, and 3, Tab 1 for the Volume, Bass, Output Level, Vibrato Intensity and Speed are all connected to the GND BUSS.

The Switching Input and Output jacks are also grounded to the GND BUSS.

I hope I'm not making this too confusing.

Your grounding is bad. That's why you're getting hum

You should use the grounding scheme I posted above with the hum-loop blocker. It is based on Merlin's daisy-chained galactic ground return scheme

[wsscott]

Tubeswell-Thanks.  Looks like I'm going to have a busy afternoon!

[tubeswell]

The loop blocker should be the only connection between the signal ground return buss and the chassis


All other ground returns (including high-current retruns) should go to the buss BUT the high current returns go to the opposite end of the buss that the low current returns go to, and the loop blocker is at the low-current return end of the buss.


At various points along the buss, there will be clusters of returns that go with respective filter cap ground for the filter cap that is supplying those components. These clustered returns need to be daisy-chained in sequence with the order that the power supply dropping resistors/filter cap nodes are set out

[wsscott]

Yea, I think at least part of my problem is that Node E is carrying 258VDC and it is grounded on the low voltage ground buss on the circuit board, which then connects at only 1 point to the GND BUSS.  All the other nodes connect to that GND Buss at the opposite end of the BUSS.  So Node E is at the far right end and is connected to the GND BUSS with all the other grounds for the pre-amp section. 

Seems like this might be fixed/improved if I disconnected the Node E cap from the GND Buss at that point and somehow get it connected all the way down to the other Nodes.  I hate to have a long wire running the length of the chassis, but I don't see any other way to do that.  Does this make sense to do?

[tubeswell]

How to daisy-chain ground returns in an amp