Wind Sensor Calibration and the Wind Tunnel

We’ve been selling a little wind sensor, which has found its way into lots of hobbyist projects and even some “harder” science projects including one by the US EPA, but we’ve never had the kind of solid data that would make the sensor really useful for a range of people. We decided that some hard data was long overdue.

At the top of this post you see the plot of wind speed vs sensor output voltage for two different temperatures. Without too much consideration you can see that the output of the sensor is temperature sensitive and that the curve is fairly flat as the wind speed increases. Both of those facts mean that it is fairly easy to confuse a shift in temperature with a shift in wind speed. Or that recalibration will be required when the temperature changes greatly.

Obviously this greatly limits the utility of the sensor, so to remedy this we have a number of ideas in mind, including an Arduino Sketch that uses the temperature output on the wind sensor for calibration and several new schematics and designs for wind sensors.

Before we could do any of that however we had to have a tool to verify wind speed, but not just any wind speed, we needed to be able to precisely control temperature along with wind speed. Building the wind tunnel has been a lot of fun and I’ll cover its construction in several blog posts. I’ll also be posting more data about the wind sensor, as I type up my first handwritten data journals (that look something the image below).

You may wonder about the manual data entry effort. I am worrying about it too, and wondering why I’m not automating all the logging procedures to save the data entry. This is really just the first flush of  excitement at having a research instrument that can provide good data – so there is lots of room for improvement in my technique. The anemometers that I’m calibrating against don’t have any connection to the computer, so if I automate the procedure I will still need to read the anemometers manually and get that column of  numbers in by hand, and also synch it with the other data. I think some kind of semi-automated system still makes sense though. Many experiments can also be done without actually entering any of the data, or even saving much data. More work to do.

So you probably want to see what a DIY research table-top wind tunnel looks like, and I’ll oblige you with a teaser photo just now. Many more details will be provided in the next week, as to the design choices that went into its construction. Yes it does still look like a table full of wires, that aspect kind of comes with designing things as you figure out how to do them. As I decide which parts of the apparatus seem good enough for  my purposes, I’ll harden them up.

Back to blog