On Thursday 31 July 2003 05:03, Alan B. Pearce scribbled:
> > > There will be a current flowing thru the divider, and therefore thr=
u
> > > the temp sensor, and the self-heating will cause the sensor to chan=
ge
> > > output resistance.
>
> Why not just do a calibration run with the LM35 while the sensor is pow=
ered
> in the intended circuit? This sounds like a KISS method to me.

Right.  That was the second option.  I really don't *need* the accuracy t=
hat=20
I'd get by compensating for the self-heating effect, but thought it would=
 be=20
a great experiment, and wanted to see how much of a difference it would=20
actually make.


> A second way to do it would be to have a (relatively) large metal clip =
on
> the sense resistor, so that the self heating energy is dissipated into =
the
> surrounding air. I am thinking in terms of a little sheet metal clip on
> heatsink for a TO92 sized transistor here, just to provide a radiating
> flag. This would keep the sensor temp closer to the temp being measured=
,
> but could provide some thermal lag, especially when the temp drops.

Well, it seems I'll have a "(relatively) large metal clip" on there alrea=
dy=20
=2E.. it's the engine block.  Now THAT's a real heatsink!
:-)

Cheers,
-Neil.

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads