![[-]](https://theultimatetone.com/images/collapse.png "[-]")
04-15-2026, 10:09 PM
Hi Guys
Precision Power Scaling: Part Gamma
Sometimes when you spend too much time immersed in a design process it can be difficult to see what turn out to be very obvious flags.
BIG FLAG: potentiometer value tolerance.
A survey of pot data sheets shows that regardless of the price and the nominal "quality" nearly all pots have a 20% tolerance just like any other resistor from <1950 or so. This means that a 1M pot can be 1,200,000 ohms down to 800,000 ohms despite being labelled as 1M (1,000,000). Numerically, this is a HUGE spread. And it makes them fairly useless where exact values matter.
Fortunately, there are ways to design circuits that accommodate pot value spread in such a way that the actual pot value is irrelevant. For example, active Volume or Gain circuits use a pot in the feedback loop of an opamp. Most of the circuits of this type have terrible gain errors, and in a stereo, the channel tracking is egregious. In the example, Peter Baxandall created a solution for the electronic gain control where the pot value is not a part of the gain equation. Brilliant as usual, Peter!
For my Precision Power Scale circuit, the core is precise and accurate. The problem remained that the imprecise pot value is in a precise voltage divider and the pot's old-timey performance throws everything into a cocked hat. This turns out to be what I have been battling since after the Super Budget kits; the "amplified pot" was imprecise because the pot tolerance is so wide. Until now.
"Third time is the charm", as they say, and the third shot at Precision Power Scaling is truly complete and precise. With a minor but highly significant addition to the circuit, the pot value no longer matters. It can in fact be changed from 50k to 1M without alteration and have the full and proper sweep of output.
Were it so sweet? Looking more closely at pot tapers from Alpha, regardless of taper there is a flat spot at both the CW end and the CCW end. These amount to 5-10 degrees of rotation at each end. The data sheet for RV4 mil-spec pots show an even worse loss of sweep and these are 100 times the price! The only pot type that does not have this characteristic are ALPS stepped resistor 27mm units. These are a 42-position switch with resistors between the steps. Not much resolution, but maybe it is practical and it has repeatable settings. The physical size is problematic for most amps as a modification.
The new SV1-Z and SV2-Z PCBs are on their way.
Precision Power Scaling: Part Gamma
Sometimes when you spend too much time immersed in a design process it can be difficult to see what turn out to be very obvious flags.
BIG FLAG: potentiometer value tolerance.
A survey of pot data sheets shows that regardless of the price and the nominal "quality" nearly all pots have a 20% tolerance just like any other resistor from <1950 or so. This means that a 1M pot can be 1,200,000 ohms down to 800,000 ohms despite being labelled as 1M (1,000,000). Numerically, this is a HUGE spread. And it makes them fairly useless where exact values matter.
Fortunately, there are ways to design circuits that accommodate pot value spread in such a way that the actual pot value is irrelevant. For example, active Volume or Gain circuits use a pot in the feedback loop of an opamp. Most of the circuits of this type have terrible gain errors, and in a stereo, the channel tracking is egregious. In the example, Peter Baxandall created a solution for the electronic gain control where the pot value is not a part of the gain equation. Brilliant as usual, Peter!
For my Precision Power Scale circuit, the core is precise and accurate. The problem remained that the imprecise pot value is in a precise voltage divider and the pot's old-timey performance throws everything into a cocked hat. This turns out to be what I have been battling since after the Super Budget kits; the "amplified pot" was imprecise because the pot tolerance is so wide. Until now.
"Third time is the charm", as they say, and the third shot at Precision Power Scaling is truly complete and precise. With a minor but highly significant addition to the circuit, the pot value no longer matters. It can in fact be changed from 50k to 1M without alteration and have the full and proper sweep of output.
Were it so sweet? Looking more closely at pot tapers from Alpha, regardless of taper there is a flat spot at both the CW end and the CCW end. These amount to 5-10 degrees of rotation at each end. The data sheet for RV4 mil-spec pots show an even worse loss of sweep and these are 100 times the price! The only pot type that does not have this characteristic are ALPS stepped resistor 27mm units. These are a 42-position switch with resistors between the steps. Not much resolution, but maybe it is practical and it has repeatable settings. The physical size is problematic for most amps as a modification.
The new SV1-Z and SV2-Z PCBs are on their way.
