Arduino + Love TS

[Pash91]

Hello,
I purchased a Love TS from Brewers Hardware and am having issues getting it to work with my Arudino. I'm following the Thermistor examples on the Arduino site; however, I can't seem to get it to read correctly. I used the calculator to get my a, b, and c values, but I can never get it to 1) read correctly, and 2) get a stable reading. I have a feeling I'm having issues with the wiring, so any help would be much appreciated. I've tried

1) white to analog0 and black to ground
2) white to analog0 and shield to ground
3) white to analog0, black to ground, red to 5+
4) white to analog0, shield to ground, red to 5+

and pretty much every permutation of that I can think of. I've also used it with and without a 1K and 10K resistor (changing the values in the program to compensate). I've also played with the thermistor values between 0, 1000, and 10000.

I would much appreciate any insight.

Thanks!

 

[jgalak]

Is there a data sheet for the Love? Any other info on it? I'm new to home brewing and don't now the Love, but I've played a fair bit with Arduinos.

 

[Pash91]

Is there a data sheet for the Love? Any other info on it? I'm new to home brewing and don't now the Love, but I've played a fair bit with Arduinos.

I bought this probe from Brewer's Hardware, so I'm not really sure besides what's on their site. I followed that, but am still having some issues

 

[jgalak]

Ok, looks like the info is probably here: http://www.dwyer-inst.com/Product/Temperature/TemperatureSwitches/DigitalPanelMount/SeriesTS/Specs

The important wires appear to be black and red. Do you have a multimeter? If not, get one you'll need it.

Put it into a mode to measure around 1000 ohm resistance, and connect the meter probes to the black and red wires of the sensor. Don't connect anything else. What resistance do you get? At a temperature of around 25C (77F) it should be around 1000 ohm. Is it? I so, we can proceed from there. If not, we will need more data.

 

[Pash91]

Ok, looks like the info is probably here: http://www.dwyer-inst.com/Product/Temperature/TemperatureSwitches/DigitalPanelMount/SeriesTS/Specs

The important wires appear to be black and red. Do you have a multimeter? If not, get one you'll need it.

Put it into a mode to measure around 1000 ohm resistance, and connect the meter proves to the black and red wires. Don't connect anything else. What resistance do you get? At a temperature of around 25C (77F) it should be around 1000 ohm. Is it? I so, we can proceed from there. If not, we will need more data.

I have done that; however, it's the white and black wires that produce this resistance

 

[jgalak]

And you get around 1k at room temperature? If so, I would follow this tutorial:

http://learn.adafruit.com/thermistor/using-a-thermistor

However, that assumes a 10k thermistor, and this one is a 1k. So when you get to the temperature conversion section, the value for THERMISTORNOMINAL will need to be 1000, not 10,000 like in the example.

You can try both a 1k and 10k series resistor. The 1k should get more accurate results, but with that little resistance, you may get self-heating. I suspect that 10k will work better. Make sure to set the value of SERIESRESISTOR as appropriate.

Note that this program take multiple readings and averages them. That's the only good way to get consistent readings from the system.

This probe can be submerged, right? If so, I'd get some water, heat it to 25C/77F (using a regular thermometer to check), and tweak TEMPERATURENOMINAL until it reads 25C.

Then, lower the temperature by a bunch (I'd just get a glass of water and ice, so you are at 0C/32F), and tweak BCOEFFICIENT until the result is again correct.

If the temperature is fluctuating significantly, but within a smallish range, increase NUMSAMPLES.

Hope this helps.

 

[Pash91]

And you get around 1k at room temperature? If so, I would follow this tutorial:

http://learn.adafruit.com/thermistor/using-a-thermistor

However, that assumes a 10k thermistor, and this one is a 1k. So when you get to the temperature conversion section, the value for THERMISTORNOMINAL will need to be 1000, not 10,000 like in the example.

You can try both a 1k and 10k series resistor. The 1k should get more accurate results, but with that little resistance, you may get self-heating. I suspect that 10k will work better. Make sure to set the value of SERIESRESISTOR as appropriate.

Note that this program take multiple readings and averages them. That's the only good way to get consistent readings from the system.

This probe can be submerged, right? If so, I'd get some water, heat it to 25C/77F (using a regular thermometer to check), and tweak TEMPERATURENOMINAL until it reads 25C.

Then, lower the temperature by a bunch (I'd just get a glass of water and ice, so you are at 0C/32F), and tweak BCOEFFICIENT until the result is again correct.

If the temperature is fluctuating significantly, but within a smallish range, increase NUMSAMPLES.

Hope this helps.

I believe I did essentially this, but I will sit down and try again now that I've had some time.

Thanks

 

[jgalak]

Correction: you should tweak THERMISTORNOMINAL, not TEMPERATURENOMINAL when calibrating at 25C.

 

[Pash91]

Correction: you should tweak THERMISTORNOMINAL, not TEMPERATURENOMINAL when calibrating at 25C.

Ok, came back to it today, still having issues with the thermistor either not reading at all, reading incredibly inaccurately, or reading closer, but not moving.

When trying it with a multimeter, I can see that as I heat it up, the resistance increases and as I cool it, it decreases, so it seems to suggest that the thermistor is actually working.

 

[Pash91]

Scratch that. After playing with the numbers in the code a little bit, it seems to be working better. Thanks for the help

 

[jgalak]

Cool.

Yeah, tweaking constants isn't unusual for this type of thing.

 

[Pash91]

I figured, but it was more than that just to get to that point. Anyways, onto the screen!

 

[Pash91]

Ok, what seems to be happening is that as it's heating, the temperature is dropping and as it's cooling, it's rising. Back to the code