circuit for RTD AND Thermistor...

Thread Starter

Gibson486

Joined Jul 20, 2012
331
I am trying to make a circuit that will take an RTD and a thermistor (not at the same time). I was thinking simple and using a simple divider into an op amp. Switching out out resistor values is not a problem. For the thermistor, I would use an op amp as a buffer, but when I switch to an rtd, i would change it to a non inverting op amp. It sounded feasible until i saw how everyone uses an active current source for an RTD usually. So...I thought to use a current source for both, but small current through the huge resistance of a thermistor does not sound right.... Anyone every try such a thing?
 

Thread Starter

Gibson486

Joined Jul 20, 2012
331
No, because RTDs and thermistors have such drastically different characteristics that designing a circuit to handle both would be a titanic PITA-- especially if decent accuracy is needed.

Not worth the trouble.
Yeah, seems like RTD are best with current sources and thermistors are best with a divider type topology. You can use the RTD with the divider/bridge...so i am kind of going towards that....
 

Thread Starter

Gibson486

Joined Jul 20, 2012
331
Take a look at this http://www.analog.com/en/products/ltc2983.html#product-overview chip or it's brother. There is an Arduino (Linduino) development kit available.

They don't really show you how to "auto-range" manually for an unsupported type. One of the chips has a EEPROM for a look-up table. Maxim has a solution too.

Remember that RTD's can sometimes use 3-wire rather than 4.
I have used that chip before. Unfortunately, it is too big for this project. I decided to use a differential amplifier as the front end. When I want to switch, I am just gonna switch resistors and jumpers. Essentially, we want to compare thermistors and rtds for use in an upcoming porject. I am thermistor guy, but the other is an rtd guy.
 
You should easily be able to make a choice. I used an IEEE-488 interfaceable RTD thermometer from -80 to 200C. Could not beat it. Basically wanted to measure the surface temperature of a 1" x 1" x .025" piece of glass in an environmental chamber.

So, I had the environmental chamber controller which I used the proportional band of the oven controller to control the surface temperature. Turned out to be really cool.

For thermocouples, you use K when you can. T for room temperature or cryogenics. T isn't the most appropriate for cryo, but it isn't bad. A diode is best for that.

You have semiconductor sensors too.

The RTD was pretty rugged. I made a surface prober and I think I only replaced the RTD once in like 10 years.

We did have an application for type R and even type C.

You can get the mass of thermocouples very low, but they become throw-away TC's. We sometimes made our own junctions. For the platinum types, a Hydrogen/Oxygen torch. For K, a carbon block or a spot welder.

Even did IR with a video camera once.
 
@Gobson486: I've been thinking about a possible project for a long time which is somewhat difficult.

It could easily be described a "thermister type changer" where you would have one thermister type in and another thermister curve out.

The final description is a little different. I want to use the same thermister, but in different ranges. Something like 100F is seen my the controller as 200F in the setpoint.

it' something like 115 to 215F is mapped to 15 to 115F. Not sure of the upper limit though, The lower limit is 115F for sure. The upper limit is at east 180F. I want 75 to 115 to be able to be achieved temporarily.

The application is a thermister probe for a microwave oven. You need about 100F to proof yeast and why can;t you heat a refigerated liquid to room temperature.

I have characterized the thermister probe and I GOT LUCKY and was able to find the same curve in another probe thats still being sold. The phone plug though is not standard on the old one.

That's the motivation anyway.

Any ideas how to do your own auto-range?
 
Top