"Mojo" Parts in Guitars, Effects, and Amplifiers Part 1: Resistors

July 29, 2010

There is a lot of speculation and opinion to be found around the internet about the "magic" of old carbon composite resistors, particularly in amplifiers. People talk about "harmonic swirl" and "breathing" and other vague but encouraging terms. This is a poorly understood concept, but we'll attempt to shine a bit of light on it.

Things to keep in mind:

First, carbon composite resistors were chosen in "the good old days" not because they sounded great, but because they were inexpensive and the only real choice for many years. When you read this whole article you'll realize Fender and Marshall and Denney wouldn't be reaching for carbon composites if they were designing their amps today.

Second, carbon composite resistors have a 5-10% tolerance range when new, which can increase to 10-20% or higher as the resistors age and are subjected to heat and voltage and humidity. For example, this means that the 220Kohm resistor used on the plate of V1 in the AC30, per the JMI schematic, may read anywhere from 176Kohm to 264Kohm in reality (they can actually swing higher or lower [usually they drift higher] though such resistors would be considered defective). This is a large variation, and you easily hear the difference in use. Even assuming most of the carbon composite resistors in an amp have stayed within the 10% tolerance range, you have a random cumulative shift in the entire amp well away from what the designer intended.

As we noted here, some of these "drifted" values actually sound good. If we want to reproduce the sound of a 240Kohm carbon composite resistor, and have it stay at 240Kohm, a modern carbon film resistor is a better choice than a carbon composite, as the carbon film resistor will stay well within it's 5% tolerance range for many many years.

Particularly in the output section, it can be desirable to have resistors as evenly matched as possible - this is difficult with carbon composites, and there is no guarantee the resistors will remain matched in coming years (it's highly unlikely, actually).

Third, while carbon composite resistors can in fact produce even ordered harmonic distortion in a manner we perceive as pleasing, they do so only when significant voltage is coursing through them. As there is only one resistor in the top boost channel of an AC30 that carries signal through it while being subject to high voltage [aside from plate resistors*], there is no need to use a carbon composite resistor anywhere but the slope resistor in the tone stack. At lower voltages, there is no pleasing distortion provided by these resistors, but there is a good bit of noise/hiss - not a good tradeoff.

* Plate resistors carry enough voltage and are so crucial to the sound of an amplifier that we do not like to use carbon composites there, as it is the one location most susceptible to drift yet exactly where one does not want drift to occur.

Fourth, carbon composite resistors are relatively large for their voltage and current capacity. The same space required for a 1/2W carbon composite resistor will also accommodate a 1W carbon film resistor. In addition to being more accurate and not subject to drift due to heat, voltage, and humidity, a 1W carbon film resistor has the added advantage of lowering background noise while also performing well above the voltage and current requirements in the circuit, ensuring a longer more reliable component life span.

Note: as a resistor's wattage rating increases, its self noise decreases. There's math behind this, but we promised you no math, so we'll just reduce this to: 1W is more beneficial than 1/2W, which is more beneficial than 1/4W, etc. Good practice is to use as high wattage a resistor as there is physical room for in a circuit (resistor size increases with increased wattage rating).

For the same reasons listed above, aside from internet-inspired "mojo" there is no good reason to use carbon composite resistors in a guitar effects pedal. They introduce wide variation and increased noise (hiss) but offer no aural benefit (high voltages are required to gain any slight benefit, and effects pedals rarely exceed 30vdc internally, more often running in the 5V-12V range).

So for an effects pedal circuit, the designer does well to choose the highest wattage resistor there is physical space for, and to choose carbon film or metal film resistors for lowest noise and the least amount of random variation.

As the voltages are so low in an effects pedal as to eliminate the sonic benefits of carbon film over metal film resistors, 1% tolerance 1/4W to 1/2W metal film resistors are our preference in effects circuits. In new and modified current production amplifiers, we prefer 1/2W to 1W carbon film resistors, save for some sections of the power supply where we choose 1W and greater metal oxide resistors. When restoring a vintage amplifier, we do use carbon composite resistors so as to maintain the highest value to a collector, but then we sort through many carbon composite resistors to select those that measure well within tolerance, and we match as closely as possible when needed (screen and grid resistors on the output tubes, for instance).

Continue reading:
"Mojo" Parts in Guitars, Effects, & Amplifiers Part 2: Capacitors