Marcelo Yamamoto wrote:
> 
> Reg wrote:
> 
> >RTD and thermocouple sensors can operate reliably at this temperature. Can
> >you tell us anything about the other requirements - physical size,
> >reliability, cost, volume etc.?
> >
> >Reg Neale
> >
> 
> Ok. I'm looking for a temperature sensor about 15mm length (or less) like a
> PT-100. Cost is my problem. I will purchase about 1k/month. I want (of
> course) a reliable piece but not so accurate as I can adjust its curve by
> software.
> 
> Thanx,
> 
> Marcelo Y.
> m_yamamoto@uol.com.br

There is a misconcept here about adjust all curves by software. I am not
saying you are wrong, in true this is the way to go, I do it too, but
the tricky thing here is not the accuracy, but the "repeatability".  If
the component curve is a pure zig-zag around what it should be but it is
steady and repeats itself all the way, then using software you can turn
a very cheap component into a high accurate and expensive one.  But
normaly low cost and low accuracy components also have poor
repeatability and lots of noise.

An easy way to test the component repeatability, is just take a
component (ambient temperature) imerse it in 0¡C (water and lots of
ice), wait 2 minutes, measure its output, then imerse it in boilling
water (100¡C) during 2 minutes (enjoy and boil two eggs along :), then
take it back to the iced water... wait 3 minutes and measure its output
again... you will see a diference, this delta is the component
repeatability factor...  Do it 10 times, you will see different values.

Hey Marcelo, o Ciro Ramos ganha? :)

Wagner