Preamp Ground Bus Hum

[gabasa]

And now for something completely different.

How does the soldering on this look to you?

View attachment 973324

I don't see a problem ... It looks great?!?

For the amount of time i spent inspecting every solder joint with my magnifying lamp, I can't believe I missed that, so thanks for spotting it. I might try to redo that connection altogether: Remove the solder, poke a thin wire through both holes and wrap around the posts, then solidly solder both in.

 

[gabasa]

Pre amp Common is not connected to PT CT when there is no wire between 3rd and 4th filters. Then pre amp current flow to chassis thru amp input and when there is ”power amp ground lug” the pre amp current return to PT CT thru it.

100% I'll connect the two and listen to the difference, thanks for the explanation.

 

[gabasa]

Does anyone know why on Rob's website, or any other for that matter, the Presence pot has its own ground point? Is there a theoretical reason that explains this, and is it ok in theory to ground it to the preamp ground bus?

 

[gabasa]

Ok, for anyone that's read through Blencowe's grounding chapter, I just did and a couple of things came to mind:

First off, regarding that fourth 20uf capacitor, I've connected it to the ground bus physically in the area between the volume pots and the input jacks. If I look at the schematic, wouldn't it be smarter to connect it to the left side of my ground bus, near where the presence control is grounded?

Second, why bother connecting that fourth filter cap to the ground bus at all? That current coming from the power supply is using the input jack ground as its return path. Isn't that bad?

 

[schmee]

The under side of the BF amps is screen like screen door screen. The SF started using metal sheet at some point. Both work fine.

 

[Lowerleftcoast]

Is there a theoretical reason that explains this, and is it ok in theory to ground it to the preamp ground bus?

I think it is just a throwback to old designs. On this design, it should theoretically connect with the last 20uF cap node. Iow, on Rob's layout, it can connect to the Mid pot terminal #1 rather than the back of the pot.

First off, regarding that fourth 20uf capacitor, I've connected it to the ground bus physically in the area between the volume pots and the input jacks. If I look at the schematic, wouldn't it be smarter to connect it to the left side of my ground bus, near where the presence control is grounded?

Following Blencoe's grounding chapter, each filter cap denotes a node. The only ground associated with that node is the ground of the presence pot. Since that ground will connect at the furthest left end of the ground bus that is where the 20uF should be.

Second, why bother connecting that fourth filter cap to the ground bus at all? That current coming from the power supply is using the input jack ground as its return path. Isn't that bad?

There is extra filtering from each cap. You have chosen Rob's two point ground scheme. If you decide to follow the advice from @2L man , your ground scheme will be more like what Blencowe has suggested, just one point at the input jacks.

All of the DC current returns to the transformer through the HT CT. In a two point ground scheme some of it flows through the chassis. It is not necessarily bad, but there is more current flowing in the chassis. More current flowing near the input can make for more noise but many many amplifiers are quiet because it is only noisy if the noise is picked up and amplified by the tube.

 

[gabasa]

This discussion has been amazing for me. I'm seeing the grounding and nodes differently now when I look at the 5F6-A schematic. I need to roll my sleeves up and do some testing, but as I said, with a 2 ohm OT secondary, I'll wait until my cabinet is done so that I have a proper load connected to the amp. I also have to fix that input jack solder.

I'm really thankful for all the help here.

 

[gabasa]

Pre amp Common is not connected to PT CT when there is no wire between 3rd and 4th filters. Then pre amp current flow to chassis thru amp input and when there is ”power amp ground lug” the pre amp current return to PT CT thru it.

Hi @2L man, I think I understand what you’re saying.

For safety, chassis should only connect to earth and input jack grounds, and never used for return currents back to power transformer CT. Return current ground lug should be isolated from chassis. Is this correct? If yes, then I learned something today because it makes sense.

However, I do have a question. With this grounding method, do you isolate the speaker jacks? Where do you normally connect the OT secondary ground?

 

[King Fan]

I think i will, asap. Thanks.

Great. Link is here if you don't have it.

Word to the wise? Merlin is brilliant, but he discusses many small shades of 'best' in best practice. He admits a *lot* of quiet amps violate some of his guidelines. I'm just saying I don't think your hum is necessarily from ground, or at least from violating his 'ideal' guidelines. But the section on the 'safety earth' is basic, necessary, rock-bottom solid. And even there, if you don't have an IEC power entry, you want to add Rob's long loop of green ground wire (so that's the last wire to come loose when your amp falls off the stage at Woodstock).

I'm not gonna speak for 2l man, who does beautiful amp builds, BTW. I know he opposes all split grounds everywhere, based on solid ground theory. Hey, it's possible a single ground could cure your hum. It might not even be all that hard to try. I'm just saying that thousands of split-ground amps are quiet, so an organized hum hunt makes more sense to me than a whack-a-mole game of one attempted fix after another.

 

[Paul G.]

Look, the idea is to try to eliminate multiple paths to ground. Any of the commonly used methods can do that, single ground, split ground, star ground, brass plate, etc. As long as your components are grounded in a manner that promotes a single flow, you're alright.

 

[2L man]

Hi @2L man, I think I understand what you’re saying.

For safety, chassis should only connect to earth and input jack grounds, and never used for return currents back to power transformer CT. Return current ground lug should be isolated from chassis. Is this correct? If yes, then I learned something today because it makes sense.

However, I do have a question. With this grounding method, do you isolate the speaker jacks? Where do you normally connect the OT secondary ground?

I use isolated loudspeaker jacks because they are isolated and cheaper than Switchcraft. I install a shielded NFB cable which shield connect loudsleaker Common to amp Common.

Chassis Safety Earth is safety feature but Safety Earth is used for "shielding" and "killing" noise from instrument signal because it is easily available. Thats why it is bad idea to use part of that to flow operative current!!! You guys are known to have lots of freedoms Here in Finland it has been forbidden in "electric law" about 50 years!

 

[gabasa]

The more I read through this, the more I'm amazed at how bad my grounding has been. Last night I read through dozens of 2L man's posts, and those ideas have really sunk in, so I'd like to try them.

I'll report back, because after some investigation, every amp I have in the house has return current flowing through the input jacks and chassis back to the HT CT. This is just the way a lot of layout diagrams are presented online so I never questioned them. I really want to see the difference it makes when this is avoided, on one amp.

To begin, for one of my new Bassman builds:

I'll connect together all the filter cap grounds; none of the negatives will be separated.
I'll try to make sure that what I've read from Merlin's chapter is implemented to the best of my abilities.
I'm lifting the ground terminals on the power amp side of my build from the chassis.
I'm going to get an isolated speaker output jack, or at least Switchcraft isolation washers.
Only the main power cord green wire and the input jack grounds will be connected to the chassis.
If it all heads in the right direction, I'll isolate three of the four input jacks.

These are easy changes to implement. Let's see what happens!

 

[Lowerleftcoast]

I also have to fix that input jack solder.

I, for one, am very interested if the hum is because of this solder joint. Like King Fan, I am skeptical the ground scheme is giving you this hum.

When diagnosing hum problems it is helpful to know the frequency of the hum. Hum from the ground scheme should be at 120Hz or double the mains frequency. Hum from that solder joint would be 60Hz. Hum from not having a Faraday cage, (no conductive chassis back panel), would be 60Hz. Hum from the 6.3VAC heater would be 60Hz.

Where do you normally connect the OT secondary ground?

The speaker jack should only have DC on it when a global negative feedback is connected. The 5F6A has this negative feedback circuit. If you are keeping the nodes together as proposed by Blencowe, the speaker ground would be with the last 20uF cap. It is the same node employed by the presence pot. The B+ side of this node connects with the plates of the phase inverter.

 

[gabasa]

I, for one, am very interested if the hum is because of this solder joint. Like King Fan, I am skeptical the ground scheme is giving you this hum.

When diagnosing hum problems it is helpful to know the frequency of the hum. Hum from the ground scheme should be at 120Hz or double the mains frequency. Hum from that solder joint would be 60Hz. Hum from not having a Faraday cage, (no conductive chassis back panel), would be 60Hz. Hum from the 6.3VAC heater would be 60Hz.

The speaker jack should only have DC on it when a global negative feedback is connected. The 5F6A has this negative feedback circuit. If you are keeping the nodes together as proposed by Blencowe, the speaker ground would be with the last 20uF cap. It is the same node employed by the presence pot. The B+ side of this node connects with the plates of the phase inverter.

I must have soldered in that joint after I took the photo because I just looked at the amp and it's nicely filled with solder. Whew.

I plugged in the amp and fired it up. The hum is 120 Hz because the note is between B and Bb, so based on your comment, it's a ground issue.

 

[2L man]

There is not much difference to hum and noise despite the "grounding" when not playing and when you play you don't hear any hum and noise because signal is amplified. The reason it is forbidden to use safety earth (ground) to flow operative current is because it cause noise to other devices which use the same Mains. Some of your guitar comes thru other band member amplifiers and it their amps use chassis for operative current some of them come thru your amp and thru PA etc... even your amp is built correct way.

Because there is no advantage to use chassis for signal amplifying why to use it?

 

[King Fan]

Isn't it nice how knowing the frequency gets rid of most of the hum suspects? Especially if the hum is 120Hz, which points at filtering. An easily treated cause is reservoir (first cap) ripple getting onto the grounds -- Merlin notes we can reduce this up to five-fold simply by anchoring the HT center tap on the reservoir negative pole (which is still also anchored to ground -- I forget if you did this already?) And of course: Is your household safety ground on a shared ground anchor?

Other filter problems: On a new build from a reliable schematic, chances are the filters and ground scheme are designed right but functioning wrong, and leaky caps aren't likely, so a wiring or soldering fault is more likely. Backside ground traces between various caps and nodes are especially suspect, and with your hum appearing before the volume pot, I'd look extra hard to make sure I split the grounds in the doghouse exactly like Rob shows. Oh, and where do you ground your presence pot?

*But* it's not always filter failure per se; the problem may be due to excessive current draw of B+ somewhere. Filter cap values are chosen for the expected load, so excessive load will cause more ripple. Thorough review of all voltages can help.

(EDIT: The following was written before I saw you're going to go to a single ground scheme. Good idea anyway, should be fun -- if it cures the hum, so much the better). Finally, if we confirm the HT CT is right, the household ground is mechanically and electrically separate, the ground soldering, wiring, and anchoring are good, and voltages are right across all sockets, for sure you could try a single bus anchored at the input end. As 2l man says, in theory it's better. But I don't start there since tons of split-bus 5F6as and even random-ground originals are "quiet."

 

[2L man]

Obviously part of tube amp hum transfer from power transformer to output transformer inductively when filter capacitors charge and PT pass the current. To lessen this there is a "headphone trick" where transformer placing and oriention on chassis is optimized connecting headphones to OT and Mains to PT.

 

[gabasa]

I checked all the connections today, they seemed good....

 

[gabasa]

PROBLEM SOLVED!!! This is what I did:

Isolated the power amp grounds onto a terminal strip by the power transformer.
Connected my ground bus to the isolated ground with a wire.
Replaced the speaker jack with an isolated Switchcraft jack.
Grounded the speaker jack negative to the terminal strip.

Now, the circuit only connects to the chassis at two points: the earth ground and the input jacks.
All the noise completely vanished and the amp is now incredibly quiet! I'm going to head over to my build journal and post an update.

 

[King Fan]

As the guy who somewhat doubted a single ground would cure it, I was wrong. Single ground is better all the time in theory, and so can be better in practice whenever a ground loop happens to be noisy. Isolated speaker jack but not isolated input jacks, eh? And your amp has NFB, and that'd be a long loop, so that may also make sense.