Linear, Low-Cost, 0-250F Temperature Solution?

You'd be best to use a platinum RTD probe with a signal conditioner and possibly an approved intrinsic safety block, depending on the class of the environment and consequences of failure. If it's just one or two, just buy the bits off the shelf and be done with it.

You could also use a thermocouple, but it would be unlikely you'd get system accuracy as good as +/-°C, however the response speed would be much faster.

I'd also add that his specifications and "low cost" are pretty much mutually exclusive.

I found an LM34 with a range of -50F to 300F and a linear 0-5V output for $14 in a TO-92 3-wire package. Would it be possible to insert this item into a thermowell with some epoxy?

Less than $100 is the goal. I definitely don't want to spend $300 an RTD and transmitter.

If it's actually going into a potentially explosive environment, be prepared to spend at least that much for a hazardous area certified transmitter.

--Gene

Good advice, that. A Pt100 RTD probe has to be the best fit for that spec. With a constant current source [1], it's easy to linearise, and there's no cold junction. 3-wire and 4-wire compensation is a piece of cake...

Steve

--

I don't know if it is a piece of cake, at least it would be difficult for me to develop it myself :-) , but I have managed to implement a 3-wire measurement with the AD7793. If you need to convert it to a digital signal, this works very good:

See page 29 for an example. Absolute accuracy is about 0.1°C with a good reference resistor and a good PT100 sensor. For +/-1°C the less expensive AD7792 should work and a less expensive reference resistor.

What about a K-type thermocouple in conjunction with a MAX6675 ? The MAX6675 takes care of all the analogue stuff, and gives out a digital reading through an SPI bus.

Regards Anton Erasmus