Peter McNulty wrote:
> Hi,
>
> What temperature sensors would you recommend for:
>
> -Fast Response Time (90% of actual temp)
> -Relatively Cheap (This is a school project, so nothing that costs
> $100)
> -EASY to program for using PIC16F877 - This is a must! Themistors are
> looking to be very difficult to program for, for a nice accuracy
> -Accuracy of sub +- 1degreeC
>
> I've looked at the LM335 solution, but the response time is very long
> compared to other things, and so is the maxim chip. I've also messed
> around with thermistors, just printing out an ADC value and they seem
> to vary the most, but they are hard to program for as they are
> non-linear, and they change over age. RTD's look good, but i can't
> get a definate price on these things yet, and they are linear with a
> fast response time. Thermocouples I haven't really thought about,
> though i don't know if  they have thermocouples for the range of -50C
> to 100C with a nice accuracy, fast response time and easy
> programming. Are thermocouples linear?
>
> Are there any other solutions i haven't thought of? What are your
> opinions of solution i need?
>
> Thanks
> -Peter


Peter, linearity can only be found by paying $, on Maxim or Dallas chips.

Any other sensor needs calibration and linearization, not always by the
sensors (case of the PT100), but mostly by your circuit out of precision.

If you think thermistors are difficult to deal with, just forget about
thermocouples. They give you millivolts (and microvolts) in a total non
linear range.

At this +-1=B0C accuracy you will need to go for calibration, no other wa=
y.

Calibration means you use any kind of sensor, but then needs to calibrate
individually each unit based on fixed temperature standards, ice point
(0=B0C), water boiling (100=B0C) and intermediate temperature points, the=
n your
device software will interpolate the non calibrated mid points.  It means=
,
as much points you calibrate, better accuracy you get.  Except of course =
if
you want to start to play with polynomial math to do linearization withou=
t
table and interpolation.

It is NOT an easy task.

Very cheap sensor?  try to measure the Voltage Drop across a 1N4148 small
glass diode and change temperature...
A 10k NTC also is a nice unit, but none will give you 1=B0C accuracy for
free.

VV46NER

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listserv@mitvma.mit.edu with SET PICList DIGEST in the body