I think there's probably more confusion there than I can sort out. If I were you I wouldn't start out by getting too into the internal details of your mixer, I would learn gain staging in general and as it applies to your mixer. Its manual should help with that. Simply learning and using best practices should minimize noise.

What mixer, by the way?

I know about gain staging. I am asking these questions to better understand the technicality of my signal flow.

Most likely there's a DC blocking cap somewhere near the input followed by a preamp stage that uses a variety of caps for different things. The preamp stage produces some noise, which you can minimize by proper gain habits. The insert will follow the preamp, usually after the HPF (which likely has more caps). The eq section (with yet more caps) might be before or after the insert, I've seen it both ways. Then there's the routing, with resistors to isolate the channel from other channels feeding buses.

Most affordable mixers don't have any transformers at all except in the power supply. Usually they use transistors rather than transformers to reject noise. That's why I asked what mixer you have.

There are some short answers and some long answers to your set of questions. I'll stick to the short(er) ones.

Firstly, a bit of physics. All resistors in a circuit generate noise (Johnson noise), the energy of which is a function of the Ohmic value, the measurement bandwidth, the temperature and a physical quantity called the Boltzmann constant. The voltage generated is proportional to the square root of the power (energy). Very roughly, a 1000 Ohm (1K) resistor generates a voltage noise a bit less than 1 microvolt r.m.s. at room temperature over the range of audio frequencies. If you had two identical circuits but one built with 1K ohm and the other 100K ohm resistors, the 100K version would have 10 times the voltage noise of the other.

In addition to Johnson noise, all active devices (transistors, valves, integrated circuits, etc) add their own noise. It's the job of the circuit designer to balance noise performance against other factors such as distortion, power consumption, audio quality (a very loose term) and manufacturing cost.

Now the mixers. All mixer designs should be different, so will have different performance properties. Some years ago there were mixers produced by a couple of companies (no names) that were pretty much exact copies of the circuit designs from other manufacturers, but built using much cheaper components and manufacturing processes. Surprise, surprise, these copies were noisier, distorted more, sounded worse and were much less reliable than the originals.

Your question about the capacitors rather misses the mark. Most mixers are able to supply 48V phantom power to microphones connected to their XLR inputs,. This power is supplied to the microphone as a common-mode voltage using ground and the same two conductors that return the differential signal voltage from the microphone to the mixer. The two main methods of isolating the 48V phantom voltage from the sensitive amplifier circuits at the front end of the mixer are the use of transformers and capacitors, both of which block d.c. voltages. This is not the only way of doing it, as one of the mixers I designed (primarily for church organ recording) had d.c. coupled inputs in a special configuration that was insensitive to common-mode input voltages up to 60V. Note that capacitors can have a very low equivalent series resistance (ESR), and so generate very little Johnson noise of their own. Equally, transformers usually have low winding resistances, and so do not generate large Johnson noise levels, but they multiply up the effective source resistance (by the square of the turns ratio) and so pass on the noise along with signal in the input circuit to the output.

As a final comment about noise, relating to your question about what creates noise in the mixer, is that everything does. However, as the signal is amplified by the pre-amp stages, the noise at the input stage gets amplified as well, so tends to swamp the noise later in the signal path. The exception to this is the mixer stage, as the noise generated by that stage is amplified by a factor proportional to the number of mix inputs. That means that designers of mixers with a large number of inputs have to pay very close attention to the noise generated by the main mix circuits, and frequently use special circuit techniques to reduce the effect of self-generated noise.

After the pre-amp in the mixer, the signal passes through the insert switches (usually push-button or jack contact operated), and then through the EQ stages and finally on to the main mix stage. After the mix stage, some designs of mixer will have transformer-coupled outputs and others will have directly-driven outputs. I don't know of any professional mixers that use capacitor-coupling on their final outputs, as that would give a frequency-dependent output impedance.

Your question about what generates (non-deliberate) harmonic distortion in a mixer is less easy to give a straighforward answer to. All active components generate some distortion, and better active components (usually more expensive) are designed to keep the distortion to an acceptably low level. In addition to this, resistors can change their value slightly with applied voltage, and this causes non-linear distortion when the resistor is used, for example, in the negative feedback circuit of an amplifier. I have mentioned before in these forums that one of the problems I have been plagued with in my designs is manufacturing engineers substituting small surface-mount components where I had explicitly specified physically larger axial-lead components to minimise the non-linear distortion. There was even one case where I designed in three resistors in series to reduce the terminal voltage on each resistor and hence the distortion. When the design came through production, these had been lumped into a single component, with a corresponding hit to the distortion figues. This goes to show that people may know Ohm's law, but non-linear effects in circuits are generally poorly understood.

It almost appears that the OP is hung up on "capacitors generate noise"? Capacitors are often used to MINIMIZE noise. Of course, weak or failing capacitors will allow noise to creep in, as they are not doing their job, as designed.

ALL electronic components in a circuit have the ability to generate, or pass, noise. A lot of that is "thermal noise" within any particular component. Anytime you try to amplify a signal through any stage, you always have the possibility of generating noise.

Capacitors are used for several different things. The first thing is in the power supply, coming from the AC wall outlet. that outlet voltage hits an input transformer, and it generally goes through a rectifier circuit. That turns the AC into DC to feed the low-voltage DC power requirements of IC opamps, possibly effects chips, etc. That voltage has been regulated down to the required voltages for each circuit, possibly +/- 12V...+/-5V...+/-3.whateverV, etc.

Those are always 'elecrtolytic polarized caps', with a definite orientation as to + or - voltages, referenced to what the circuit dictates. (To make sense of that statement, if you are running a -12V regulator, and a cap is connected...UNINTUITIVELY, the + of the cap will actually go to ground, and the - to -12V, since, obviously, -12V is LESS than 0). If any cap(s) in the power supply section are funky (or the circuit is poorly-designed), noise could generate from that. The filter caps in the power supply are meant to 'smooth out" any remaining ripple that may remain after rectification.

So, in these cases, especially, capacitors are actually preventing (or removing) noise. If any of those are bad or weak...noise can get to the rest, as things aren't performing as designed (assuming it was designed well, in the first place).

The input coupling caps that may be in series with the input (some may be direct-coupled) are NOT electrolytic polarized caps, and so can be inserted without regard to orientation. Their job is to block any DC voltage that an input device may be passing to the input, and allowing only AC to pass (music, vocals, etc., which is always an varied AC signal). IF those aren't there, (or funky, or poorly-designed) any DC that may be present from the input device would ride along through the entire chain of circuitry, being amplified along with the musical signal. In this case, also...it's NOT "the cap" causing noise. It may be the cap not doing its job to PREVENT the noise.

Caps are also used in EQ sections, whether they're in graphic EQ, tone knobs, etc. A proper selection of a cap-and/or-resistor circuit acts as a 'filter', but not a 'noise-filter', as in the power supply. It's an 'audio filter', and determines the frequencies to block, attenuate, or boost. Those are non-polarized caps, also, allowing AC audio. If any of those are funky, it's likely more of "that (knob or slider) for that frequency doesn't sound right", as opposed to actually creating noise, though it could, especially if it's not doing its job to block high frequencies (allowing hiss), or low frequencies (rumble, etc.)

SOOO...the point is that capacitors are not the primary "generators" of any noise. They are actually remedies for noise generated more by other components, whether it's thermal noise inherent in the other components, noise entering at any input (including the AC power input), or in coupling sections of the device, that need AC coupling/DC decoupling.

ANY mixer will have SOME noise. The better designed ones will have less, and that noise is generally "masked" by the signal anyway. Unless, of course, there is a delicate passage of a lone, softly-played instrument. Even at that, depending on the range of timbres of the instrument, most of that noise can be reduced by wise and careful EQ, to only pass what it's capable of, thus rejecting unnecessary frequencies.

IF you KNOW you are gain-staging properly, but there is ALWAYS 'too much' noise (hiss, rumble, buzz, etc.), AND you know for certain that you are not feeding it into the mixer from another device...(if you have noise riding in the signal, itself, the input cap isn't going to get rid of that...it just sees that as part of the audio)...then you may have old, weak, leaky, cheap caps...or a poorly-designed circuit, to begin with. There are ways to try to isolate what section is causing it, and that takes some experimentation, and learning just what kind of noise you hear. "60-cycle hum". "Hiss". Etc.

That's all I got. It just seemed the OP was kind of...'infatuated'...with capacitors, in his post? Well-designed circuits, with good capacitors, help tame noise. Poorly-designed circuits can't be helped. Well-designed circuits, with faulty components (including caps) can allow the creation...or passing...of noise.

Kapt.Krunch

Looks like Boswell and I posted about the same time! Except...he explained more technically, which is never a bad thing!

I was just trying to explain in 'layman terms' what it SEEMED that he may have been driving towards...and it SEEMED that he was kind of obsessed (not saying that in a bad way!) with "capacitors generating noise"?

Kapt.Krunch

I hope the OP appreciates the incredible amount of knowledge and wisdom that you gentlemen have taken the time to share. Very well done.

I'd advise that you stop thinking about a 'signal' entering the mixer and then 'hitting' anything. If you look at the circuit diagram fro the product you have, the first thing you will notice is that it is not a case that the signal passes through each component sequentially, but the components are grouped into little small clusters of parts that do something. So you actually have passive components offering resistance, capacitance and inductance, provided by the individual components, and the configuration makes them what they are. Maybe a pad to reduce signal level, maybe a power supply to power the microphone. It could be a filter - designed for a specific frequency, or band of frequencies that maybe adjustable. Other chunks of components will provide amplification - the chips you see add gain, controlled by external components that set how much, and perhaps at what frequency. Other clusters may include more gain too enable amounts of the signal to be hived off and sent somewhere else. Some mixers might have little modules that can act as limiters, which again use the components to control what they do. More modules might simply be there to measure how strong the signal is.

So don't think of the audio passing through each component, it passes through the mixer and the capacitors, resistors, inductors and other active components just massage the signal as it goes though. The others explained how the noise components work, but in general, the little clusters of parts can be very well designed and complex and very noise friendly. Other designs may be simpler, use fewer and less expensive components and be less good - but they may be used in a different manner so their performance is still usable.

My own rather simplistic view is that the difference between a cheap capacitor an an expensive one with the same value is minimal, if measurable at all! A cheap mass produced resistor is still a resistor, and does the same job in the circuit as the expensive one. Maybe the cheap resistors are less strictly accurate. is that 470Ohm resistor actually 462Ohms? Does this make a difference when it's in a circuit providing a non-critical function. Audio passing through capacitors can get filtered, the HF being gently rolled off as it passes. If the component is within spec, then I cannot truly subscribe to the theory that the expensive one sounds better. I do not believe cables exhibit sound improvements as long as they too are within spec. I cannot ever see me having capacitor changes in an amp like I'd change strings. If they dry out, or do the bulgy thing, then they need changing because they are faulty.

There are poorly designed circuits, but they'd still be poor with the worlds most expensive components, and good circuits can be excellent with cheap components. That's my own reading.

we sure have come a long way. from a place and time where engineers with degrees who actually built their gear were doing the recording to this, not even a rudimentary block diagram understanding of how amplifiers work. that's what happens when a person skips over all the 101 stuff.

"i are a mixer!" ....... lol.

Does anyone remember Tony Waldron who engineered many of the Cadac consoles before he retired - he was always willing to explain things, but I was alway left way behind he got excited and started to talk about micro-detail, but I learned lots from him, even though often he'd get really uptight that people got the wrong end of the stick, and he HAD to put them right.