Tempco measurements on ASM-1

[Joachim Verghese]

Jörgen wrote:

> But if you *are* going to use a tempco, it is best to have the
> stable spot is high as possible, isn't it? Because above this spot,
> the tempco will worsen the error instead of reduce it.

I don't think so. If you do the maths for the expo converter
you will arrive at something like

Io = Iref * exp(-Vi/Vt)

where Iref is the reference current (PWR+/R15), Vi the control
voltage at the base of Q1, and Vt the thermal voltage. It is also
assumed that the two transistors are "Vbe-matched".

As far as I can see, the polarity of Vi doesn't affect operation
even if you use a tempco resistor.

Just as a side note, if the transistors were not Vbe-matched,
the output current could be expressed as

Io = K * Iref * exp(-Vi/Vt)

where K is the ratio (Icq1/Icq2) of the collector currents of
the two transistors measured at a nominal Vbe.

I think an Exponential Converter Tutorial would be a great thing
to have online. Gene Z?

-joachim