Thermistor thermometer

[bob91343]

I have an old Taylor indoor/outdoor electronic thermometer.  The outdoor sensor has opened and I would like to replace it.  I asked Taylor and they said the model is so old they have no information or parts.

I put an adjustable resistor in its place and the thermometer reads 25C with 4000 Ohms.  A few other points are 46.6C at 2000 Ohms, and -17.5C at 20k Ohms.

Does anyone have an idea where I can buy a thermistor like this?

[floobydust]

I would redo the measurements, it's an oddball thermistor 4k at 25°C and usually the beta is much higher.
3k, 4.7k and 5k at 25°C are available, and my math your beta is around 2900-3000 whereas 3100 is the lowest you usually see.
Sometime AC hum pickup or extra cable capacitance can give wrong readings. These thermistors are usually Cantherm MF52 family or in glass MF58, from china or at Digi-Key.
Common is 10k at 25°C and beta 3950.

I have some old Taylor Thermor transmitters, they die if you connect them to a bench power supply at 3V, the IC latches up unless you add 47R resistor to lower dV/dt.
Most of the Taylors Thermors I had died, either they drifted off frequency when cold or IC shorted when I ran them on the bench.
So far LaCrosse temperature transmitters are absolute garbage, what a waste of money.
Oregon Scientific/Radio Shack is very good but hard to find in town.

[bob91343]

Well I used an old but still accurate GR decade resistance box to run a curve.  Here are some more data points:

Resistance - degrees C
2.5k             39.8
3k                33.5
4k                25.0
5k                18.4
6k                13.2
7k                  9.0
8k                  5.4
9k                  2.2
2k                46.8
1.81k            50.0

[floobydust]

The 4kohms at 25°C is so odd and a resistance change Rt/25 at -17.5°C of 5x which is not much at all, beta low under 3100. So both values are oddball.
I wonder if the stock sensor had a thermistor and resistor in parallel, for calibration or open sensor detection? Is it just an epoxy bead?

Amphenol material type A2 or A3 (page27) or Vishay Curve 14 match what you have.
But finding the part is another load of work.

[bob91343]

Good questions but the original failed sensor is gone.  I had replaced it with a  5k unit that gave me wrong readings but I never thought to try a parallel resistor.  However that would really mess up low temperature readings.

I suppose one could do some math to find out what parallel (or series) combination would give the values I measured.  I have more data but didn't try to type all of it in.

The unit still measures local temperature fine, so I could see what that thermistor is.  Or remove it and use it as the outside unit.

Meanwhile I bought a Chinese unit that only measures outside temp and only has C calibration.  It basically does what I want but still I would like to solve the mystery of this Taylor unit.

[james_s]

I used to have some of those inexpensive indoor/outdoor thermometers and I've had a bunch of different weather stations. I finally splurged and bought a Davis Vantage Pro and soon wished I had just got one of those in the first place. It was expensive but looking back I spent more on cheap ones that didn't last.

I also have a bunch of zigbee temperature/humidity senders that I have in various places around my house and outside. They work great and in the event that one fails it's under 10 bucks for a new one. Moteino is an option if you want to try rolling your own, probably less effort than trying to replace that oddball thermistor. I totally get the desire to fix something though, really drives me nuts when I know exactly what is wrong with something but I can't get the part I need.

[floobydust]

I did some math and most likely your original sensor is a thermistor with resistor in parallel. If you can give some cold readings with >20kohms, it would help determine the resistor value. If you can get -40°C or so to display, should be able to figure it out. Some systems just clip below -40°C so they stop reading any lower even if it is colder, or display a sensor error message.

I have a really good Oregon Scientific outdoor thermometer, TX uses a single 1.5V AA battery and it's the only thing that reads a cold Canadian winter in the -30°C's (with a lithium battery). But the 433MHz range is poor, it really needs 3V. When I played with ZigBee it uses too much power.

Just bought a LaCrosse TX141 thermometer and what a POS it reads up to 10° high in cold weather. It stupidly uses a 50kohm/25°C thermistor which is almost 1MEG in cold weather! LaCrosse replaced the temp transmitter under warranty... with an older PC board, and it was only 8° out in cold weather. Total junk, the firmware is not good. Bin it.

[bob91343]

58.7k gives -40C reading.
50.0k gives -37.1C.

[floobydust]

Closest match to your readings is a Semitec 402ET 4k@25°C B=3100. You can look at the datasheet table. $2.25 at Mouser 954-402ET-1. I would try one. What model is your Taylor TX?
Semitec AT line up is the same as Radio Shack 271-110 (Semitec 103AT-2, 10k@25°C B25/85=3435), good quality part 1% accuracy.

I tried a bunch of other thermistors in the market including parallel resistor- but your sensor is some low sensitivity, beta is under 3100 which rules out everybody.

[bob91343]

Taylor model 5367.