1970s Design Indulgence

Justification for the panel meters and front panel Vq trimmers has been happening quite a lot at my location.

We had two brief power outages yesterday morning, in addition to the ones a few months ago. Mains voltages have now varied between 235V and 254V.

The 245V transformer tap is probably the safest one for me to use, but it does mean that mag-current either falls and there's less to "swamp out" power supply harmonics, or it increases and there's a bit of chassis noise due to the increase in magnetostriction of the transformer.

The resultant change in HT voltage is approximately 4%, and that affects the VAS collector current by the same percentage. NFB prevents DC voltages drifting astray, but the increase or decrease in VAS collector current leads to a inversely proportionate increase or decrease in output stage standing current.

This results in a quiescent voltage drift of +/- 0.75mV and then you have drift from changes in ambient room temperature. I've noted a variation between 14mV and 18mV.

The exact setting is where 1W THD is at its lowest and the higher harmonics are seen to be falling on an FFT display, and is always at approx. 16mV.

The lowest 1W THD is approx. 0.008%, but at the extremes of variation of the power supply, will rise to approx. 0.02% which is about -74dB and so should be inaudible.

Even using the base spreader circuits proposed by such as Self, there is still going to be equivalent drift for the amount of power supply difference.

There really isn't any set it and forget it point, and the only way to prevent 1W THD rising under these circumstances is to make it disappearingly low in the first place.

The exception being voltage regulation to the voltage stage, but as that needs to be at "the point of use", which requires the integration of an IC voltage regulator - which are now no longer manufactured - then, without any PSU regulation, some degradation of the 1W THD is to be expected.

The front panel trimmers and the meters allow the required micro-adjustment to keep the 16mV setting, which will then stay there for as long as the mains voltage stays put, and there aren't any major temperature swings in the listening environment.

Can a change of +/- 2mV be heard? Not that I've noticed! But if it were to drift too far, then maybe.

Another thing that's slowing down work on the amplifier is the third world mains voltage provided by Berkshire Hathaway Energy through its Northern Powergrid.

Sound quality depends on there being sufficient mains voltage for the modified toroidal transformer to attain a magnetising current of circa 100mA.

The prototypes had untapped primaries to suit my supply, which varied between 240 and 250 volts, so the primary was rated 245 volts.

The mains voltage here is now 228 volts.

I wonder if Mr Buffett realises how much money his company is losing him?

(P = V * I)