Noise Issue on Note Decay

[FourT6and2]

This just started happening (as far as I can tell). Modern Hiwatt Super-Hi 50. Based on DR504, but with 4 cascaded stages and standard Marshall TMB tone stack. Has regular Hiwatt PI and power section.

After a note is struck, there is an electrical "crack" once the note decays long enough. And during the decay there's a fizzing/buzzing arcing sound as well, allmost like a Jacob's Ladder arcing. It's easier to hear during palm mutes. So in the video, I'm just tapping the strings with my finger since I'm holding the camera with my other hand. I've verified my setup (cables, power, guitar, cab, etc.) with other amps and no issues. I've tried a Furman voltage regulator and a Furman power conditioner to make sure it's not a mains issue.

Filtering problem?
Tube problem?
Signal/coupling cap?

I have no idea where to look. Don't have a scope. But I do have new tubes and new filter caps on the way just in case.

Video:
https://www.youtube.com/shorts/RwOfVFCYtSs

[solderhead]

It's always good to post the schematic when requesting help so that we don't have to expend our time hunting it down.

[Stevem]

I have had similar issues like this.
A filter will not make a fuzz sound.

1) if you have a fuzz note showing up as the main note is decaying away that has always been a bad preamp tube in my experience.

2) in terms of a fuzz sound mixed in with a note you may also have a oscillation taking place due to your wiring or coupling cap layout.

3) A snap and crackle must be a bad connection either solder wise or tube pin to tube socket wise and most likely with a B+ connection being involved, this can include arcing output tube socket.


Don't rule out the possibility of having all 3 of these taking place at the same time in a new build.

[FourT6and2]

It's always good to post the schematic when requesting help so that we don't have to expend our time hunting it down.

I can PM you a photo, but I can't share it publicly. Here are some photos:

IMG_7883.jpg

IMG_6954.jpeg

[FourT6and2]

1) if you have a fuzz note showing up as the main note is decaying away that has always been a bad preamp tube in my experience.

No noise as main note/fundamental.

2) in terms of a fuzz sound mixed in with a note you may also have a oscillation taking place due to your wiring or coupling cap layout.

This isn't happening either... Listen to clip?

3) A snap and crackle must be a bad connection either solder wise or tube pin to tube socket wise and most likely with a B+ connection being involved, this can include arcing output tube socket.

I've gone through every solder joint in the amp that I can access, but I'll do it again.

[mhuss]

I may be oscillation, these sometime pop up during note decay on multi-stage high gain preamps. Do you have a scope?

[FourT6and2]

I may be oscillation, these sometime pop up during note decay on multi-stage high gain preamps. Do you have a scope?

I do not. But it sounds like legit arcing lol, not typical oscillation.

[FourT6and2]

I think I figured it out. Fortin Zuul+ noise gate.

When running the gate in front of the amp, no problem.
When running the gate in the loop, no problem.
When running the gate using 4 cable method (as it was designed) in both the loop and out front, using gate's Key input, the gate is causing this noise when it closes.
If I run two gates, one in the loop and one out front, no problem.

Glad it's not the amp!
But at the same time, I wish this thing wasn't so noisy to begin with. Then I wouldn't need a gate.

[Phil_S]

You may be using the noise gate to compensate for a ground scheme that doesn't do what it should. Cure the problem. Don't put a bandaid over it.

[FourT6and2]

You may be using the noise gate to compensate for a ground scheme that doesn't do what it should. Cure the problem. Don't put a bandaid over it.

Well yeah. But I can't completely rewire an amp without a concrete game plan. If you'd like to offer some suggestions, I'm all ears. This amp is inherently quite noisy.
The entire preamp is grounded to a single star point by the input jack/Gain 1 (on the ground side of the first cathode bypass). So everything on the board flows to the ground of V1a cathode, and then to chassis ground.
NFB has a low-pass after the presence circuit that connects to chassis ground by one of the 220uF filter caps. And all filter cap grounds are connected together to that same point.
I'm not able to clearly trace the PT center taps because they're inside wiring looms.

4 cascaded gain stages, with the 4th stage switchable (relays)
Single channel, with two switchable gains and masters
DC heaters via SMPS (probably contributing to the noise)

The main noise issues are:

1. With master volumes at zero, lots of noise. Hum, hiss, radio. Raise master volume so signal begins to pass and that noise vanishes. i.e. there's more noise with masters at zero than with masters at 1.
2. Connecting the TRS footswitch (for gain/channel switching) raises the noise floor considerably.
3. Low-frequency hum coming through the speakers. Master volumes, gains, EQ has no effect on this. Running a noise gate has no effect on this. I believe it's transformer hum that's somehow making its way to the speakers.

[Phil_S]

Without a schematic and layout diagram, it would be challenging to understand where all the grounds start and end, and where the go in between. You've chosen not to post these. This sets a limit on how much help will be offered. While the build is elegantly tidy, that does not mean tidy = quiet. Also, without actually being able to see the amp, it is hard to tell some things.

I suspect the low frequency hum is from the heaters. You should be able to lick this one. Also, it is easy enough to confirm 60Hz or 50Hz. If you aren't sure what that sounds like find an on-line tone generator.

For the heaters, for example, on the third tube socket from the left (yellow wire across the socket,) there is a neatly bent yellow wire crossing over the heater wires. I can't judge the vertical distance between the yellow and heaters. My instinct says to reroute the yellow so it does not cross the heaters or to bend it up at the tube socket to create more distance. I didn't see a faux CT with a pair of 100Ω resistors. Maybe I missed that. If not present, installing the faux CT might go a long way. BTW, don't combine that with the signal ground. Make a separate connection to the chassis.

For the filter caps, each one typically feeds a section or stage of the amp. The cathode(s) from each stage and any other related grounds should tie to the related filter cap ground.

There are many good resources on ground schemes, including on this forum. Search for them. There should be easy to find.

Do I see the mains ground in the lower left corner? The short green wire soldered to the chassis? Soldered to the chassis near the entry point is good. The short wire is not. In case the mains supply wire pulls out there needs to be extra length on the ground so it is the last wire to be stressed by the pull out. This is a safety issue. Also, it looks like the mains supply is mixed with the PT primary and secondary wires. This is not a good practice. The primary and secondary should not be running through the same hole.

Just because the amp is high gain and four stages, etc., does not mean you must accept excessive noise. A scope will help to see where noise originates so you can focus on the problem. You might also try isolating the noise(s) origin through the simple tube pulling diagnostic technique.

Good luck on reducing the noise floor.

[R.G.]

Well yeah. But I can't completely rewire an amp without a concrete game plan. If you'd like to offer some suggestions, I'm all ears. This amp is inherently quite noisy.

You might like reading https://www.dropbox.com/scl/fi/vpijj3al ... wq2wm&dl=0
which is a lot of suggestions for quiet wiring and grounding, with drawings and explanations.

The entire preamp is grounded to a single star point by the input jack/Gain 1 (on the ground side of the first cathode bypass). So everything on the board flows to the ground of V1a cathode, and then to chassis ground.

That's an OK start, but other issues can remain even with that starting point. A better solution is often to ground the entire amp to the negative terminal of the first filter cap, isolate the input and output jacks from chassis, and run the jacks' "ground" wire to the circuits they connect to. This is for reasons explained in the article.

NFB has a low-pass after the presence circuit that connects to chassis ground by one of the 220uF filter caps. And all filter cap grounds are connected together to that same point.
I'm not able to clearly trace the PT center taps because they're inside wiring looms.

Ohmmeter on the ends of the wire loom??

1. With master volumes at zero, lots of noise. Hum, hiss, radio. Raise master volume so signal begins to pass and that noise vanishes. i.e. there's more noise with masters at zero than with masters at 1.

This tells you that the master volumes are changing the input impedance seen by the following stages, and probably that the following stages are going into RF oscillation with the MV at zero. Changing the MV setting changes the impedance and damps the oscillation. A second possibility is that poor ground-chaining is setting the whole thing up to oscillate, and the MV is changing the currents that get to/from the grounding scheme.

2. Connecting the TRS footswitch (for gain/channel switching) raises the noise floor considerably.

Another vote for ground induced oscillation if the footswitch jack uses chassis for grounding things.

3. Low-frequency hum coming through the speakers. Master volumes, gains, EQ has no effect on this. Running a noise gate has no effect on this. I believe it's transformer hum that's somehow making its way to the speakers.

Probably PT CT related.

The negative tab of the first filter cap is where all of the circuit power ground and signal ground current goes eventually, after winding its way through the circuit and wires (=resistors) in the grounding path. It helps this be quiet if this is also the place that all ground wires go directly. Grounding to the chassis at an input jack can be a quiet grounding scheme, but it carries with it some other requirements. Try this test: temporarily remove the jack single star point by the input jack/Gain 1 from the chassis. Now ohmmeter the chassis to that floating-in-mid-air ground. Does it really show no connection? If not, there are other sneaky grounds.

The other comments about possibly poor wiring techniques are also places to start looking.

There's an ad-hoc rule in design that says that anything will oscillate if you just give it enough gain. Four gain stages is a lot; it makes any wiring issues very touchy.

[R.G.]

Well yeah. But I can't completely rewire an amp without a concrete game plan. If you'd like to offer some suggestions, I'm all ears. This amp is inherently quite noisy.

You might like reading https://www.dropbox.com/scl/fi/vpijj3al ... wq2wm&dl=0
which is a lot of suggestions for quiet wiring and grounding, with drawings and explanations.

The entire preamp is grounded to a single star point by the input jack/Gain 1 (on the ground side of the first cathode bypass). So everything on the board flows to the ground of V1a cathode, and then to chassis ground.

That's an OK start, but other issues can remain even with that starting point. A better solution is often to ground the entire amp to the negative terminal of the first filter cap, isolate the input and output jacks from chassis, and run the jacks' "ground" wire to the circuits they connect to. This is for reasons explained in the article.

NFB has a low-pass after the presence circuit that connects to chassis ground by one of the 220uF filter caps. And all filter cap grounds are connected together to that same point.
I'm not able to clearly trace the PT center taps because they're inside wiring looms.

Ohmmeter on the ends of the wire loom??

1. With master volumes at zero, lots of noise. Hum, hiss, radio. Raise master volume so signal begins to pass and that noise vanishes. i.e. there's more noise with masters at zero than with masters at 1.

This tells you that the master volumes are changing the input impedance seen by the following stages, and probably that the following stages are going into RF oscillation with the MV at zero. Changing the MV setting changes the impedance and damps the oscillation. A second possibility is that poor ground-chaining is setting the whole thing up to oscillate, and the MV is changing the currents that get to/from the grounding scheme.

2. Connecting the TRS footswitch (for gain/channel switching) raises the noise floor considerably.

Another vote for ground induced oscillation if the footswitch jack uses chassis for grounding things.

3. Low-frequency hum coming through the speakers. Master volumes, gains, EQ has no effect on this. Running a noise gate has no effect on this. I believe it's transformer hum that's somehow making its way to the speakers.

Probably PT CT related.

The negative tab of the first filter cap is where all of the circuit power ground and signal ground current goes eventually, after winding its way through the circuit and wires (=resistors) in the grounding path. It helps this be quiet if this is also the place that all ground wires go directly. Grounding to the chassis at an input jack can be a quiet grounding scheme, but it carries with it some other requirements. Try this test: temporarily remove the jack single star point by the input jack/Gain 1 from the chassis. Now ohmmeter the chassis to that floating-in-mid-air ground. Does it really show no connection? If not, there are other sneaky grounds.

The other comments about possibly poor wiring techniques are also places to start looking.

There's an ad-hoc rule in design that says that anything will oscillate if you just give it enough gain. Four gain stages is a lot; it makes any wiring issues very touchy.

[FourT6and2]

Without a schematic and layout diagram, it would be challenging to understand where all the grounds start and end, and where the go in between. You've chosen not to post these. This sets a limit on how much help will be offered. While the build is elegantly tidy, that does not mean tidy = quiet. Also, without actually being able to see the amp, it is hard to tell some things.

Alright. Here's a schematic. It's mostly correct. No 47pF snubber in PI and PI plates are 91K/82K. The rest is fairly accurate.

Hiwatt 4-Stage.pdf

I suspect the low frequency hum is from the heaters. You should be able to lick this one. Also, it is easy enough to confirm 60Hz or 50Hz. If you aren't sure what that sounds like find an on-line tone generator.

It could be heater hum. But the amp has DC heaters. I can pull up a tone generator next to the amp to see. If anything, I think it's more likely PT vibrating. The hum doesn't get louder/softer with the master volume. It's always there at the same volume, regardless of amp volume.

For the heaters, for example, on the third tube socket from the left (yellow wire across the socket,) there is a neatly bent yellow wire crossing over the heater wires. I can't judge the vertical distance between the yellow and heaters. My instinct says to reroute the yellow so it does not cross the heaters or to bend it up at the tube socket to create more distance. I didn't see a faux CT with a pair of 100Ω resistors. Maybe I missed that. If not present, installing the faux CT might go a long way. BTW, don't combine that with the signal ground. Make a separate connection to the chassis.

I'll look into it. Moving that wire with a chopstick does nothing. And if there is heater hum, it's the least of my worries. The noise issues with this amp are much more than just the underlying hum. There's all sorts of static, beeping/morse code, swishing/swooshing/ocean noises, etc. Plugging in the footswitch causes even more noise. Using the loop causes more noise, etc. It's just a bad design in terms of grounding and noise bucking it seems. I've had two of these amps. So I know it's not unique to this specific one as both amps did it.

There are many good resources on ground schemes, including on this forum. Search for them. There should be easy to find.

Yes, I know plenty of ground schemes. I didn't build this amp, though. Hiwatt did. And I don't think I can rip it apart to extensively redo the grounds. But I'll look into it.

Do I see the mains ground in the lower left corner? The short green wire soldered to the chassis? Soldered to the chassis near the entry point is good. The short wire is not. In case the mains supply wire pulls out there needs to be extra length on the ground so it is the last wire to be stressed by the pull out. This is a safety issue.

It's an IEC connector. It's not going to pull out.

Also, it looks like the mains supply is mixed with the PT primary and secondary wires. This is not a good practice. The primary and secondary should not be running through the same hole.

I'm not following. I'll post more photos so you can see everything better because I don't think I'm seeing what you're seeing.

Just because the amp is high gain and four stages, etc., does not mean you must accept excessive noise

I totaly agree!

A scope will help to see where noise originates so you can focus on the problem. You might also try isolating the noise(s) origin through the simple tube pulling diagnostic technique.

Unfortunately, I don't have access to a scope. But I'll start pulling tubes.

[FourT6and2]

That's an OK start, but other issues can remain even with that starting point. A better solution is often to ground the entire amp to the negative terminal of the first filter cap, isolate the input and output jacks from chassis, and run the jacks' "ground" wire to the circuits they connect to. This is for reasons explained in the article.

Jacks are already isolated. I can try moving that star ground to the first filter cap. Easy enough to lift that start point and run a temporary wire.

Ohmmeter on the ends of the wire loom??

Will try

1. With master volumes at zero, lots of noise. Hum, hiss, radio. Raise master volume so signal begins to pass and that noise vanishes. i.e. there's more noise with masters at zero than with masters at 1.

This tells you that the master volumes are changing the input impedance seen by the following stages, and probably that the following stages are going into RF oscillation with the MV at zero. Changing the MV setting changes the impedance and damps the oscillation. A second possibility is that poor ground-chaining is setting the whole thing up to oscillate, and the MV is changing the currents that get to/from the grounding scheme.[/quote]

Have a look at the schematic I just posted. Maybe that will help illuminate something?

2. Connecting the TRS footswitch (for gain/channel switching) raises the noise floor considerably.

Another vote for ground induced oscillation if the footswitch jack uses chassis for grounding things.

I'll double check!

3. Low-frequency hum coming through the speakers. Master volumes, gains, EQ has no effect on this. Running a noise gate has no effect on this. I believe it's transformer hum that's somehow making its way to the speakers.

Probably PT CT related.

I'll check!

The negative tab of the first filter cap is where all of the circuit power ground and signal ground current goes eventually, after winding its way through the circuit and wires (=resistors) in the grounding path. It helps this be quiet if this is also the place that all ground wires go directly. Grounding to the chassis at an input jack can be a quiet grounding scheme, but it carries with it some other requirements. Try this test: temporarily remove the jack single star point by the input jack/Gain 1 from the chassis. Now ohmmeter the chassis to that floating-in-mid-air ground. Does it really show no connection? If not, there are other sneaky grounds.

Good idea!