relation between ampere, voltage and distortion or linearity

[oliverniekamp]

In the original zen artikel Nelson mentioned the relation between amperes and distortion, double amperes means half the distortion. Its like a reverse proportionality. In general, increasing voltage and increasing amperes both can increase linearity.
What is the relation between voltage and distortion?
It would be interesting to make some experiments to find out what differences are between, for example 25 volt 3 amps and 37 volt 2 amps, in relation to a given speaker impedance of 4 or 8 ohms. Im not asking for the watts coming out from the amplifier but for the objektive and subjektive differences between the voltage and ampere combinations.
Does someone have ideas or experiences on this question?

[jan.didden]

I don't remember Nelson's article verbatim, but I would gamble he said one of two things (or maybe both):

- "more amps CAPABILITY" means less distortion, as in doubling the number of output devices increases the current capability (or halves the current per device, which comes out to the same) hence less nonlinearity (distortion) everything else being the same;

- "more quiescent current" (as in Class A) means less distortion.

Are you sure you read his statement correctly (the devil, as always, is in the details)

Cheers, Jan Didden

[13DoW]

The distortion will be related to the amount of current change in the output stage of the amplifier to supply the loaudspeaker load. In the case of Zen the output stage is also the gain stage.

To drive a given voltage across a loudspeaker a corresponding current will be drawn depending on the impedance of the speaker so the bias current must be set high enough to supply the maximum expected load current. However, the larger the proportion of the bias current supplied to load then the higher the distortion becasue the gain transistor does not have a linear voltage-in to current-out characteristic. So, if you double, or more, the bias current then signal current excursions around the bias level are proportionately lower and so is distortion.

Nelson Pass takes this argument to the extreme in his description of single-ended amplifiers by comparing class-A biasing to air pressure. Air pressure has a (reasonably) static value and sound waves cause an excursion about that static level. However, the static value of air pressure is ~10^5 N/m^2 and the loudest sound (threshold of pain) is ~30 N/m^2. So, the loudest sound casues an air pressure excursion of 0.003% of the static (bias) value. Translating that into amplifier biasing terms, 10W into 8Ohms is a 0.5A peak excursion so for the analogy to hold the bias current should be 1,666A (I trust Mr. Pass had his tongue somewhat, if not wholly, in his cheek when he wrote this).

Regards
13th Duke of Wymbourne

[Nelson Pass]

Actually I believe my tongue was extended from my
cheek at the time.

[PMA]

relation ...

- the higher quiescent current means less distortion,

- the higher MOSFET transconductance means less distortion, too,

- 4 Ohms speaker will have more distortion compared to 8 Ohms one.