Hum issues in a EH-150/185 build

[Jerry garrcia]

I have posted this in TDPRI but posting this question here since this seems to be the place for this amp due to many previous builds by its members.
This summer I built a EH-185 variant with an 6SC7 instead of two of the 6SQ7’s. Used a 6SN7 as a PI, PT 300-0-300V, OT 1760JB (6K), skipped the 20W bleeder, a preamp choke, single ground bus and also included a mosfet LND-150 reverb driver. 281V on the plates of the 6L6’s.
Did some more modding after completion due to I wanted to even out the mic and instrument channel a bit. Main purpose for the amp is bebop and the reason I built it was that I heard that the clean channel is really clean. It surely was that from the beginning but the instrument channel had a bit to low volume and the mic channel was an untamed beast. Removed the bypass cap on V1 to tame it a bit and added a bypass cap on the output tubes to increase the instrument channel a bit.
Last mod I did was to add a 1M grid leak to the instrument channel and skipped the volume pot. The function of that is what the volume pot on the ES-150 does anyway I was thinking. Will probably mix a bit before I decide what the best set up is for my son, who is the best guitarist in the family.
Now to the problem. Hum! Been there the whole time. The mean, evil hum. In Sweden, so 50Hz. I really hate it.
The lead dress is not optimal and will fix it but I don’t think that’s the main reason for the hum issue.
Been using the chop stick like crazy without success. Resoldered the filament wires and added shielded cables in all inputs to grid. Grid resistors in the sockets. Disconnected the reverb driver and still the same hum (have a 470K grid leak after the reverb driver and no 100K before). Also if I press down on the OT ground connection on the ground bus there is a high noise (also on other parts on the ground bus but more prominent there).

I’ve read that this is a hummy amp and shielded cables is a good thing.
So with your vast experience in this amp, where should I go next? Add shielded cables on the whole signal path? Isolate them switch craft jacks except the one that’s connected to the ground bus? Any suggestions are highly appreciated!

FA1AE1CD-651C-4C4A-82B4-6C5DB461254D.jpeg

137BCE08-30CE-4F71-A0D9-BCC0BA34D12D.jpeg

[Stevem]

Your bussing of all the grounds is not a good thing without question!

If anything your preamp grounds need to get landed to the chassis where the B+ center tap wire comes back to.

Also the B+ center tap needs to go right to the negative of the first filter, not to the chassis first.

I would also pull all the tubes and confirm that your power transformer is not coupling into your output transformer.
This will make for a 50 hz hum out of the speaker even with all the tubes out and might be able to be corrected by just pivoting the OT the other way without having to move it further away.

Is the secondary side of your OT grounded, and does that ground go right to the speaker jack first?

You don’t show the filiments on your schematic, how is the ground reference of that being made?

Also your rectifier is drawn wrong.
A 5U4 does not have a separate cathode, it’s directly heated.
Not easy on the filters these types of recto’s are!

[R.G.]

1. Is the hum 50hz, and therefore caused by the power line mains, or 100Hz, and therefore derived from the rectification of B+?
2. Where >>exactly<< does the high voltage power centertap wire go? The only really low-hum place it can go is to the negative terminal of the first filter cap. Any other place will cause 2x power line frequency hum by wobbling the "ground" voltage around.
Note that there are many more "where >>exactly<< questions about wiring that follow. These constitute some basics of how to wire for low hum. Also note that the schematic is almost useless for the purposes of reducing hum. Schematics assume that all wires are 0.00000000 ohms, and that all "ground" symbols are truly zero ohms from one another. This is not true in the real world, and much hum comes out of it. You have to wire your amp assuming that wires are just low value resistors.
4. Where exactly is the AC mains safety ground attached to the chassis metal? It should attach at one and only one place, and that should be a bolt or screw position that holds no other wires.
5. How many "ground" connection wires attach to the chassis? The ideal answer is 2 : one for the safety ground and one for grounding the input jacks, or optionally a single wire going to the negative terminal of the first filter cap. Any other connections of "ground" to chassis make a place that hum and oscillation problems can creep in.
6. Yes, including speaker ground. The speaker jack should be isolated from chassis, and its return/ground run back to the OT secondary common. If your power amp uses feedback or you just want your speaker outputs to be grounded, you can run a wire from the OT common back to the negative terminal of the first filter cap. Using the chassis for speaker return currents chances oscillation and feedback problems.
7. How are your volume/control pots grounded? The "ground" on a volume control or tone control should be a single wire leading back to the stage that is creating the signal that it's modifying. Using a heavy brass or copper strip, perhaps soldered to the cases of the pots, can let hum creep in.
8. How is your heater circuit wired? The heater secondary on the PT should have a center tap, preferably made by two equal-value resistors (100 ohms each is good) and that center tap tied... yes, back to the negative terminal of the first filter cap. The individual heater wires can be run side by side, but for truly lowest noise should be twisted together. This keeps the magnetic field the current makes captive between the wires, not broadcasting out to grids.
9. Where are your bypass cap negative leads wired? Ideally, yes, back to the star ground at the negative terminal of the first filter cap.
10. There are more, but this ought to help. These guidelines are an introduction to low ground-noise wiring.
11. Now there is magnetic pickup: Where exactly do your AC mains wire run and connect? The AC mains, like the heater wires, really ought to be twisted pair everywhere they go. The PT primary wires should be twisted together and end close to the PT body on terminal lugs, and the switched AC mains current should go to these two terminals after fuse, power switch, EMI caps, etc. No caps from either AC mains wire connect to the chassis. Switch both sides of the AC mains, not just one; this makes you safer when you work on the amp.

There is more, but I'm tired of typing. Schematics are almost useless for low noise wiring.

[Jerry garrcia]

Your bussing of all the grounds is not a good thing without question!

If anything your preamp grounds need to get landed to the chassis where the B+ center tap wire comes back to.

Also the B+ center tap needs to go right to the negative of the first filter, not to the chassis first.

I would also pull all the tubes and confirm that your power transformer is not coupling into your output transformer.
This will make for a 50 hz hum out of the speaker even with all the tubes out and might be able to be corrected by just pivoting the OT the other way without having to move it further away.

Is the secondary side of your OT grounded, and does that ground go right to the speaker jack first?

You don’t show the filiments on your schematic, how is the ground reference of that being made?

Also your rectifier is drawn wrong.
A 5U4 does not have a separate cathode, it’s directly heated.
Not easy on the filters these types of recto’s are!

The HT CT goes to the ground bus but to the point where the negative side of the first filter cap and the same point it attaches to.

The OT ground goes first to the output jack then a ground wire to the input jack where the ground bus connects to. That’s a stupid way to do it since the output jack is unisolated. Will redo the OT ground directly to the first input jack.

The reverb driver is grounded to the bus and also the isolated RCA Jacks.

The filaments have a artificial CT with two 100R connected to the output tubes cathode pin.

Will try the PT trick.

Will redraw the rectifier.

The preamp is also grounded at the bus.
So with my conclusions I have the ground points.
1) the main safety earth on a separate lug on the chassis.
2) OT grounded on the speaker jack.
3) all other stuff including HT CT to first input jack through the ground bus.

One thing is that the tone control ain’t grounded. Looked at a lot of home made schematics and almost non had that. Will need to try to ground it.

Regarding the use of a common ground bus I’ve heard and read so much about it it is hard to find the best way. A lot of different opinions and appreciate yours.
Thanks for all your inputs!

[Jerry garrcia]

1. Is the hum 50hz, and therefore caused by the power line mains, or 100Hz, and therefore derived from the rectification of B+?
2. Where >>exactly<< does the high voltage power centertap wire go? The only really low-hum place it can go is to the negative terminal of the first filter cap. Any other place will cause 2x power line frequency hum by wobbling the "ground" voltage around.
Note that there are many more "where >>exactly<< questions about wiring that follow. These constitute some basics of how to wire for low hum. Also note that the schematic is almost useless for the purposes of reducing hum. Schematics assume that all wires are 0.00000000 ohms, and that all "ground" symbols are truly zero ohms from one another. This is not true in the real world, and much hum comes out of it. You have to wire your amp assuming that wires are just low value resistors.
4. Where exactly is the AC mains safety ground attached to the chassis metal? It should attach at one and only one place, and that should be a bolt or screw position that holds no other wires.
5. How many "ground" connection wires attach to the chassis? The ideal answer is 2 : one for the safety ground and one for grounding the input jacks, or optionally a single wire going to the negative terminal of the first filter cap. Any other connections of "ground" to chassis make a place that hum and oscillation problems can creep in.
6. Yes, including speaker ground. The speaker jack should be isolated from chassis, and its return/ground run back to the OT secondary common. If your power amp uses feedback or you just want your speaker outputs to be grounded, you can run a wire from the OT common back to the negative terminal of the first filter cap. Using the chassis for speaker return currents chances oscillation and feedback problems.
7. How are your volume/control pots grounded? The "ground" on a volume control or tone control should be a single wire leading back to the stage that is creating the signal that it's modifying. Using a heavy brass or copper strip, perhaps soldered to the cases of the pots, can let hum creep in.
8. How is your heater circuit wired? The heater secondary on the PT should have a center tap, preferably made by two equal-value resistors (100 ohms each is good) and that center tap tied... yes, back to the negative terminal of the first filter cap. The individual heater wires can be run side by side, but for truly lowest noise should be twisted together. This keeps the magnetic field the current makes captive between the wires, not broadcasting out to grids.
9. Where are your bypass cap negative leads wired? Ideally, yes, back to the star ground at the negative terminal of the first filter cap.
10. There are more, but this ought to help. These guidelines are an introduction to low ground-noise wiring.
11. Now there is magnetic pickup: Where exactly do your AC mains wire run and connect? The AC mains, like the heater wires, really ought to be twisted pair everywhere they go. The PT primary wires should be twisted together and end close to the PT body on terminal lugs, and the switched AC mains current should go to these two terminals after fuse, power switch, EMI caps, etc. No caps from either AC mains wire connect to the chassis. Switch both sides of the AC mains, not just one; this makes you safer when you work on the amp.

There is more, but I'm tired of typing. Schematics are almost useless for low noise wiring.

1) will measure for real. Sounds like a 50Hz hum but even the sun has its spots.
2) yes. CT connects to the same point as the first filter cap connects to.
4) main safety earth connects to a single point in the chassis. Nothing else is connected to that place. One concern I have is that the power jack is placed between both 6L6’s. Made a mistake and was aiming for the place by the rectifier.
5-6) three since I’m using unisolated switch craft jacks and the speaker jack is the ground for the OT. Need to fix that but no isolation “rings” at home…
7) all pots are connected to the common ground bus except fire the tone control that is without a ground wire from one of the lugs.
8 ) HT CT on negative side of the first filter cap. Filament ground on 2x100R to each of the cathodes of the output tubes.
9) goes to ground bus.
10) thanks. The ac wires from PT are twisted but maybe not the extra primary leads. Think I just used heat shrink and let them in there. Same with the unused OT secondaries. Will need to check that. Only a single pole switch for mains and protected with heat shrink over the solder joints on the input jack and switch.

So the OT ground and the speaker jack might be the problem.
Will of course check all your tips. THANKS!

[Stevem]

No need to redraw the recto tube, I was just making note of it for you.

[pompeiisneaks]

It's been a while since I built my EH-185 but I forgot that some of those metal tubes needed a ground pin connected or they hummed pretty badly. you may not be using metal tubes, don't know.

Otherwise, I think I also ended up using some shielded signal wires to reduce hum ingress.

~Phil

[pompeiisneaks]

Oh I'll add I didn't see you've got an LND mosfet driven reverb, that maybe another source of hum? that's 'new' to the amp.

~Phil

[R.G.]

1) will measure for real. Sounds like a 50Hz hum but even the sun has its spots.
2) yes. CT connects to the same point as the first filter cap connects to.
4) main safety earth connects to a single point in the chassis. Nothing else is connected to that place. One concern I have is that the power jack is placed between both 6L6’s. Made a mistake and was aiming for the place by the rectifier.
5-6) three since I’m using unisolated switch craft jacks and the speaker jack is the ground for the OT. Need to fix that but no isolation “rings” at home…
7) all pots are connected to the common ground bus except fire the tone control that is without a ground wire from one of the lugs.
8 ) HT CT on negative side of the first filter cap. Filament ground on 2x100R to each of the cathodes of the output tubes.
9) goes to ground bus.
10) thanks. The ac wires from PT are twisted but maybe not the extra primary leads. Think I just used heat shrink and let them in there. Same with the unused OT secondaries. Will need to check that. Only a single pole switch for mains and protected with heat shrink over the solder joints on the input jack and switch.

So the OT ground and the speaker jack might be the problem.
Will of course check all your tips. THANKS!

2) That's not quite the same thing. The PT CT needs to connect to the actual first filter cap, as close to the body of the cap as you can reasonably get it, not just to a point that the first cap negative connects to. The difference is that the big pulses of current feeding the first cap from the rectifiers travel through any wire between the CT feed and the first cap negative and induce V = I * R pulses. The longer the wire between the first filter cap and the CT wire, the bigger the V = I * R pulses. Your signal ground can be on a wire to a terminal block removed from the first filter cap and only see the V = I * R losses of the DC and signal currents to/from the first filter cap. This removes the cap/rectifier pulses from wobbling the signal ground.
4) Good! That's something a lot of people miss. In an ideal world, there would be only two connections to the chassis: one for safety ground, and one for signal ground. This leaves the chassis acting only as a shield.
5-6) You can get away with that wiring, and most people do. Having the OT common go only to the speaker jack "ground" confines the return current to the OT wire, not letting it run through the chassis. Not ideal, but then we don't live in an ideal world. It will work until you have every other source of noise and hum killed out.
7) "Common bus" is a poor idea from a grounding/hum view point. A bus implies that there's a long wire with "ground" tapped off along the way. That lets every current that moves to "ground'" cause a V = I * R loss in the wire resistance on its way back to the place that drives the current. The only way to get as close to a zero resistance ground as possible is with a star ground: the common bus contracts to a single point. This is massively difficult to wire up in a real world amp, so a more practical way is to make local circuit "clumps", where all the "grounds" for perhaps one tube or one gain stage are gathered together for a local ground for that piece of the circuit, including any tone/volume controls it drives. Then one wire per "clump" goes to the one star ground point. It localizes the ground offsets to the single clump, and allows any cancellation of ground offsets to take place local in the clump.
8 ) HT CT connect to the first filter cap negative terminal as close to the body of the cap as you can practically get it; Filament CT is a reference voltage, and should carry no current. It would be much better to run the filament CT on a wire to the central star point. The output tube cathodes have high currents flowing. Any wire between them and the first filter cap negative terminal has a V = I * R voltage across the wire due to the current through it. The output tube cathodes have to have current flowing out of them, the filament CT does not, so it's better from a hum and noise standpoint to run the heater CT and output tube "grounds" on separate wires to the One True Ground point.
9) same comments - ground bus introduces the possibility of noise/hum; single ground >point< removes many possible ways to let hum/noise in

Congratulations on having already thought about many of the issues on your own. Low noise wiring in tube gear is HARD.

[Jerry garrcia]

Oh I'll add I didn't see you've got an LND mosfet driven reverb, that maybe another source of hum? that's 'new' to the amp.

~Phil

I thought so to so I disconnected the LND driver without any success in reducing hum. “Working” from home tomorrow so will try to isolate that speaker jac.

[Jerry garrcia]

1) will measure for real. Sounds like a 50Hz hum but even the sun has its spots.
2) yes. CT connects to the same point as the first filter cap connects to.
4) main safety earth connects to a single point in the chassis. Nothing else is connected to that place. One concern I have is that the power jack is placed between both 6L6’s. Made a mistake and was aiming for the place by the rectifier.
5-6) three since I’m using unisolated switch craft jacks and the speaker jack is the ground for the OT. Need to fix that but no isolation “rings” at home…
7) all pots are connected to the common ground bus except fire the tone control that is without a ground wire from one of the lugs.
8 ) HT CT on negative side of the first filter cap. Filament ground on 2x100R to each of the cathodes of the output tubes.
9) goes to ground bus.
10) thanks. The ac wires from PT are twisted but maybe not the extra primary leads. Think I just used heat shrink and let them in there. Same with the unused OT secondaries. Will need to check that. Only a single pole switch for mains and protected with heat shrink over the solder joints on the input jack and switch.

So the OT ground and the speaker jack might be the problem.
Will of course check all your tips. THANKS!

2) That's not quite the same thing. The PT CT needs to connect to the actual first filter cap, as close to the body of the cap as you can reasonably get it, not just to a point that the first cap negative connects to. The difference is that the big pulses of current feeding the first cap from the rectifiers travel through any wire between the CT feed and the first cap negative and induce V = I * R pulses. The longer the wire between the first filter cap and the CT wire, the bigger the V = I * R pulses. Your signal ground can be on a wire to a terminal block removed from the first filter cap and only see the V = I * R losses of the DC and signal currents to/from the first filter cap. This removes the cap/rectifier pulses from wobbling the signal ground.
4) Good! That's something a lot of people miss. In an ideal world, there would be only two connections to the chassis: one for safety ground, and one for signal ground. This leaves the chassis acting only as a shield.
5-6) You can get away with that wiring, and most people do. Having the OT common go only to the speaker jack "ground" confines the return current to the OT wire, not letting it run through the chassis. Not ideal, but then we don't live in an ideal world. It will work until you have every other source of noise and hum killed out.
7) "Common bus" is a poor idea from a grounding/hum view point. A bus implies that there's a long wire with "ground" tapped off along the way. That lets every current that moves to "ground'" cause a V = I * R loss in the wire resistance on its way back to the place that drives the current. The only way to get as close to a zero resistance ground as possible is with a star ground: the common bus contracts to a single point. This is massively difficult to wire up in a real world amp, so a more practical way is to make local circuit "clumps", where all the "grounds" for perhaps one tube or one gain stage are gathered together for a local ground for that piece of the circuit, including any tone/volume controls it drives. Then one wire per "clump" goes to the one star ground point. It localizes the ground offsets to the single clump, and allows any cancellation of ground offsets to take place local in the clump.
8 ) HT CT connect to the first filter cap negative terminal as close to the body of the cap as you can practically get it; Filament CT is a reference voltage, and should carry no current. It would be much better to run the filament CT on a wire to the central star point. The output tube cathodes have high currents flowing. Any wire between them and the first filter cap negative terminal has a V = I * R voltage across the wire due to the current through it. The output tube cathodes have to have current flowing out of them, the filament CT does not, so it's better from a hum and noise standpoint to run the heater CT and output tube "grounds" on separate wires to the One True Ground point.
9) same comments - ground bus introduces the possibility of noise/hum; single ground >point< removes many possible ways to let hum/noise in

Congratulations on having already thought about many of the issues on your own. Low noise wiring in tube gear is HARD.

Thanks for the inputs and comments! The Hum stuff is hard. Especially since, in this amp, everything is wired backwards from what I’m used to. Used to have the power section to the left and the inputs to the right. Did it the other way this time to get a more appealing look from the rear. Mistake since my spatial function isn’t that great.

Will try to connect the PT CT to the first filter cap on a common turret and then run it to the ground bus. If that doesn’t work I’ll try to separate the power section to a single ground point and the signal/gain section to the input jack. If that doesn’t fix it I’ll isolate the speaker jack and run the OT ground directly to the speaker jack.
If that doesn’t work I’ll install those dreadful shielded cables along the signal path. Really hate them.
If that doesn’t work I’ll bring my father in law’s oscilloscope home. How do one chase down the hum with that?

Since this amp isn’t a true clone of the ES-185 I’m thinking of removing the tone circuit and install a 5F2A tone stack for the instrument channel. This one is so and so…

[Stevem]

First lets start off a check out session with this.

Set your meter for low AC volts and note what you have for ripple voltage on the last filter node that feeds the preamp section.

This would be a indication of the amp's 100 hz power supply noise level.

Next go to the speaker jack and make the same test and note what the voltage is and report back with both numbers.

[Jerry garrcia]

First lets start off a check out session with this.

Set your meter for low AC volts and note what you have for ripple voltage on the last filter node that feeds the preamp section.

This would be a indication of the amp's 100 hz power supply noise level.

Next go to the speaker jack and make the same test and note what the voltage is and report back with both numbers.

7 mVAC on both. No tubes in.
A picture of the first filter cap and the PT CT/ground bus as it is with the hum. Haven’t redone anything. Unfortunately needed to work

41AA245D-DE68-4149-AF8A-4F392F65AA38.jpeg

[Jerry garrcia]

I made a mistake in a previous post. Since the PT has a heater CT I didn’t use the artificial CT on this build. Sorry for messing up info.
The PT has a HV CT and the heater CT and both are connected to the negative node of the first filter cap as it is now with the hum. As I have been reading the ripple current of 0.007VAC seams pretty low. No tubes in and 263VDC on the node that I measured the AC on.
One more thing. The power jack has a fuse in it and also the high voltage CT wire.

[Stevem]

Please redo the test with the tubes in.