Neptune Apex Temp Probe used for air temp? Anyone?

[halmus]

I'll skip all the extra details, but I have an application where I would like to use a temp probe attached to the Apex to monitor air temp. Has anyone out there done this, or heard of it being done?

From my experience, the temp probe has a pretty high range. I have one on my fresh-water reservoir in the attic with is running about 60 degrees right now (with a heater and circulation pump) I want the temp probe to monitor the air temp in the section of my attic where the water lines run. I have them insulated, but I want to monitor them to make sure things don't freeze this winter. That means, I want the probe to monitor an air temp range of 30-70 degrees. A little crazy right?

For the tech geeks out there, I know I can engineer a little circuit with a thermistor which would communicate with the I/O switches. I have a few free, I just don't have time to put the circuit together. I wish I did. Sounds like a fun project.

 

[halmus]

So, I went ahead and ordered the add-on's needed to try to implement the project. Turns out, like I should have guessed, that using the Apex temp probe for air-temp measurements (especially in the range I was targeting) is completely inappropriate. I took a number of readings with the probe just hanging out the window alongside a dedicated digital thermometer that is intended for outdoor use. I did this over a period of a few days where the temps were 30-40 deg F, so the probe had plenty of time to acclimate. I recorded the actual readings I was getting from the Apex and plotted them (in blue data points) against the "Ideal" case (in green). I did a quick and dirty linear regression on the data points I recorded for the Apex readings. You can see it fits that for the small range I recorded. A higher degree regression (Quadratic) would fit better if I recorded a full set up into the range that it is intended to measure (70-85 deg F) as it's plotting would leave the projected (in red) line I graphed as it approached the intended operating range. Still, it would be less than reliable. It's intended to be submerged.

Bottom line, for anyone finding this on a search engine, the answer to my initial question I couldn't find anywhere else is NO. Don't try to use the Apex temp probe to track ambient temps especially down at 30-45 deg F.

 

[eartaker]

Odd, you would think the readings would be the same.

 

[halmus]

I copied this from the site:

http://www.ilxlightwave.com

"The price of the thermistor’s high sensitivity is paid with non-linearity, a factor which makes their selection and use a bit more challenging than might be otherwise expected. Thermistors are generally two-terminal semiconductor devices that have an electrical resistance that varies non-linearly with temperature. The non-linearity of thermistors complicates their calibration and use."

If you check out the Resistance-Temperature response curves on that site, you'll see that the curve is not linear at all. The compensation, I guess, comes from the software. I'm guessing to get an output response curve from the "black box" full of hardware, you're going to need some sort of reference chart, a matrix, or a conversion formula:

Steinhart-Hart equation:
1/T = A + B*(Ln R) + C*(Ln R)3

Which they list on that site. The variables and constants are listed on the site. They say it leaves only +-0.1°C over the range -30°C to +125°C. The Apex probably just didn't need to take that wide range into consideration thinking that the target range would be very small.

 

[eartaker]

Ahh I guess that would make since. You can order a cheap enclose thermocouple that would work for what you want but if you already have something on hand you could just go ahead a build it. If you need help with the soldering or circuit design I can help you out.