signal loss

[{ antonio }]

how much cable loss is really happening between your guitar and the amp if you use several effects?

a friend and i got into a discussion about this the other day.  and he was discussing hwo low the resistance is on monster cable and how great the signal loss was on using several effects.  

i just wanted to see what everyones thoughts were on that and see if anyone had sone any research or found any sites on the matter.

thanks in advance.

[Paul Marossy]

I have read that the typical straight (non-coiled) guitar cable has a capacitance of around 5-10pF. Not much, IMO.

In fact, using a coiled guitar cable like Hendrix used, can soften up an otherwise overly bright sounding guitar amp. Don't know the capacitance is on one of those - maybe 30pf or so?

Now, if you want to talk "oxygen-free copper", well I think that is non-sense. There is something to it, a real effect, but I think it would only matter on the microchip/nano technology level. On a guitar cord, I just don't buy into the hype. I think it just gives the manufacturers an excuse to charge more.

[R.G.]

how much cable loss is really happening between your guitar and the amp if you use several effects?

a friend and i got into a discussion about this the other day. and he was discussing hwo low the resistance is on monster cable and how great the signal loss was on using several effects.

i just wanted to see what everyones thoughts were on that and see if anyone had sone any research or found any sites on the matter.

You know, the great thing about tehcnical learning is that it helps convert "I wonder" and "has anyone ever" kinds of questions into things you can calculate and actually know, instead of gathering opinions.

The short answer is that the cable capacitance, not the cable resistance matters, and that you can mitigate the effect of cable capacitance with an on board buffer. Cable resistance makes so close to no difference at all that if someone tells you about low resistance cable for your guitar, that statement is a reliable guide that they don't know what they're talking about.

The signal losses involved with guitar cables is fully knowable and easily calculable. To calculate, you need to know
(a) the impedance of the pickup and how it changes with frequency
(b) the cable's distributed capacitance and inductance as well as resistance
(c) the input impedance of the amp or effect that the cable connects to.

If you use a series of effects, then you need to know both the output and input impedance of each of them, as well as the cable specs for each cable between them.

In general, you model the signal as a signal generator of some signal voltage (which doesn't matter much, only the ratio of generated to lost matters in this) and some source impedance. That generator is connected through a cable of some series and parallel impedance to a load, which is the input of an effect or an amp.

The generator voltage is divided down in the ratio predicted by ohm's law by the generator impedance, the cable impedance, and the amp impedance.

If the cable and amp impedance is greater than about ten times the generator source impedance at all frequencies you care about, then you have no problem - the losses will be inaudable to most people. It's only when the amp or cable impedance to ground gets down near the generator source impedance that signal gets eaten.

Tube amps are quite good. They're usually an input impedance of 1M or greater, which means that the cable and pickup can have a resistance of up to about 100K before you'd start noticing loss, if that were the only loading. That's where I get the "don't know what they're talking about" - anyone thinking low resistance in a guitar cord is necessary is ill informed.

The problem with signal loss is with the pickup inductance and the cable capacitance.

The pickup is a coil of wire around some magnetic stuff - that's one good definition of an inductor. An inductor has an impedance that goes up with frequency. The cable is a wire surrounded by a tube of ground shielding. That make a pretty good capacitor. As the signal frequency goes up, the pickups have to push through more and more pickup inductance impedance, and what gets through there is shunted to ground by more and more cable capacitive impedance which goes *down* as frequency rises. Since you can't change your pickup inductance much, more cable capacitance give more treble loss, simple as that.

Shielded cable is all rated in capacitance per foot. "normal" values are 10pF/ft (very, very good) to 150 or more pf/ft. And that is "per foot" so the longer the cable, the more you lose treble.

There are things you can do. If you put a buffer in the guitar, the buffer will take the pickup signal before the cable capacitance gets a chance to degrade it and re-output it as the same signal level but from the buffer's very low output impedance. The buffer now supplies the muscle to overpower the cable capacitance and keep the treble frequencies in the signal, not let them bleed away.

Any time someone wants to sell you a radically expensive piece of wire, be really, really sure you understand what is going on at a technical level. In the audio world, almost all such claims are the purest nonsense; at best there may be a tiny grain of truth inside a 100 pound bag of nonsense.

[Paul Marossy]

I think I meant to say 5-10pF per foot.  :oops:

[Hal]

wouldn't a coiled cable add some inductance, and actually cancel out some of the normal capacatince of a cable... ?? ??

[R.G.]

wouldn't a coiled cable add some inductance, and actually cancel out some of the normal capacatince of a cable... ?? ??

1. Inductance and capacitance do not cancel. They're different.

If resistance is straight ahead, and capacitance is a turn to the right, inductance is *up*. Literally. They all act at right angles to each other in the electronic sense.

Inductive impedance and capacitive impedance may be equal and opposite, thereby "cancelling" each other at one frequency for any given values of inductor and capacitor. Audio is notoriously not one-frequency, so what you get with any inductor/capacitor in audio is either a notch filter or a resonance.

2. The inductance of a coiled cord is nominally of no consequence to the signal inside the cord because the magnetic field generated by the signal going one way through the axial conductor is countered by an equal and opposite magnetic field generated by the signal going the other way through the shield, for a net field of zero. No field, no inductance.

If you paralleled the axial lead and the shield and ran current through them in the same direction, then the coiled cord would indeed look like an inductor, just like if you ran signal through only one at a time.

[Hal]

mmmmm i understand the 2nd part, but not the first.  I always thought that if resistance was forwards, inuctance is left and capacatince is right...that they're opposite vector quanitites.  I need to study up more :-D hahaha

[{ antonio }]

wow thanks for the reply.  these are great.  javascript:emoticon(':D')
javascript:emoticon(':D')

[niftydog]

I always thought that if resistance was forwards, inuctance is left and capacatince is right...that they're opposite vector quanitites.

In the theoretical ideal world, that might be so.  But reality bites.

It's more like, resistance is forward, inductance is clockwise rotation and capacitance is anti-clockwise rotation!

Look up "imaginary algebra" or "complex algebra" and vectors etc. for a more detailed explanation... but only if you're keen!  It can be quite a mindbender!

[Brian Marshall]

resistance is only important for your speaker cables.....

what is usually the first of second thing your signal sees in a pedal.... usually a 100k plus resistor. the cable should be only a couple ohms at the most.

[{ antonio }]

i just read this on pete cornish site

"Take for instance a 15 ft guitar cable linked to ten pedals, each linked by a 2 ft cable, and then onto the amp by a 30 ft cable. If all pedals have "true bypass", and are off, then the total cable length hanging on the guitar output will be 63 ft. This will cause a huge loss of tone and signal level particularly if the guitar is a vintage type with low output and high impedance. The amp volume is then turned up and the treble control increased to compensate for the losses. The inherent background noise now increases by the amount of the gain and treble increase and is usually, in my experience, too bad for serious work. If one of the pedals is now switched on, then it's (hopefully) high input impedance (and usually low output impedance) will buffer all the output cables from the guitar and the signal level will rise due to the removal of some of the load on the pickups (i.e.: 17 ft instead of 63 ft of cable). The treble will rise and the tone and volume will not be as before. If that pedal was say a chorus or delay, devices which are usually unity gain, then your overall signal level and tone will vary each time an effect is added...not a very good idea."

in it he goes through and discusses the idea of signal loss through the mass length of cable.  granted i kind of see where he is coming from but i wanted to know what you'all thought about the tone variation idea.  i couldnt imagine it being dramatic.

[Paul Marossy]

It could be pretty noticeable to someone with a good ear. Figure this: suppose your cable is 20pF per foot. Out of the 63' feet of cable quoted, that would be 1260pF or approx. 0.0013uF. Enough to take off some of the high end of the signal. With weaker vintage pickups, it could be a factor.

[Hal]

i realize that 'good' cables are ususally gimicks, but bad cables cause more of a problem...i bought a 30' cable for $10, just for occational gigging use...and it kills my signal so bad its terrible  :twisted:  :twisted:

but...i dont notice anythigh with any of my other cables or setups.

[smoguzbenjamin]

See, I don't worry about treble loss. My old guitar is too trebly anyway so I actually like my lo-grade cable/lo-grade PU combo

[Ben N]

Since you can't change your pickup inductance much...

Isn't that what the Bill Lawrence Q-filter thingie is supposed to do?

Ben