Re: An analog challenge (and a real problem)


Don Black <donald_black@...>
 

I missed the start of this thread. Are you trying to measure the temperature inside an IC that's performing other functions with a spare junction or just using a junction in an IC as a temperature sensor? If it's just a thermometer why not use one of the temperature measuring ICs such as the LM335 or something newer.
If it's the wire resistance you're worried about, why not either just measure its resistance or calculate it from the gauge and length. Since it's a pure resistance you should be able to compensate for it with a simulated negative resistance (effectively a bucking voltage equal to the current by its resistance).
Of course I might have missed something in your question and I'll just go hide in the corner.

Don Black.

On 15-Jan-14 9:22 PM, keantoken@... wrote:
 

The inductance of the wires may skew that result. If the diode capacitance is small, the transmission line behavior might even be a problem.


Even if we take into account inductance and all that - we can't even be sure of the result because of skin effect and proximity effect which raise loss/resistance at high frequencies.


Perhaps the capacitance is large enough for a reliable measurement. You could theoretically calculate the influence of skin and proximity effect if you were good at math and knew the geometry of the wires.


If you built a simulation model with all known factors, you could tweak the unknown factors until you could reproduce exactly the pulse response of the sensor assembly. I've done this all the time and it is one way for a non-technical person to "cheat" (but also to learn).


A constant current measurement technique would only require an additional offset voltage for calibration between the two diodes.


And if the wire is long enough that its resistance is an issue, then you should really consider whether the wire's own tempco could skew your results.





This email is free from viruses and malware because avast! Antivirus protection is active.


Join TekScopes@groups.io to automatically receive all group messages.