Hum and Buzz in Cordovox Build

[scstill]

I have a frustrating noise issue, both Hum and Buzz in my Cordovox Conversion project
The build is somewhat complicated in that there are three chassis and an umbilical cable connecting them.
I have tried everything - haha - except for the solution. Hoping you might provide some wisdom.

What "good practices" I have done...
1a) all signal cables are coax, grounded only one end
1b) all electrolytics replaced, some coupling caps replaced (others have no dc leakage)
2a) all grounds are star with only one chassis connection for each star
2b) some of the original design grounds were kept (indicated in ground schematic)
3) good cable layout, twisting, separation
4) Hum increases with MV  Hum with buzz; Buzz increases with Vol and Tone
5) design calls for 3.2 preamp stages 16v on 7591 grid/ 50mv guitar input = 320

What I have done to troubleshoot
6a) tried different tubes except for 7591 (no spare)
6b) Bypass umbilical with Audio  - no change
6c) Bypass umbilical with AC line - no change
7) Bypass original jacks with another jack - no change, less buzz without PreAmp1 grid stopper, no change with lower grid stopper
8) Bypass PreAmp1 & Vol & Tone - no change, hum lower makes sense w/one less stage
9) Bypass PA2 & MV - no change, hum controlled by Vol
10)Bypass PA1 and PA2 and MV  hum low but so too is the audio,
11) bypassed the unusual speaker circuitry - no change
12) bypassed the 6v CT circuitry - no change

the diagram had the 7591 cathode cap incorrectly drawn edited below
the story of the build here   https://stillampd.com/cordovox-conversion

[shooter]

that's a lot to unpack!


got a scope?


drawing error on the cap?  or is it wired wrong like schematic?


add a 1 ohm R to the PA cathodes n verify current, not just grid vdc.


Is the Buzz audible inside the amp or just at speakers?

[scstill]

I really tried not to bring this to the forum, trying the stuff learned here, but I'm at wits end, not giving up...

The buzz is clearly at the speakers and increases with MV and to lesser extent with Vol. So getting into audio somehow, but cables are shielded. If it was a amp buzz like mechanical transformer plates would be present with audio down. Because MV controls hum and buzz it suggests that it occurs before MV and V but then when I bypass those stages and pots it is still present.

The cap is drawn wrong - thanks.

[scstill]

When the volume/MV is all the way down there is a very slight noise coming from the PT.
When I add finger pressure the noise changes, so there is some slight electromechanical noise from PT.
When I briefly disconnect speaker that "PT noise" goes away.
But would that slight noise be amplified to uncomfortable levels with volume up? Its really loud.
I cannot change the pitch of the audio noise at all with finger pressure on PT.

[tubeswell]

drawing error on the cap?  or is it wired wrong like schematic?

[scstill]

drawing error on the cap?  or is it wired wrong like schematic?

Cap is wired correctly, drawn wrong. Will change the diagram.

[shooter]

with amp on and annoying, measure both AC n DC on BOTH sides of the 100 ohm between rectifier n 1st PS tap.


do the ohms law for current across that 100 ohm.


post it all

[scstill]

100ohm
rectifier side: 0vac; 442vdc
1st tap side: 0vac; 428vdc
14v drop; 0.14amp; ~2w
no voltage change whether MV low or annoying

[Platefire]

Just some things to check:

check continuity form Signal 5 to B7 on 9 pin connector

Try disconnecting NFB

Check continuity on V2a from junction of 1m & 1.2K resistors to G4 Ground

[scstill]

Just some things to check:
check continuity form Signal 5 to B7 on 9 pin connector
Try disconnecting NFB
Check continuity on V2a from junction of 1m & 1.2K resistors to G4 Ground

Continuity check; checked all resistors on V2, all check fine. Voltages check fine too.
Disconnect NFB - when disconnected noise gets slightly louder

Interesting that even when removing preamp tube V1 the noise persists implying it is in power tube section
But with V1 removed MV controls the noise volume which implies kin the preamp section ???

[shooter]

you need to "standardize" your hums n noises


ie;
I have 60hz hum
I have white noise


otherwise I just glaze over thinking about the humming bird that buzzed me the other day and the noise the wings made 


I think you're fighting on 2 fronts, maybe 3 but

[scstill]

Sorry thought I described noise in the first entry...

There is both buzz (sharp like when you say buzzzzzzzz) and 60hz hum (mellow like when you hummmmm)
I think most of the buzz might be from LED lights as when I'm in the dark it is mostly gone. But the Hum is always present and increases with MV. Even though I'm plugged into AC noise filter. Have also tried outlets at the other end of house with same results.

The hum is noticeably reduced going directly into PreAmp 3, but that could be because two stage amplification is gone too.

BTW - the guitar tone is very nice. Just the hum is quite annoying.
bench setup below. All three chassis have ground continuity. I was careful to keep star grounds as shown in the ground schematic in first entry

[Platefire]

I don't know if this is contributing to the noise but it appears to me the 5E3 tone stack is wired wrong on the schematic?? Double check that anyway

[shooter]

Even though I'm plugged into AC noise filter.

60hz hum

the amps 60hz can couple into the signal path from INSIDE the system, a noise filter only keeps the outside noise out.
60hz is filaments "based" so anywhere filament wires and sensitive signal get close, crosstalk can happen, I know you used shielded wire, is it properly connected, grounded at one end only.


do filaments n signals pass through your connectors at pins close to each other?

[scstill]

Thanks for the ideas, I will check these ideas next week as I'm off to the mountains for the weekend.
yes signal lines shield are only grounded at one end.

[scstill]

I don't know if this is contributing to the noise but it appears to me the 5E3 tone stack is wired wrong on the schematic?? Double check that anyway

Why did you draw the vol pot before the tone pot? Doesn't 5E3 have the tone first?

Anyway I did correct this tone/vol circuit to match the 5E3. It did not make any difference to the hum.
I corrected this in the original post schematic.

Still hopeful for a solution...

[Platefire]

I'm Sorry!   I sketched the Princeton 5F2A Tone Control.  I thought the 5E3 and the 5F2A tone stacks were the same, but they are a little different

[scstill]

Hey NP, I'm still learning (always)

For me it begs the question what is the difference volume first vs tone first.
Based on what Gemini says (below) I might change to Volume first.
What do you think?
Plus I do not think any of my builds have Volume first

"In a guitar preamp circuit, placing the volume control first allows the tone controls to work with a more consistent and controlled signal, while putting the tone control first can lead to more dramatic or thinner tones, as it applies the filter before the signal is fully boosted, and the later volume control then adjusts the overall output. Essentially, tone-first shapes the signal's frequency content early, while volume-first allows for more subtle tone shaping that is less affected by distortion.

Volume Control First
When the volume control is placed first in the circuit, it acts as the primary input control for the preamp stage. This allows the subsequent tone controls to operate on a more predictable and stable signal level, which can lead to more controlled and predictable shaping of the tone. It can also result in less distortion, as the signal isn't being driven as hard into the first gain stage. Think of it like adjusting a water hose before the sprinkler. If the volume control is first, it's controlling the overall water pressure going into a filter, ensuring the filter receives a consistent stream to work with.

Tone Control First
Placing the tone control first means the signal's high frequencies are filtered out (shunted to ground) very early in the circuit, even before the primary signal level is set. This can result in a more dramatic change in tone, potentially leading to thinner or more muted sounds, as the tone-shaping is done on a very raw and uncompressed signal. This is like adjusting the water hose's spray pattern (the tone) before you've even turned on the main faucet (volume). The filter is applied to a very broad or direct stream, and then the overall output is controlled. "

[shooter]

what Gemini says

so most of the major amps designed/made last century got it wrong according to AI?? 


from a Double E perspective, AI might be correct, from 80 years of guitarist Rockstar's, they might disagree.


when I 1st can to tube amps from the engineering world, I was horrified at the waveforms guitarist said "Now That's how a Great Amp should sound"  If my boss seen that kind of sheet coming from one of my systems, I'd be selling beer bait n gas long ago.

[Platefire]

I almost hate to answer the question because when I do, I prove more and more what I don't know
 
This is the way I think of it. First you have the raw signal divided into two paths. One is the Raw volume path, the other the with the treble cap and tone control pot.Then the two paths are mixed back together prior to going to PI or in a SE amp, the power tube.

So this leaves the question on tone adjustment side of the circuit, rather to put the 500pf prior to the 1M tone control like the 5E3, or after the 1M tone like the 5F2A.

It seems to me the idea of the 5F2A arrangement for the tone control with the raw signal traveling through the 1M tone pot first controlling the amount of raw signal effected by the 500pF cap before being mixed back in with output of the volume pot would offer more control?

The 5E3 way with the 500pF cap first, a 100% of the raw signal is effected and the 1M tone pot just controls the amount of 100% effected signal is mixed back in with the raw volume signal sent to the PI or/power tube.

However skewed my thinking is electronically on this, this is the way I think of it Seems like AI and I tend to agree on this one

[scstill]

Back to hum Troubleshooting
1- Heater Voltage measures 6.6v; any concern being too high?
2- Bias on 7591 blue (-18.12v)  on 7591 brown (-17.73v); any concern with imbalance?
3- with preamp tube removed hum is still controlled by MV. shouldn't the the hum be in front of MV since it is adjusting hum? but with Preamp removed is there anything in front of MV to create hum?
4- with PI out hum is gone.

[scstill]

The PT in the original Cordovox design powered 60 tubes in the tone generator that are not in the circuit anymore. Could this possibly have something to do with the hum?

[tubeswell]

Sorry thought I described noise in the first entry...

There is both buzz (sharp like when you say buzzzzzzzz) and 60hz hum (mellow like when you hummmmm)
I think most of the buzz might be from LED lights as when I'm in the dark it is mostly gone. But the Hum is always present and increases with MV. Even though I'm plugged into AC noise filter. Have also tried outlets at the other end of house with same results.

The hum is noticeably reduced going directly into PreAmp 3, but that could be because two stage amplification is gone too.

BTW - the guitar tone is very nice. Just the hum is quite annoying.
bench setup below. All three chassis have ground continuity. I was careful to keep star grounds as shown in the ground schematic in first entry

That wiring is pretty chaotic looking, and the filter cap can looks really old. 2 places to start.

[scstill]

That wiring is pretty chaotic looking, and the filter cap can looks really old. 2 places to start.

Thanks
Yeah chaotic for sure, but that is how they did it in the day.
I am assuming that the original design and layout is ok because it worked at that time.
Have had success with this approach in the past. Chop sticking reveals nothing on wiring.
Thinking the hum is from something I did in the conversion.
The filter caps have been replaced with new.

[shooter]

with preamp tube removed hum is still controlled by MV.

with PI out hum is gone.

so V1 out, 60hz still there
V2 out it's gone.


look at the schematic, the MV is basically a VOL control for the V2 driver that feeds the PI. 


lift the "In" side of the 100k R
run a gator-clip from the "IN" side of 100k to ground. 60hz gone??

[scstill]

lift the "In" side of the 100k R
run a gator-clip from the "IN" side of 100k to ground. 60hz gone??

Yes hum gone.

In fact when I connect guitar at that point it plays well without any hum
of course without any tone, volume or MV control and missing 2 preamp stages.
So I guess the hum is before that point. Of coarse everything before that point is my conversion changes

Prior to your thought, I modified the power supply and grounds of PA3 and PI so they are from the G2 star. Thinking there might have been a loop. It did not change the hum. Mod reflected in the "edit3" schematic in original post.

[shooter]

So I guess the hum is before that point.

no guessing required 


once DC checks are made, 1/2-splitting the circuit follows, now 1/2-split what is between the in-jack and the 100K R

[scstill]

Thanks
I have not yet gone half split but I did discover thru this exercise, that the original coax signal cable running through the umbilical had its one sided ground on the end that I cut off to shorten it. When grounding the coax properly it noticeably reduced the hum :-)  Still louder than I would like, so will continue to progress adding sections and evaluating.

[scstill]

I replaced the original pots with new CTS to eliminate the scratchiness
Smooth operation but at the max setting there is a severe crackle both tone and vol.
Plus the hum is still loud with max MV (reduced from earlier sig cable fix but still too loud for my liking)

Tried a couple of other 12ax7 in preamp slot, same result.
But when I tried 12au7, it removed all issues. No Crackle in pots, min hum with MV.
The amp is still very loud. I'm thinking this design with three AX stages is too much gain.
So now I have 2AU 1AX and the PI

Any advice on other approach to taming the gain for AX's in Preamp1/2 appreciated.

[AlNewman]

Wouldn't lowering the value of the pots reduce gain?  Or running a resistor in parallel with the pots to ground?

The key would be to find out where the signal is not happy.  Marshall did it with their treble boost/cut dividers.

[scstill]

Wouldn't lowering the value of the pots reduce gain?  Or running a resistor in parallel with the pots to ground?
The key would be to find out where the signal is not happy.  Marshall did it with their treble boost/cut dividers.

These options are worth a try, thanks     
I often use 1M pots like 5E3 so using 500k seemed low.
will look at Marshall, is there a model you'd recommend?
                             

[AlNewman]

You were asking how to reduce gain.

Really, there shouldn't be much gain in that circuit.  You could try just a regular guitar input by removing the 2.2M grid leak, and jumping the .0047 cap.

From there, I would maybe try increasing the 100k divider to the PI, with possibly a cap in parallel, to reduce gain and maintain high frequencies.

What was the amp designed for?  The feedback loop is also pretty wild.

[scstill]

Really, there shouldn't be much gain in that circuit.
What was the amp designed for?  The feedback loop is also pretty wild.

There are three 12AX7 stages before the PI, why do you say this shouldn't be much gain?

Originally Cordovox was designed for accordion and mic inputs. My plan was to convert to guitar amp with lots of headroom (7591) into a 4x12 speaker cab keeping as much of the Cordovox design as possible.

[shooter]

So now I have 2AU 1AX and the PIAny advice on other approach to taming the gain for AX's in Preamp1/2 appreciated.

signal path "tuning" really really needs to be done with sig-gen n scope, guessing just leads down paths to nowhere.


here's the data sheet for AU7, with cheats for proper biasing.  I try n set V1a for a "centered swing"  boosting the S/N, no real "tonal" additives, V1b I setup somewhat asymmetrical for some preamp flavor.

I would recommend 1-10uF bypass cap on V1a, then experiment on V1b and V2a

[AlNewman]

Really, there shouldn't be much gain in that circuit.
What was the amp designed for?  The feedback loop is also pretty wild.

There are three 12AX7 stages before the PI, why do you say this shouldn't be much gain?

Originally Cordovox was designed for accordion and mic inputs. My plan was to convert to guitar amp with lots of headroom (7591) into a 4x12 speaker cab keeping as much of the Cordovox design as possible.

It looks very similar to a Princeton.  I think Shooter is right, and I mentioned earlier, you need to find where the signal gets snakey.

[scstill]

So I put the 12ax7 back into preamp slot to try AlNewman and Shooter ideas.
Nothing really worked and in many cases made the pot crackle and hum worse.
What I did notice during these tuning experiments, is  that when I clipped a wire to the Signal 1 at the .0047 cap input unconnected at the other end, the pot crackle was gone.  a little more buzzy though
Maybe this is acting as an antenna of sorts. Baffling to me....

I believe the V1a (with 12ax7) is close to center swing: Vplate 60, 160, 260v corresponding to Vgrid 0, 1.4, 3v

Thinking that maybe I should rebuild the preamp to more of a guitar circuit vs original mic circuit. The original preamp chassis is so tiny.

[scstill]

I had reassembled the amp into its cabs to see how it sounded complete vs on bench. No improvement, so I put it back on the bench to keep trouble shooting...

Tried a couple of other ideas:

1) High filament Voltage
received some insight from another thread that maybe the heater voltage was too high because the Cordovox Tone Generator with ~60 tubes was no longer part of the circuit.
The heater voltage measures 6.8v, when I added a dropping resistor to get it closer to 6.3v nothing really happened to the hum.

2) Power Tube Fixed Bias with only reservoir cap
The original Cordovox bias circuit only had one filter cap before the dropping resistor and no filter cap at the node.  Another conversion design (noisemakers) added a filter cap at the Vbias node. When this was tried it did nothing for the hum. The attached current schem show the original Bias and the revised bias (although I only jumpered in the filter cap to test did not add the adjusting pot)

The sound bite link below is the hum with the MV turned up and the Tone adjusted in the middle of the bite and the Vol adjusted toward the end of the bite. Seems to be a 60hz hum (LMK if you think otherwise)> All pots have a click when they are at full (not sure why but the new CTS pots do have a resistance skip at max as well).
https://drive.google.com/file/d/1gSXfBiblGQhzKY9qJP8gj2QojfC7alxJ/view?usp=sharing

[AlNewman]

That sounds like bad connection, like an open ground, open switch to ground, etc.   Sounds like you have your thumb on the tip of the cable.

[AlNewman]

I remember I rebuilt a radio a few years ago, and had a similar buzz.  I can't remember if I maybe shorted a cathode direct to ground, or maybe left one of the tubes disconnected to ground.  It definitely let me know in a hurry.  I was only rebuilding in sections, so I was able to find it fairly quickly.

[Platefire]

You may have already done this, but have you checked for any DCV on the pots?

[Wonder2121]

What purpose does the 100k/22uf circuit from heater CT to ground do?

[dogburn]

That sounds like bad connection, like an open ground, open switch to ground, etc.   Sounds like you have your thumb on the tip of the cable.

I had a recent project where I got a similar buzz, and it turned out to be a loose wire on the input jack. Have you tried re-wiring the inputs or replacing the jacks altogether?

[scstill]

What purpose does the 100k/22uf circuit from heater CT to ground do?

This was original to the Cordovox circuit.
I believe the 100k is to elevate the heaters to reduce hum.
The 22uf is to filter any remaining ripple.
Maybe a dated approach...
Does anyone think this could be an issue? Rarely see it in guitar amps right?

below is a more detail explanation from Gemini
"In a guitar amp, a 100kΩ resistor and a 22μF bypass capacitor on the 6.3V heater center tap serve to "elevate" the heater voltage, which is a key technique for reducing noise and hum. This is most beneficial for sensitive high-gain preamp tubes, as it helps create a cleaner, less noisy signal.

The purpose of heater elevation
While a simple grounded center tap helps cancel some AC hum, it does not prevent capacitive coupling between the tube's heater and cathode, which can introduce a 60Hz hum into the audio signal. By elevating the heater to a positive DC potential (usually sourced from the high-voltage B+ supply), this coupling is significantly reduced.

In a guitar amp, this technique is particularly important in the following ways:
Silence high-gain stages: The first gain stages of a guitar amp are the most sensitive to noise, and any heater hum introduced here is amplified by every subsequent stage. Elevating the heater voltage is one of the most effective ways to eliminate this type of hum.

Manage cathode-biased tubes: In cathode-biased stages, the cathode is already at a positive DC voltage. Elevating the heater voltage ensures that the heater-to-cathode voltage stays within the tube's maximum rated limits, preventing damage and further reducing noise.

Reduce parasitic noise: The high resistance of the 100kΩ resistor, combined with the bypass capacitor, helps isolate the heater circuit from the potentially noisy main signal ground. This prevents ground loops and ensures a quieter operation.

The role of the 22μF bypass capacitor...
The capacitor is placed in parallel with the 100kΩ resistor to form a low-pass filter, which is critical for making the elevation effective.

Filters out residual noise: The 22μF capacitor filters any remaining AC ripple from the high-voltage B+ power supply. This ensures that the elevated potential applied to the heaters is a pure DC voltage.

Creates a stable reference: By providing a low-impedance path to ground for AC frequencies, the capacitor ensures that the DC elevation voltage remains stable, even when there are small fluctuations on the B+ rail. This prevents noise from the power supply from modulating the heaters and causing additional noise in the preamp."

[scstill]

I had a recent project where I got a similar buzz, and it turned out to be a loose wire on the input jack. Have you tried re-wiring the inputs or replacing the jacks altogether?

Early on I bypassed the jack it made no difference.
BTW - these original cordovox jacks have double switches, the first to ground when no plug, the second is not connected. Odd since this was the time when manufacturers were trying to save bucks. Maybe they had these lying around.

[scstill]

You may have already done this, but have you checked for any DCV on the pots?

Yes I measured for DC some time ago.
Doubled check again just now same, no DC.
BUT, my ground side meter lead was sitting close to the preamp tube.
I discovered that when a ground wire is close to the preamp glass the hum/buzz was noticeably reduced.
Thoughts?

Have to further troubleshoot this. But possibly a good pathway...

[Wonder2121]

I dont see how its elevating anything, it's not connected to a DC source...
Read this https://www.valvewizard.co.uk/heater.html

[shooter]

not a theory guy so;


the "charged" cap n R form an artificial ground, above actual chassis ground.  easy enough to verify, set up your meter for VDC, measure the top of cap/R juntion to chassis ground.


me, I would lift that connection from CT n temp ground CT to chassis, just to eliminate it as a problem, when all the good things are verified, then only the bad thing is left.   

[scstill]

Yeah that circuit is not elevating the CT so something not working correctly, unless it is not why its there.
BTW the cap is 0.22uf wrong on the schematic.
But, there is no change in the hum when jumping xfmr CT to ground. So that is not the issue apparently.

There is another 7vct coil on the transformer (not used in this guitar conversion) that was used for the Cordovox Tone Generator Tubes and its CT was also connected thru 100k to the CT of the 6.3v winding (diagram attached). But when I reconnected that CT it changed nothing.

The biggest (really big) reduction in hum is when a ground wire is close to the preamp glass. Also eliminates the loud pot click at full on. Maybe a tube shield is in order. Or a redo of the preamp circuit. Most everything in this build is as it was in the Cordovox except new filter caps and revised signal path adding tone stack and MV in Preamp and some improved star grounding.

[AlNewman]

I'm sure you've already tried the simplest option, which is to remove all the tubes, and install them from the power section back to isolate where the problem originates.