Temperature accuracy with auxiliary close-up lens

[JohnG]

Folks,

I have a question about the use of auxiliary lenses for close-up temperature measurements. The application is the measurement of die temperatures of GaN power FETs, and the smallest transistors of interest are about 0.9 mm x 0.9 mm. In the ideal case, I'd want at least 9 pixels on the FET.

If I use an auxiliary lens to get close-up measurement capability, how might this affect the accuracy of the temperature measurement, i.e. how much might the absolute temperature accuracy be degraded?

Thanks,
John

[Bill W]

The simple answer is quite a bit.  However you can easily 'recalibrate' the camera and generate a lookup table from 'camera reading' back to 'real reading'.

Ambient will still be ambient.
You will have changed the 'gain' and cameras are linear in thermal radiance difference - not temperature difference.  At a guess you are looking at over 100°C so this begins to matter.  A bit of maths (there's a couple of spreadsheets somewhere on the forum already) will sort this out.

So a few measurements of known hot temperature 'black' bodies through your new close-up optics set will give you the scaling.  You may get some centre to edge shading, so stick to calibrating the centre.

regards
Bill

[JohnG]

Thanks, Bill.

Looks like it is doable, but I will have to review the math. I have to do it for lenses, anyway...

John