resonant frequencies in high gain amps

[marshallguy]

Hello everyone,

With the all the pots at 10, preamp, vol, eq, on the OD, with fluke on hz setting, no input, I'm getting a resonate freq around 2.6k +/- 100hz, might be what I'm hearing under the notes while playing the guitar.
It shows up in recordings and a lot of notching to try and limit it hurting the overall tone.

Any other ideas?

"failure is not an option”

[sluckey]

white on black background hurts my eyes.

[PRR]

white on black background hurts my eyes.

It  is also huger than my screen or eyes. Flipped and sized.

[tubeswell]

So did you build something?, or is this a factory amp?, or a modded amp?

[marshallguy]

So did you build something?, or is this a factory amp?, or a modded amp?

I was stationed in the U.K. seems like a life time ago not far from the Marshal factory. Became friends with a lot of them there. Ken Brad gave me this combo which was sitting in his foyer at home while I was there to pick up 3, JCM800 2203 100w heads he gave in exchange for a mini stereo system I got for him at our audio club back on base. It was their first channel switching amp and a combo, JCM800 4210. The schematic floating around the net is close but a bit off, or at least to this one. Nice thing was the KT77 Golden Lyon tubes swapped in which I still have. Ken told me I just missed Eric Clapton’s manager whom dropped off the combo. Word was he was using it in a studio but was now under contract promoting Music Man amps and under the contract, he wasn’t allowed to use other manufacturer amps. He said it was used but you can have it too if you want it. That’s the story and I have no reason why he’d make up such a story.

 Later on the 2205 which is the head only has the updated, racially changed final circuit design to the 4210. Pull up the schematic and compare it to mine. It’s a completely different animal now.

[pdf64]

It looks the nice design, lots of nice touches, eg heater elevation.

Is the 2k6Hz resonance with all control on full definitely something new, or might it always have been that way?

If the former, might there have been some coincidental changes when it became a problem, eg replaced some valves, adjusted bias, use of different cab, moved rig to new environment

[shooter]

With the all the pots at 10, preamp, vol, eq, on the OD, with fluke on hz setting, no input, I'm getting a resonate freq around 2.6k +/- 100hz,

where are you looking?  The speaker?
IF so;
set ALL the stuff to 5 and repeat


V3 is a good candidate for adding stuff you don't want when driven into clipping.  The vale Wizard has a good right-up on DCCF n the fixes 

[marshallguy]

It looks the nice design, lots of nice touches, eg heater elevation.

Is the 2k6Hz resonance with all control on full definitely something new, or might it always have been that way?

If the former, might there have been some coincidental changes when it became a problem, eg replaced some valves, adjusted bias, use of different cab, moved rig to new environment

Thank you, a lot of trail and error. If I can only figure out what’s causing this chiming riding over the guitar notes.
I was mastering a few songs recorded with my home studio and when I did some stereo widening, all sort of unnecessary artifacts came blooming out in the mix.This amp is so quiet now, it’s easy to hear everything.

I eliminated the tubes, grounding scheme, tried 250p across 82k P.I., then across the 100k cc at V5b, then a cross the snubber cap, 240p from power tube grids to grounds, tube dampers, tried Vox cut mod, redressed some wiring. have shielded audio grid wires, tried different cab, adj. for different biasing…..

What exactly it is I’m hearing has to have an explanation. Tube rattle, ghost notes, ringing, not sure what to call it but it comes out all over the guitar neck. Easier to hear when playing through the clean channel.

[marshallguy]

With the all the pots at 10, preamp, vol, eq, on the OD, with fluke on hz setting, no input, I'm getting a resonate freq around 2.6k +/- 100hz,

where are you looking?  The speaker?
IF so;
set ALL the stuff to 5 and repeat

V3 is a good candidate for adding stuff you don't want when driven into clipping.  The vale Wizard has a good right-up on DCCF n the fixes

Thanks, I’ll take a look.

I’m looking on the output of V5b with O.D. Channel On. Huge gain there. Seen it on the output jacks too really well only after turning up the master vol. and amplifying. Isolated the output is quiet and clean.

 I first notice something on the scope on the output of V2a, pre-amp around 5. These spikes over the noise, about 120hz apart, looking like a sign waves moving across the noise floor but their too short width wise to measure freq. Not sure what I’m seeing, but maybe what I’m hearing…

[marshallguy]

It looks the nice design, lots of nice touches, eg heater elevation.

Is the 2k6Hz resonance with all control on full definitely something new, or might it always have been that way?

If the former, might there have been some coincidental changes when it became a problem, eg replaced some valves, adjusted bias, use of different cab, moved rig to new environment

******I found a trend about filament ringing or something to that effect and Kevin O’Conner suggested heater elevation of 70 ~ 80v. These guys were going on about sometimes at a lower voltage 35 ~ 50v isn’t enough. I have 50v right now so going to change a resistor and try around 80v too, maybe that will cure it.******

[tubeswell]

A standard 6-string 21-fret guitar fretboard in standard tuning goes up to about 1kHz. So it's possibly a harmonic resonation of that.


Have you checked for
1) a microphonic tube?
2) a slightly loose input jack or FX-loop jack tip-switch contact?

[marshallguy]

A standard 6-string 21-fret guitar fretboard in standard tuning goes up to about 1kHz. So it's possibly a harmonic resonation of that.


Have you checked for
1) a microphonic tube?
2) a slightly loose input jack or FX-loop jack tip-switch contact?

The guitar harmonic theory sounds interesting. If that’s the case….how do you cure that?

The guitars I tried are 24 fret if that makes a difference. Standard tuning. They’re both made by Carvin, DC160 (1980) solid eastern hard rock maple wood and the other mahogany with maple top (2002). Both lefty (I’m a southpaw) with ebony fret boards. Right now using the maple wood one for troubleshooting. I’ll try the mahogany again to compare.

The jacks look tight and solid, but I’ll place a short jumper cable in them from send to return just to be sure. I’m going to check again one by one the preamp tubes. V1 was microphonic and replaced earlier but problem persists. Sounds like a tube rattling higher pitch resonance ringing on the note played but now have those big silicone tube dampers on them all by Euro tubes. It didn’t cure the problem.

I upped the dc elevate to 68v, Kevin O’Conner suggests 70 ~ 80v for complete noise suppression so I figured wtf try it but honestly the amp is already hum free and quiet. Bumping it about 20v more volts didn’t change anything. Changed the tail resistor call it R2 from 33k to 47k but 56k would get me closer to 75v. The cap filter there is 100v/100uf , so either I have to change that too, 150v more or move the connection pass the Standby switch to keep DC there below 100v incase those Zeners ever shorted, I calculated just above 100v on power on (previously measured with 33k installed 585vdc B+ Stby off with Zeners bypassed @ 125vac input).

Of course in that scenario the capacitor exploding at power on would be my saving Grace not to turn on Stby thus save my output tubes from screen meltdown. 

I’ll keep you posted on results.

“Failure is not an option”

[shooter]

a sign waves moving across the noise floor

I've called them many things, gremlins, creepers, &&^&%^$ , they are coherent signals creeping along the power wires.  they typically get "induced" or "coupled" there from iffy power connections, iffy grounds, strong signals in || with power/ground wires.


can originate from any source, the "noisy" air around the system to sketchy parts.  most are harmless, some are catastrophic, depending the system they get into.
If you get paid by the hour to kill 'em, you'll have a new Bass boat in no time.


find your 2.6Khz at the speaker, then work to eliminate it in the system, if you monitor it "inside" the problem, then try and fix it "inside" the problem, you pull out lots of hair and speak Sailor

[marshallguy]

A standard 6-string 21-fret guitar fretboard in standard tuning goes up to about 1kHz. So it's possibly a harmonic resonation of that.


Have you checked for
1) a microphonic tube?
2) a slightly loose input jack or FX-loop jack tip-switch contact?

Well I swapped each tube one at a time, tried different type P.I. tubes, 12AY7, 5751, etc, no change.

Before I retired I maintained Navcomm, communications and some radar systems for the FAA. One thing I learned was when you have a problem, sometimes you have to regroup and start over. It’s easy to go down the wrong path and miss something.

I decided to try this, my other amp, a modded Marshall 2203 has an almost identical preamp, only different is the grounding scheme and wiring dressing. Other differences are I’m using a discrete power chip to drive and 2N2222A to recover reverb instead of a 12AT7 & transformer. Same reverb tank being used in both. The last difference are a pair of EHKT90’s instead of the 6CA7/EL34’s and slight difference in power tube grid resistor size, 120k instead of 150k.

Ok, I wanted to eliminate the output power section as the cause so I connected a short cable from 4210 FX Loop send to 2203 return using just the preamp up to the FX Loop of the 4210 and power amp section of the 2203 (this includes V4A/B too and reverb). Result? No change.

Ok, now I reversed the connection scheme and used the 4210 power section and 2203 preamp section via FX Loop send & return. Result? No real change but slightly a little less ringing but that could just be the EQ settings. So either both amps are fooked or it’s the cabinet.

I’m now focusing on the speaker cabinet. I used the combo open cabinet with both amps on the bench. It’s a 12 inch Celestion T3135 / 8 ohm G125 - 125W speaker. I have the extension cabinet for that combo too with the identical speaker.

Tomorrow just for chits and giggles I’ll repeat the test only with the extension closed back cabinet. My hope is it’s the open back combo only and then I can work it out from there.

I’m also reinstalling the 33k DC elevate R2 resistor. 50vdc elevate is good enough and works well.  Just because Kevin O’Conner thinks you need 70 ~ 80v doesn’t always makes it so….

[marshallguy]

a sign waves moving across the noise floor

I've called them many things, gremlins, creepers, &&^&%^$ , they are coherent signals creeping along the power wires.  they typically get "induced" or "coupled" there from iffy power connections, iffy grounds, strong signals in || with power/ground wires.


can originate from any source, the "noisy" air around the system to sketchy parts.  most are harmless, some are catastrophic, depending the system they get into.
If you get paid by the hour to kill 'em, you'll have a new Bass boat in no time.


find your 2.6Khz at the speaker, then work to eliminate it in the system, if you monitor it "inside" the problem, then try and fix it "inside" the problem, you pull out lots of hair and speak Sailor

My friend, I agree with your assessment 100%. Tomorrow I’m going to open up my other amp with same preamp and check for those gremlins. If I see them, not going to worry about it. If I don’t? Then it’s back to looking for iffy grounds, coupling signal, dressing, etc. I soldered a lot of components directly to the tube sockets. Wither or not they’re too close to one another in the chassis only time will tell, but at least I’ll have a reference.

Thanks to everyone for your on going support. Cheers!

[tubeswell]

looking for … coupling signal, dressing, etc. I soldered a lot of components directly to the tube sockets.

If this is factory PCB construction, and you’ve modded it, and the noise wasn’t there before you did the mods, then the noise is occurring because of something you did.


PCB designs from factories should ideally be designed with a layout (parts layout and trace layout etc) to address unwanted signal coupling. If you mod it to ‘bypass’ the PCB traces and jumper cables etc (by shifting components and soldering directly to tube socket pins), you could be leaving all sorts of ‘antennas’ behind that cause unwanted coupling. Moreover, if your modding got a little rough (and you inadvertently tore some trace pads of caused minute hairline cracks in the traces when removing parts) you could create sources of noise.

[marshallguy]

looking for … coupling signal, dressing, etc. I soldered a lot of components directly to the tube sockets.

If this is factory PCB construction, and you’ve modded it, and the noise wasn’t there before you did the mods, then the noise is occurring because of something you did.


PCB designs from factories should ideally be designed with a layout (parts layout and trace layout etc) to address unwanted signal coupling. If you mod it to ‘bypass’ the PCB traces and jumper cables etc (by shifting components and soldering directly to tube socket pins), you could be leaving all sorts of ‘antennas’ behind that cause unwanted coupling. Moreover, if your modding got a little rough (and you inadvertently tore some trace pads of caused minute hairline cracks in the traces when removing parts) you could create sources of noise.

I’ve been exploring that scenario doing my best to isolate the grounds and been working on top of the board, especially since it was the type with the pots soldered to it. I over time removed the pots from the board and point to point soldered them. I also used a bit of shielded cable for longer grid runs. Most of the component are either soldered to the tube sockets or on top of the board. I’m using 5vdc to operate relays, LED’s and LDR’s but there is no ripple what so ever I filtered the heck out of them.

I pulled V1 and the stuff riding on the noise disappeared. I decided to redo my grounding and got it really quiet but still have the spikes. Removed the DC elevate and 120 becomes 60hz spacing spikes. I zeroed the humdinger evened the filaments but still had some slight hum. Reapplied the DC elevate and hum completely gone and no 120 buzz either. Problem isn’t noise, it’d that damn ringing riding the notes. My theory is if I can remove those spikes or reduce them into the noise floor that I’m seeing on my oscope then maybe the ringing stops too. I also tried a close cabinet speaker but the ringing persists.

Since my other amp has a similar set up and heard some ringing too, I was looking for what I might have done to both to cause it. My guess it’s the grounding and one might be a little better done than the other but in reality they are both very quiet hum or buzz wise.

It also could be how I did the preamp B+ nodes for V1 -  3 and looking to redo that and see if that improves it.

[shooter]

I pulled V1 and the stuff riding on the noise disappeared.

it’d that damn ringing riding the notes.

^^^ so those are 2 different things you're seeing, or the same thing?

[marshallguy]

I pulled V1 and the stuff riding on the noise disappeared.

it’d that damn ringing riding the notes.

^^^ so those are 2 different things you're seeing, or the same thing?

Hi, one was with no input just static noise, the other I see when I pluck or pic a note. I’m trying to determine if the 2 are related.

 Amp in ringing over any note I play.

Injecting 150mv @500hz what I’m seeing is the pure sign wave has this jagged rippling look to it. I removed the 180p Clean Ch. MV bypass treble peaking cap and the the volume at less then 2 there is improvement but not until I back the treble pot down to below 2 does it disappear. I also removed the 4.7uf cap @ V4a temporarily to reduce gain for now but it seems be near the input where the problem starts. I thought I cured that problem with the 390p to ground but injecting 1000hz instead of 500hz made it harder to detect.

[shooter]

just static noise

words matter, this tells me "there's nothing there" static noise = noise floor, normal in all electronics.
now if noise has coherency, "repeating pattern inside the noise", that's "a signal"


by pulling V1 you have removed the most likely point where a "signal" can creep in at V1a b, you also broke the large signal path by removing V1b a.

the other I see when I pluck or pic a note

I'm assuming you're monitoring at the same point with both "problems"

can you hear the problem out the speaker, or are you just seeing it?
How physically close is the clean/boost transistor that feeds V1a's cathode to V1a's socket?

[marshallguy]

the other I see when I pluck or pic a note

I'm assuming you're monitoring at the same point with both "problems"

can you hear the problem out the speaker, or are you just seeing it?
How physically close is the clean/boost transistor that feeds V1a's cathode to V1a's socket?
[/quote]

Yes, same point. Agree now with just static noise because the amp is quiet, hum free and low hiss.

The transitor is several inches away, on the board, tubes are separate socket mounted.

I snubbed a 500p from plate to cathode on V1B…no change. Next up, across the 220k plate. If that doesn’t work, going to lower plate to 110k, parallel another 220k, both metal and see if that changes things. Filter node cap checks ok. Picking up some CRC contact cleaner in the morning and cleaning the sockets and tube pins.

Any idea what’s causing this? 500hz @150mV injected at input, looking at the audio sign wave, what I’m seeing instead of a nice pure 500hz sign wave, it is wavy or a bit jagged looking. Get’s worse with more treble added. Using polystyrene 630vdc coupling caps there, both 250p tone stack and 180p / 330k treble peaking. Suppose to be high quality audio caps. Suppose to….

[marshallguy]

So did you build something?, or is this a factory amp?, or a modded amp?

I found the source of the problem and want to close this discussion and open a new one with the appropriate diagnosis. Found injected audio from generator oscillating on the plates of the output tubes only.