Easy way to mesure temperature at 20 m+ , 0.1 °C accuracy. Thanks to Brian and

inservi

Hello,

I trying to mesure water temperature from my piscine from some years. The best result i get was with LM35 or AD592 but these ICs give only ~0.5°C acuracy and special cabling.
Now, i want to try with PT100 or the DS2760 kit. The problem is that i need a good trustworthy with about 70 foots of cable. Is it a 3x2.5 cables between the skeemer and the technical room. I dont like a radio comunication because it is not reliable. I already make a complete controle box for filtration and heating with a Microchip 16F252 but the mesure of water temperature was not reliable enough. Now i want restart all with the Propeller for fun.

Thanks an sorry for my poor English.

dro

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
in medio virtus

Post Edited (inservi) : 1/28/2007 9:00:46 AM GMT

Brian Riley

Dallas Semi One-Wire chips would be the answer for the 20m+ run. Specificially the DS18B20 offers 12 bit accuracy and can easily do .1C. There is already a Spin Object and routines that would cover most of what you would need. The only thing you might have to do is the specifics of turning the 12 bit value into a number you can conveniently worth with, which should not be very difficult. The data sheet is readily available.


I just purchased a large quantity for my business. If you are interested would be happy to let you have a couple for my cost plus postage. Contact me via private message.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
cheers ... brian riley, n1bq, underhill center, vermont
See the K107 Serial LCD Controller at
www.wulfden.org/k107/

Tracy Allen

The LM35 or the AD592 can be very accurate, especially with a one or two point calibration. I am curious, what is the issue with the accuracy? Is there noise on the cable due to the long run and interference sources? Analog to digital converter limitations?

Accuracy is different from resolution or precision. A good quality PT100 is capable of that kind of accuracy, but you have to make the interface circuit, and that is going to have to be calibrated. And if there are issues of transmitting analog data at this site, the signal from the PT100 will face the same issues.

I would NOT recommend a thermocouple (DS2760) if you require absolute accuracy. Thermocouples have many nice features, but accuracy is not one of them.

Direct to digital sensors like the DS18B20 that Brian recommended should work fine. It gives plenty of resolution, but will require calibration to get the 0.1°C accuracy. One-wire protocol has issues with long cables, and it is necessary to use a relatively low value of pullup resistor to drive the cable capacitance.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

inservi

Hello Brian, Hello Tracy,

Thank you for your answer. I am very glad to read your suggestions. The DS18B20 look to be the best for me. I did not hope to receive a so simple solution.
I precisely bought DS18B20 this week but I did not believed in it too much. I am glad to learn that One-Wire work on long distance.

In two days, you have solved a problem that I track for several year.

Brian, thank you for the offer but I already have 10 DS18B20.

I am in a hurry to test that.

Live the forum. I hope to also be able to bring of the assistance to community.

Best regards,
dro

(sorry for my poor English)

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
in medio virtus

inservi

Hello,

I tried the DS18B20 and it is exactly what I sought.

Thank-you,

dro.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
in medio virtus