12AX7 HUM

[Chuck H]

Only a few builders have tried to do the logical thing of creating an all common mode path, balanced line, pickups, H-attenuation and gain stages but it never caught on despite all its advantages.

Ha ha. That's because it almost cant be done without the need for at least a little user awareness. Some guitar players have no trouble with this. But most, well, most get confused by hi/lo inputs or an adjustable effects loop. I have customers that are stupified by the noise of single coil pickups or try to use a speaker cable for an instrument cable, or a battery dies in their guitar or effect pedal and think their amp is broken. I even had one customer who thought something was wrong with his amp because it didn't have the same gain level anymore. He had flipped the gain switch on the back of the amp in transport and had forgotten it was there. I'm just sayin', put more than four knobs on an amp and the average player needs a lot of time to dial in and understand what to do just to use an amp.

[Jazz P Bass]

Ha Ha!.
I forget exactly which model it is, Mesa Boogie has a speaker mute switch tucked under the rear of the chassis.
Pick up the amp the wrong way & "click".
(It worked last night & now nothing!)

[Chuck H]

Bwaaaahaahahahaha.

That happened to ME. After extensive mods to a Subway model. It sounded great and I meant to loan it to a friend. I brought it to his house and plugged it in to show it off and "What the, ???, I don't understand. It worked just an hour ago."... "I" had forgotten about the switch and had flipped it during transport. My friend, who knows I build and mod amps just looked over his nose at me. "Well, I guess you'll have to go in there and find out what you broke." Ha.

[Jazz P Bass]

Something about "never show it off" or "never brag about a repair" or "never ask how it is working"
Karma or somesuch thing.

[EFK]

I have to agree with the ceramic cap comments - I certainly don't know nearly the theoretical aspect of it that soundguruman does, but I know for a fact that anytime I've had a ceramic in a signal path I can get all kinds of weird noises and especially more hum that disappears when those ceramics are replaced. Not just old ones either. The only place I'll use them now is in locations where they are going to ground - doesn't seem to have the same effect.

I love the Tung Sol 12AX7 tubes, and I think they sound fantastic, but I have had a lot of problems the past year finding quiet ones to use in V1: problems with humming and problems with microphonics. Really, I think the quietest new 12AX7 tubes on a consistent basis are the JJ, but they don;t seem to have quite the "life" in them that some of the other new tubes have (I don;t mean longevity, I mean in reference to the sound).

[big_teee]

Question:
Are the Shorter Plate 12AX7s Quieter?
Is that's to be my Take from this Thread.
So far I'm still about where I started.
I have spent low and high amounts on different ones, and so Far seems to be a Crap Shoot.
I do like the texture and tone of the Tung Sol's.
B_T

[Chuck H]

I've had more hum problems with short plate tubes. Long plate tubes often seem to be brighter and/or higher gain so perhaps more hiss? I haven't noticed.

[soundguruman]

Have you measured this effect? I find it hard to believe that considering differential mode active stages have so many "difference" noise sources that such minuscule contribution by ceramic caps several diameters away from the twisted lead heater wires. The only effective difference field is the imbalance of current between the two twisted wires. That is why they are twisted, to cancel such field induction. There are too many real difference mode sources of higher amplitude in the circuit to be that concerned with tiny common mode imbalances.
There are other reasons that can be measured as to why ceramic caps are not advised for audio signal paths such as their piezoelectric properties producing varying instantaneous values of capacitive reactance. The non-linearity that results increases inter-modulation in complex signals. The same piezo property of ceramic caps results in micro-phonics.
I would be interested in seeing and studies or measurements that was the basis for the conclusion that ceramics are any more susceptible to fields than say wires or other passive components.

By any criteria, differential signal paths in such high gain circuits as guitar amps, which usually strive for very high input impedance, it is a miracle that amps are so quiet. Only a few builders have tried to do the logical thing of creating an all common mode path, balanced line, pickups, H-attenuation and gain stages but it never caught on despite all its advantages.

The problem being that tiny amounts of inducted 60 cycle enter the signal path, especially in the first gain stage.
This is amplified thousands of times, and ends up as 60 cycle hum in the final product.
By eliminating or reducing that 60 cycle, even by a small amount, the amp produces a much more pleasing result.

Formally, amp makers paid no attention to this. Amps hummed, all of them.
After all, it takes the highest setting of the O scope to even see a hint of this 60 cycle in the first gain stage.
But, as we added more and more gain stages, the hum became louder and louder. We could no longer ignore the problem.

We started out with DC heaters, like in the early hi fi designs. Look at the McIntosh tube preamp schematics.
Then somebody, like PV, took the hint, they started using 1/2 wave rectification on the preamp tubes.
This then progressed into full wave DC to heat the preamp tubes, look at the earlier Carvin amps.
Now, about half the manufacturers out there are using DC heaters. Some have even started using non inductive components.
It's about time that the manufacturers started paying attention to the signal to noise ratio.

And others eliminated the low frequency response of the amplifier all together, just so the hum would not be passed to the output.

But still there is no high gain low noise amplifier for guitar-you are going to see that arrive, somebody is going to take the hint.

And then if you have a classic amp, like a black face, silver face, there is a hum in it, that Fender ignored. The gain was not high enough to be bothered by the problem.
But the hum can be reduced by installing silver mica caps, and the difference is measurable, if you care to measure it.

You find a tech that plays guitar, instead of just working on amplifiers, you are headed in the right direction.
The ones who do not play the guitar are oblivious to the problem. They see the amp as "working" or "non-working" they do not see the subtle differences in noise and performance.
It takes a musician to hear the difference....although it is measurable, these measurements are typically ignored. To the uninitiated, it make no difference.

It's true that only the most critical listeners are bothered by this noise. The average Joe does not even notice it.
But if it was in your Hi Fi amp, you would start to notice it.

It's easy to theorize about common mode rejection, and why there "shouldn't" be noise in the stages. But it does not work out that way in a high gain amplifier. ANY 60 cycle source in the vicinity of the preamp is another potential source of noise.
Just as the pickup in a stratocaster will induct hum from any 60 cycle source (including the amp power transformer), and introduce it into the signal path...(like electrical wires in the house wall, fluorescent lights)
the parts in the amplifier will produce the very same results.

[soundguruman]

Have you measured this effect? I find it hard to believe that considering differential mode active stages have so many "difference" noise sources that such minuscule contribution by ceramic caps several diameters away from the twisted lead heater wires. The only effective difference field is the imbalance of current between the two twisted wires. That is why they are twisted, to cancel such field induction. There are too many real difference mode sources of higher amplitude in the circuit to be that concerned with tiny common mode imbalances.
There are other reasons that can be measured as to why ceramic caps are not advised for audio signal paths such as their piezoelectric properties producing varying instantaneous values of capacitive reactance. The non-linearity that results increases inter-modulation in complex signals. The same piezo property of ceramic caps results in micro-phonics.
I would be interested in seeing and studies or measurements that was the basis for the conclusion that ceramics are any more susceptible to fields than say wires or other passive components.

By any criteria, differential signal paths in such high gain circuits as guitar amps, which usually strive for very high input impedance, it is a miracle that amps are so quiet. Only a few builders have tried to do the logical thing of creating an all common mode path, balanced line, pickups, H-attenuation and gain stages but it never caught on despite all its advantages.

The criteria is not that of musicians, this is the criteria of manufacturers.
Sure, the common mode reduced the noise considerably, but not enough for a musician.

The problem being that tiny amounts of inducted 60 cycle enter the signal path, especially in the first gain stage.
This is amplified thousands of times, and ends up as 60 cycle hum in the final product.
By eliminating or reducing that 60 cycle, even by a small amount, the amp produces a much more pleasing result.

Formally, amp makers paid no attention to this. Amps hummed, all of them.
After all, it takes the highest setting of the O scope to even see a hint of this 60 cycle in the first gain stage.
But, as we added more and more gain stages, the hum became louder and louder. We could no longer ignore the problem.

We started out with DC heaters, like in the early hi fi designs. Look at the McIntosh tube preamp schematics.
Then somebody, like PV, took the hint, they started using 1/2 wave rectification on the preamp tubes.
This then progressed into full wave DC to heat the preamp tubes, look at the earlier Carvin amps.
Now, about half the manufacturers out there are using DC heaters. Some have even started using non inductive components.
It's about time that the manufacturers started paying attention to the signal to noise ratio.

And others eliminated the low frequency response of the amplifier all together, just so the hum would not be passed to the output.

But still there is no high gain low noise amplifier for guitar-you are going to see that arrive, somebody is going to take the hint.

And then if you have a classic amp, like a black face, silver face, there is a hum in it, that Fender ignored. The gain was not high enough to be bothered by the problem.
But the hum can be reduced by installing silver mica caps, and the difference is measurable, if you care to measure it.

You find a tech that plays guitar, instead of just working on amplifiers, you are headed in the right direction.
The ones who do not play the guitar are oblivious to the problem. They see the amp as "working" or "non-working" they do not see the subtle differences in noise and performance.
It takes a musician to hear the difference....although it is measurable, these measurements are typically ignored.

It's true that only the most critical listeners are bothered by this noise. The average Joe does not even notice it.
But if it was in your Hi Fi amp, you would start to notice it.

NOW, remove the shielding from the input stage of your O scope. Does the 60 cycle hum bother you NOW?
You are not listening to it, you are LOOKING at it!
Just as all the interference screws up your O scope trace, the same problem is interfering with our music.
You would not want an oscilloscope with 60 cycle bleeding into the waveform. I do not want 60 cycle bleeding into my music.
Same thing folks, different application.

NOW, remove the metal film resistors, and silver mica capacitors from the input of your oscilloscope.
Install carbon film resistors and ceramic disk capacitors.
Now turn your scope on, and tell me what the waveform looks like...
Garbage?

You want the lowest noise components in your oscilloscope. You want the cleanest waveform. Your O scope IS an amplifier.
I want the lowest noise components in MY amplifier too. I want clean, garbage free music.
Both are amplifiers, both are artists, attempting to achieve a similar result!

If we put the concept into terms that a technician can understand, prehaps we have achieved communication.

[km6xz]

My interest in low noise is from the music production end, as a recording engineer for decades, design engineer and also having a large full service repair facility handling the work for many top selling artists. You never heard hum an any of my recordings, I would venture a guess, and I dare say we were more sensitive in a major league studio than a gigging guitarist would be.
Noise is one of my interests and I am interested in data and reproducible experiments. Noise can be measured and is everyday in many fields where it is important. Dial in a wave analyzer and have lots of measurement headroom, which of course would be compromised by broadband noise on a scope.
You made a statement that seemed counter intuitive and I asked for your measurement data. Do you have any, and under what conditions? Or is this claim anecdotal?
I raise this because after years of debunking claims by "golden ears" in hi end hi-fi and music production, I hear claims that do not have supporting evidence, my curiosity is aroused. As I said, there are lots of reasons to avoid disc ceramic capacitors in signal chains for reasoned which have been measured but hum sensitivity is not one I have heard or measured.

Your suggestions on how to show it "screws up the waveform" introduces a dozen or more uncontrolled variables so is no reason to attempt that when the hum pickup is actually quite well within range of the proper test instruments.

"If we put the concept into terms that a technician can understand, prehaps we have achieved communication" Humm, I am a tech and a damn good one, also a degree'd engineer with design and patent-approved history....so why am I not seeing any thing convincing in your claim? Maybe I don't speak the right "technician" language.

[soundguruman]

"As I said, there are lots of reasons to avoid disc ceramic capacitors in signal chains for reasoned which have been measured but hum sensitivity is not one I have heard or measured."

You just learned another reason not to use them, I guess.

I think I can explain it another way, the grid of the first gain stage is extremely sensitive.
Just putting your hand in the vicinity of the grid or grid wire will produce a hum. This is why we need to use well shielded guitar cords. But anyway...
So now I hook up a sensitive part (disk cap) to the grid. That part becomes essentially an antenna.
Put your hand near that cap and you will get plenty loud hum.
Now try the same with a silver mica cap. You will notice the immediate difference in sensitivity. It hardly cares at all what is in the vicinity, no hum, or greatly reduced sensitivity to external 60 cycle fields.
It's the construction of the thing. Must be voodoo or something.
I don't build them (the sm caps) , but I can tell the difference in performance.

[km6xz]

Learn something from that? You must have a very low threshold of analyzing evidence.
So you have no controlled measurements to show that your claim that these caps are sensitive to electro-magnetic fields. By your logic, all wires need to be eliminated because they are much more sensitive than any capacitor for fields surrounding because the cap has a high reactance and has self sheilding by design. You suggest greatly increasing the field strength by introducing a large antenna to the immediate area.
Mounting position and lead length are factors that make a larger difference in field sensitivity, just like every passive part, so tiny differences in mounting position, height over the chassis, proximity to field generators, such as poorly twisted heater leads running too close.

These caps do contribute to inter-modulation distortion however due to piezioelectric effects with ceramic. But that is to be avoiding in reproduction, but is not nessilarily a bad thing in production of music where distortion is needed to create a complex tone from a dominate fundamental. We use tubes due to their distortion character, not because they are hi-fi. They can be hi-fi to the input signal but not run in the wide range of drive and output swing used in guitar amps. A very clean amp is almost unusable for music production and you know that. That is why the caps are not used by designers of audio gear, but are used by MI designers. They do sound different and that might be just what the designer wanted as part of their tone signature. Hum and noise pickup by any design or measurement criteria is insignificant and swamped by noise induction mechanism that ARE present. Unbalanced chassis current flow, unbalanced heater wiring, parts placements, transformer design and orientation and many more factors are all orders of magnitude higher amplitude hum sources. Mentioning the guitar cable was a redherring, as you know a cable has higher delta capacitance than any fixed cap, due to cable movement, they are micro-phonic machines, translating mechanical movement of the proximity of the core to shield into a modulated shunt capacitance. The delta C will be in normal use, greater than any variability of fixed caps.
All caps have some short comings but some of those are useful in creating a unique sound. Electrolytics are often given a bad rap as path components but their self shielding character, and unique signal influence on asymetrical waveforms can be just what a designer with a goal of a particular tone might seek.

A designer should understand the fundamentals plus dynamic variables of components at their disposal when designing for a specific intended goal. Without understanding the real nature of the parts available and their influence on the outcome, creating an amp that sounds like intended would be relying on magic and legends but mostly luck if the results are near the goal. If you need the influences on signal that ceramic caps impart,, and which design of the caps have what characteristics, by all means use them because of them fitting goal, but to discount them due to internet myth and unproven claims is shrinking the toolbox a designer has to use in accomplishing the intended goal. It is not voodoo except for people without the facts and theory. For them, anything can be magic...airplanes and TV remote controls are surely mysterious things for some that only magic or gods could be the cause. High end hi-fi and hobbyist guitar amp discussions usually center around unproven myth and magic. Isolate, measure, understand.

[Alex R]

The '12AX7B' Chinese tubes, sold as specially selected for balance and low noise, are useful to have around, I find, for these problem amps that are just a bit too noisy. They have definitely reduced hum and hiss in a number of amps for me when in the first/second stage. I keep a stock of the standard Chinese 12AX7, and a handful of the 12AX7B for this particular issue.

I too am a bit mystified by the idea that ceramic caps are specially bad for picking up atmospheric noise. Film caps and resistors do it too, as does everything in the signal path that gets amplified a lot. In quite a few years of passing my hands over the insides of amps to look for induced noise I haven't noticed ceramics being any worse for that than any other component. I have found and replaced quite a few film caps that have begun to pick up enough induced noise to cause oscillation - found one in a Roland Bolt only last week - but not a ceramic yet. I wonder whether the film wraps start to loosen up and make microscopic physical responses to induced AC, as these failing caps are also noisy when tapped.

[Jazz P Bass]

I have never understood the failure mechanism in old film caps.
The ones that are noisy when tapped "leak" Vdc when tapped.
Maybe the film does break down.

[soundguruman]

Learn something from that? You must have a very low threshold of analyzing evidence.
So you have no controlled measurements to show that your claim that these caps are sensitive to electro-magnetic fields. By your logic, all wires need to be eliminated because they are much more sensitive than any capacitor for fields surrounding because the cap has a high reactance and has self sheilding by design. You suggest greatly increasing the field strength by introducing a large antenna to the immediate area.
Mounting position and lead length are factors that make a larger difference in field sensitivity, just like every passive part, so tiny differences in mounting position, height over the chassis, proximity to field generators, such as poorly twisted heater leads running too close.

These caps do contribute to inter-modulation distortion however due to piezioelectric effects with ceramic. But that is to be avoiding in reproduction, but is not nessilarily a bad thing in production of music where distortion is needed to create a complex tone from a dominate fundamental. We use tubes due to their distortion character, not because they are hi-fi. They can be hi-fi to the input signal but not run in the wide range of drive and output swing used in guitar amps. A very clean amp is almost unusable for music production and you know that. That is why the caps are not used by designers of audio gear, but are used by MI designers. They do sound different and that might be just what the designer wanted as part of their tone signature. Hum and noise pickup by any design or measurement criteria is insignificant and swamped by noise induction mechanism that ARE present. Unbalanced chassis current flow, unbalanced heater wiring, parts placements, transformer design and orientation and many more factors are all orders of magnitude higher amplitude hum sources. Mentioning the guitar cable was a redherring, as you know a cable has higher delta capacitance than any fixed cap, due to cable movement, they are micro-phonic machines, translating mechanical movement of the proximity of the core to shield into a modulated shunt capacitance. The delta C will be in normal use, greater than any variability of fixed caps.
All caps have some short comings but some of those are useful in creating a unique sound. Electrolytics are often given a bad rap as path components but their self shielding character, and unique signal influence on asymetrical waveforms can be just what a designer with a goal of a particular tone might seek.

A designer should understand the fundamentals plus dynamic variables of components at their disposal when designing for a specific intended goal. Without understanding the real nature of the parts available and their influence on the outcome, creating an amp that sounds like intended would be relying on magic and legends but mostly luck if the results are near the goal. If you need the influences on signal that ceramic caps impart,, and which design of the caps have what characteristics, by all means use them because of them fitting goal, but to discount them due to internet myth and unproven claims is shrinking the toolbox a designer has to use in accomplishing the intended goal. It is not voodoo except for people without the facts and theory. For them, anything can be magic...airplanes and TV remote controls are surely mysterious things for some that only magic or gods could be the cause. High end hi-fi and hobbyist guitar amp discussions usually center around unproven myth and magic. Isolate, measure, understand.

That's OK, you can build amps that hum, if you want to.
But you are still looking at it like a technician, not as a guitar player...
Guitar amp has thousands of times more gain than any other conventional audio equipment.
You are still looking at the guitar amp as audio equipment. When you consider the gain, it's no longer conventional.
There is enough gain to drive the amp into non-linear operation, where the stages are no longer inverting or non inverting. The gain is what you are missing...
This amount of gain will make the smallest little noise-- very loud at the output. 1/2 mv of hum in the first stage is quite loud in the speakers, although it's almost too small to measure with your test equipment.
In a Marshall, 1/2 mv audio input is supposed to drive the amp to clipping, or so say the engineers at the marshall factory. When you have an amp that is this sensitive, you start nit-picking the noise sources. That's what you are missing.