Why loading a guitar pickup loose high frequencies

A few years ago I did some experiments with guitar preamp input impedance. I built a high impedance preamp -- input impedance of 3 megohms or so -- and then tried different shunt resistors across the input to see how they affected the sound. The guitars I plugged into it were a Fender USA Stratocaster of 1980s vintage, and a Gibson Les Paul of about the same age. With the Strat I was barely able to detect a loss of brightness (high frequency production) with about 200 or 300 kilohms for the shunt resistance (i.e. a 200-300k loading of the guitar and guitar cord). The volume and tone controls on the guitars were all full on. This was a barely detectable loss of highs, listened to through good quality headphones. Through a guitar speaker, you wouldn't notice it.

Interestingly, the Gibson, which had humbuckers and therefore, I would expect, a higher inductance, could drive a lower load resistance without noticeable loss of highs.

At the other end of things, you don't want too high an input impedance because it results in more noise, especially EMF interference. Things get noticably worse past 1 megohm.

So my conclusion was that, if you're only plugging the guitar into a single input, that input should have an impedance of about 330 - 470 kilohms. That's high enough not to kill much treble, but no higher, so that noise is minimised. In practice, a higher input impedance may be desirable because some people plug a guitar into two amps at once using a simple Y splitter on the cord; then you'd want about double the input impedance on each amp. So perhaps the traditional Fender amp input impedance of 680 kilohms is about right.
 
The pickup, the volume control, the tone capacitor, and the cable, essentially reduce down to a RLC resonant circuit. Increasing 'C' (via the cable) lowers (e.g. shifts down in frequency) the resonant peak of the combination of components, which makes the perceived tone 'darker'.

An active buffer between the output of the guitar and the cable will shield the RLC circuit from the capacitance of the cable, because the output Z of the buffer is small in comparison to the RLC circuit.
 
Consider the factory wiring of a fender strat and you will find it loaded by the vol pot (250k or so), and by the tone pot as well. Thus input impedance above 500k will make no difference.

But if you rewire the guitar and put your pre-amp close to the P.U., i.e. you create an active p.u, you will notice differences upto 1meg load resistance. This is what I did.
 
Guitar pickups have massive inductance but others including the Rhodes piano are worse with many coils in series. The inductance is so high that the impedance of the guitar cable produces a noticeable effect. This is why Les Paul made Low-Z instruments.
GIBSON Les Paul: Low-Impedance Series electric guitars

Unused inputs should short themselves and loading a pickup is not going to improve the s/n. This is an advantage of valve amplifiers that is rarely discussed, but can be achieved in solid state amps if the issue is well understood. A input bootstrap circuit produces a hi-Z input and may add some 2nd order harmonics.
 
Last edited:
Hi Steveu, you say "loading a pickup is not going to improve the s/n." I'm not sure about this. If you model the pickup as an inductor and resistor in series, it has limited ability to shunt induced EMF noise voltage from the environment to ground, because of its inductance. Seems to me an added shunt resistor can help to ground out some noise, especially at high frequencies.
 
Buffer the pickup at the guitar, if possible.

This is the best approach.

However just buffering the high Z guitar pickup in guitar is no ticket to great sound. Every single high Z pickup want's to see a C + R loading to finely tune
it's response and feel. There is a sweet spot or to be more accurate, many sweet spots and conversely many loadings that don't sound particularly good.

For example you can get a very high OP humbucker to sound very bright but it will definitely not sound very good.

TCD
 
The interesting thing is that no other guitarist went on to emulate Les Paul, and neither did any guitar pickup manufacturer.

They've had many decades in which to do so if they wanted to. The obvious conclusion is that nobody wants to.

Listening to old Les Paul recordings and videos on You TUbe, to my ears, the guitar tone that Les Paul got with his low-Z pickups is quite unpleasant. Thin, cold, steely, and very unattractive.

I suspect that most other guitarists who came after Les Paul (his actual name was Lester Polsfuss) feel the same way I do. Which is why nobody else found low-Z pickups with extended treble response pleasant. Polsfuss was the last guitarist who thought they were a good idea.

(IMO Les Paul's tone was awful, but his playing, on the other hand was fantastic. He was a heck of a guitarist. Even though he constantly lied about being the inventor of the electric guitar. That was actually George Beauchamp, who was selling his Rickenbacker A-22 "Frying pan" electric guitars in 1931, while Polfuss only put together his "Log" some eight years later, around 1939.)

There are actually still low-Z magnetic guitar pickups manufactured today, but they are specialty items used only on electro-acoustic guitars (like some of Taylor's offerings), or on solid-body guitars trying to sound like acoustic guitars (like the Shadow Nanomag pickup installed as a third pickup on some Epiphone Les Paul Ultra models). Their purpose is to sound more like an acoustic guitar than an electric guitar.

I once did some experiments with a guitar, an extended-response piezo pickup, and a graphic equalizer. I used the EQ to filter out all frequencies below 5 kHz or so, so I could listen to just the remaining high-frequency output.

I found that what remained sounded harsh, scratchy, ear-fatiguing, and unmusical / out of tune. At least to my ears, the output of a guitar above 5 or 6 kHz is not musically attractive at all. This applies to both electric and acoustic guitars.

The "out of tune" part, by the way, is predicted by the mathematics of vibrating-string theory; because actual guitar strings have stiffness (they are not mathematically ideal perfectly flexible strings), their output is not a pure harmonic series - higher frequency overtones are increasingly "out of tune" and more dissonant with the lower-frequency tones.

Somewhat shockingly, the electric guitar is ninety years old this year. Like most century-old inventions, it has been nearly perfected along the way. There's always room for new ideas, but guitar pickups and amps have limited frequency response for a very good reason - they sound better that way.

These days, many guitarists play through a digitally modelled guitar amp and an digitally modelled speaker/cab simulator. It would be trivially easy to make a cab sim with a wide frequency response out to, say 20 kHz, if you wanted to.

But nobody does that, because the results sound horrible, like being stabbed in the ears with needles.

-Gnobuddy
 
  • Like
Reactions: 1 user