Alternative offset correction for Alephs.

[Netlist]

After reading William's post about the zener in the current source of the Alephx (Grey’s schematic) I was thinking if it would be possible to design a simple circuit to compensate the zener temperature.
I can confirm that warming up or cooling down that diode has a great influence on DC behavior. Suppose we tighten a PTC to the zener and control the amount of current through it with the offset voltage of the amp. At startup, the DC is rather high so a higher current could run and warm the PTC and thus the zener. The offset diminishes due to the warmth, voltage and current go down, PTC gets cooler. There you have it, an alternative DC servo.

Pro’s and con’s if at all doable?

/Hugo

[Magura]

I have been doing something like that, just applied to mosfets instead.

I simply ran a PTC off the rails to heat up the fets to working temperature as fast as possible, with the right PTC it will quit heating as the component reaches working temperature. To apply that to a zener I would make the PTC heat the zener a bit all the time, so it will be the PTC determining the temperature at all times.

Magura

[Netlist]

So, a simple resistor as current limiter and a PTC would do the job? Guess I'll give it a try.

/Hugo

[Blues]

Another alternative if you want to keep that zener warm is to keep the front-end powered up. To do this you will need a switch (dual On-On with the right DC V/I rating) that disconnects R14 and R31 (referencing Grey's AX) from the +rail and at the same time disconnects the -rail and grounding the front end Drain resistors R23 and R25. This takes the output section out of the equation. It will also keep the 25mA or so bias running and depending on your +rail value it'll use up less than a watt to keep it warm. That's now your nice standby switch.

[wuffwaff]

Blues,

keeping the front end warm won´t help as this will barely raise the temperature inside the amp case. Even taking of the lid when warm has a big influence.

The easiest way without PTC would be to find a zener with the right temperature coefficient. This should be possible cause when changing from standard zener to ref. zener the abs. dc off´set started positive instead of negative......

I´m thinking of using an active offset correction to make it possible to set different bias points with a simple switch.


William