[sdiy] Temperature compensation results

[Ian Fritz]

At 02:31 AM 6/11/2003, Czech Martin wrote:
>Ian,
>
>finally someone comes up with the question of offset voltage
>in the transistor pair and if it is perhaps usefull, or
>harmfull to the stability.
>
>My observation is that across a few cm 0.4C temperature difference
>can exist even with forced air motion and thus heat exchange.
>I'd expect (have no measurements) that in the real application
>the temperature difference will be larger, depending on board orientation
>faceplate and everything else. There may be even draft or convection
>inside the instrument.
>
>The dual transistor expo circuit is a local approach, i.e. the reason
>of drift is known to be the transistor, so another transistor
>is integrated, or glued to that very source of error in order
>to compensate. In the next level a tempco resistor is glued
>to the pair in order to kill the remaining effect.
>This works quite well, because it is quite well known what is happening.
>
>I *assume* that a non local compensation approach (i.e. compensation
>circuit not near to physical thermo process that you want to compensate,
>and perhaps physically not even related to the bunch of residual tempcos
>in the circuit) will suffer from such temperature gradients.
>I could imagine that even the sign of compensation may be wrong
>under certain circumstances.
>
>So I wonder if a drift in the range of 10ppm/K is really possible.

Martin --

All these are things I have been thinking about.  My circuit was 
constructed in two parts: converter on one board, VCO core on 
another.  Thus I could test the drift of each part separately.  After that 
they were bolted together with spacers (like a sandwich) with the component 
sides of the boards on the inside, facing each other.  I envisioned putting 
thermal insulation around the edges, but I haven't done so yet.  This 
configuration is meant to reduce thermal currents and to keep all the parts 
fairly close together.  The whole sandwich is mounted in an Al box to make 
an isothermal surround and to keep out drafts.  It seems to work quite 
well, but there is more work needed to check reproducibility.  I worry more 
about turbulent drafts than gradients, but I'm not sure about this yet.

All these factors make a good case for going to a miniature construction.

Anybody know of a decent monolithic surface-mount transistor array?

   Ian