The mosfet boosters (and others) have very high input impedances, but isn't this limited by the pot in the guitar (typically 250K or 500K)? For example, the pickup would see an impedance of a 500K pot with a pedal of 5M, which is 455K - the parallel result of the two impedances. I must be missing something.

No, you're absolutely right. That's why there is no point in having greater than a 1M resistor in parallel with the input of an amp or pedal.

I think the 1M thing is largely historic - it being the approximate input impedance of a valve amp. I notice that many people are now talking about using 1M pots on guitars rather than 250k or 500k.

1M pots, as I remember, had virtually no sweep to them. Almost like a switch. First 70% of rotation gave no/next-to-none effect, and then a sharp and abrupt decay in volume.

The law of diminishing returns gets you in almost every instance, doesn't it?

Your analysis is correct, mostly. Guitar pickups lose treble to any loading, including the guitar's onboard controls.  Keeping all the sparkle requires keeping the impedance on the pickups high everywhere.

The rule of 10:1 for significance applies here. That is, something is significant compared to something else if it's within a factor of 10 of the something else. If it's effect is less than 1/10 of the thing you're comparing it to, you can neglect it to a first order approximation.

For instance; input impedance of an effect loading a guitar is insignificant if it's greater than 10 times the guitar's output impedance, as your calculated example shows. The most significant loading is the volume control (and probably tone control as well).

If you really want sparkle from your guitar, you'll put in an onboard buffer with greater than 1M  impedance as seen by the pickups and an output impedance of less than 10K. That makes volume control and cable loading insignificant, and the treble all comes through.

Quote from: Constantin Necrasov1M pots, as I remember, had virtually no sweep to them. Almost like a switch. First 70% of rotation gave no/next-to-none effect, and then a sharp and abrupt decay in volume.

I don't understand that - if the volume pot is acting as a simple voltage divider, it's effect on 'volume' is independent of the pot value. What you describe can only indicate a faulty pot or a change in taper.

Quote1M pots, as I remember, had virtually no sweep to them. Almost like a switch. First 70% of rotation gave no/next-to-none effect, and then a sharp and abrupt decay in volume.

My Jazzmaster has a 1M pot and the sweep is fine.

There is definately a sparkle added by the high impedance boosters. In my simple example a pedal with a 500K impedance (and the 500K pot) would be 250K vs. 455K. So maybe that's what I'm hearing.

An onboard preamp will definately do the trick. What about adding a switch that bypasses the guitar's pot(s) with a 5M resistor? Now the impedance with the pedal would be 2.5M. I'm assuming there would be an issue with noise?

I have a 1M pot in my guitar. Since all my stompboxes have about 2M input impedance, the sweep is quite comfortable.
At the same time I have to mention that it requires ten times more thorough shielding relatively to 250k volume pot and 1M stompbox input impedance.

То есть заземлять надо лучше цепь с мегаомными потенциометрами? Или наоборот с 250кило?

Константин,
изначально у меня в гитаре стояли поты 250 кОм. Я перепахал все гитарные потроха - кроме звукоснимателей - и поставил на громкость 1 МОм пот.
Гитара стала ловить всякие помехи В РАЗЫ сильнее. Эта разница явно проявилось, когда я сижу перед тем же самым монитором, что и раньше, а шуму и треску стало столько, что струны еле слышно.
Мне пришлось заново тщательно все заэкранировать, и это вылечило ситуацию.
Резюме: чем выше сопротивление цепей в гитаре, тем лучше надо экранировать.
Если будешь ставить 1МОм пот, то экранируй все внутри гитары сплошной латунной фольгой, и шнур юзай с хорошим экраном тоже. С плохим шнуром будут сильные помехи, когда ГРОМКОСТЬ СТОИТ НА ПОЛОВИНЕ - это из-за наводок на шнур.

Олрайти! Пасиб!

it seems crazy, but a 5 Megohm input impedance on a booster has a dramatically higher sparkle content than a 1 Megohm input impedance.  this is a result of the output impedance of the guitar pickup varying dramatically with frequency.  you start with a pickup that measures 5K-8K ohms or so and connect it to a 250K or 500K pot.  that pot is a fixed resistance load which brings down the high frequencies of the pickup by a certain amount (quite a bit, actually), but doesn't affect the mids and lows much, because the pickup's output is lower-impedance at those frequencies, so it can more easily "drive" the pot.  what's left of the highs is still at a pretty high impedance (essentially that of the pot), while the mids and lows remain at a lower impedance.  now you drive the cable with the output of the guitar, and more of the highs are lost to cable capacitance.  when you get to the end of the cable, the ideal destination would be at least 10 times the output impedance from the guitar... if you have a 250k pot you'd need a load of 2.5M or more, and if you have a 500K pot you'd need a load of 5M or more to protect from significant signal losses.  if you use a destination impedance of 1Megohm like a typical amp, you lose a lot of high-end that's coming from a delicate source, but you don't lose much mids and lows.  if you use a mosfet booster with a 5Megohm input impedance, you'll hear more of what the guitar's output really sounds like.

if guitars had been originally designed with lower-impedance pickups with smaller windings, the pots could have been lower (100k or 50k) and the 1 meg input impedance of the amps would have made more sense.  it's not an ideal system... the history of guitars and amps is really a lot of compromised circuitry that basically has to "make do."  8^)

what it really comes down to is, most guitarists don't need that extra sparkle they get from a 5Megohm input impedance.  there are a few who really love it... but a lot of guitarists feel it's too hard and icy.

To be frank, I’m not sure that there is a great difference between 1M input impedance and 5M impedance in parallel with the volume pot that is 500k.
500k impedance in parallel with 5M gives 420k, 500k in parallel with 1M gives 330k.  420 divided by 330 is 1.25 approx., that is less than 3dB (3dB is about 1.4).
So we have to assume that our ears easily catch this 1.25 times difference in hi-frequency levels.
Or, maybe, the effect of the “sparkling highs” is just imaginary phenomenon caused by some specific of the MOSFET booster stage.
At the same time it’s not a problem to make another type preamp with 5M input impedance using not MOSFET but simply JFET. Maybe it will help to reveal the secret…

QuoteAt the same time it’s not a problem to make another type preamp with 5M input impedance using not MOSFET but simply JFET. Maybe it will help to reveal the secret…

It's not a problem to make a 5M - or 50M - input impedance with an opamp or other amplifier. You just bootstrap the input.

I believe the "sparkle" effect Z refers to is in fact accentuated by the inherent residual distortion of the MOSFET setup.

So, we should decide that excessively high input impedance is not the most important  factor to achieve sparkling highs?
It seems to be true, and elementary arithmetics gives quite convincing proof.
Besides, one can easily add 3 dB of highs with an extremely simple circuit such as one capacitor and one resistor. But it won't give any "ice" or "glass" to the tone.
...enigmatic MOSFETS...unpredictable little things...

Quote from: R.G.

QuoteAt the same time it’s not a problem to make another type preamp with 5M input impedance using not MOSFET but simply JFET. Maybe it will help to reveal the secret…

It's not a problem to make a 5M - or 50M - input impedance with an opamp or other amplifier. You just bootstrap the input.

I believe the "sparkle" effect Z refers to is in fact accentuated by the inherent residual distortion of the MOSFET setup.

believe what you will.  the sparkle effect features just as prominently in my lo-fi loop junky's direct sound... and guess what it uses as an input device?  a 2n3904, bootstrapped.

the same effect occurs every time you give the guitar breathing room.  give it an input impedance of 5Megs, and it will sound sparkly.

sometimes theory is just...  theory.  i'm a practical man... i MAKE stuff for a living.  8^)

Quote from: DDDSo, we should decide that excessively high input impedance is not the most important  factor to achieve sparkling highs?
It seems to be true, and elementary arithmetics gives quite convincing proof.
Besides, one can easily add 3 dB of highs with an extremely simple circuit such as one capacitor and one resistor. But it won't give any "ice" or "glass" to the tone.
...enigmatic MOSFETS...unpredictable little things...

the words i chose are subjective descriptions.  to my ear, it's sparkle.  to someone else's, it's ice.  you say tomato...

Quotebelieve what you will.

Thanks, zachary. I always do anyway, but it is a great comfort knowing that I have your full support.

Quotesometimes theory is just... theory. i'm a practical man... i MAKE stuff for a living. 8^)

Yep. Sometimes theory is just... theory.

But sometimes, just occasionally, it lets you figure out something without wasting days or weeks swapping parts in and out in search of your own beautiful reality.

not to be snarky about it, but RG, i've built a lot of mosfet booster circuits and i've compared them to a lot of bootstrap circuits, and the sparkle is very similar.  that's all i'm saying.  you gotta bulid it and listen if you're ever gonna know.  8^)  in the end, theory genuflects at the feet of the empirical truth.  ask Einstein.

time to bookmark this thread... :wink:


btw: it makes a big difference, if you have 1M vol-pot in the guitar,
and then a receiving 5M input-Z "miles" away - somewhere on the floor at the other side of the stage;
or if you have those 5M built-in your guitar.

Same difference goals for 1M of input-Z...

it`s that unknown parallelled capacitor named "cable" ...

(unless I wanna build a lossy low-pass, I too prefer a high impedance to follow such an RC-member).