## Wind sensor REV P - adding compensation

Questions about Modern Device and JeeLabs Sensors
yoc1983
Posts: 4
Joined: Wed Aug 26, 2020 6:29 am

### Wind sensor REV P - adding compensation

Hi,
so i am trying to calculate the wind speed from the equation derived from the regression
(ie with temperature compensation)

i don't get good values.. always something very very small.

exapmle:
Volts=2.444
ZeroWind_V=1.446
*is that an OK value (measured for 28[C]? , i tried also 1.3692 as noted but still no good
Temp_C=32.875
i get for WS_MPH very very small number

from the equation without compensation i get 30.33[MPH]
do i need to do something else?
paul
Posts: 775
Joined: Mon May 12, 2008 4:19 pm

### Re: Wind sensor REV P - adding compensation

For the zero wind value - use a volt meter and measure the voltage with glass over the sensor, but don't lay the sensor flat against a table, angle it upward so sensor is in air on all sides.

Measure the voltage from the temp sensor and run it through the equation

VOUT = TC * TA + V0°C

The Rev P uses MCP9701A temp sensor
Where:
TA = Ambient Temperature
VOUT = Sensor Output Voltage
V0°C = Sensor Output Voltage at 0°C
+ 400 mV (0.400)
TC = Temperature Coefficient
19.5 mV/°C (0.0195)

I've got some crisis to address in shop for the next few days but will try to see if that algorithm in the blog stacks up against reality. It could be my math too.
yoc1983
Posts: 4
Joined: Wed Aug 26, 2020 6:29 am

### Re: Wind sensor REV P - adding compensation

yeah that's exactly what i did.
but then when taking those values and use them
to calc the wind speed with the temperature compensation what do i do?

as i understand i need to put them in the regression equation:
WS_MPH = (((Volts – ZeroWind_V) / (3.038517 * (Temp_C ^ 0.115157 ))) / 0.087288 ) ^ 3.009364
but when doing this i get from this equation really small value and not a speed.

crisis, i remember it we had it back then in the arcade.. nice game
whenever you can
paul
Posts: 775
Joined: Mon May 12, 2008 4:19 pm

### Re: Wind sensor REV P - adding compensation

yoc1983

I just ran some dummy numbers through that expression and it can't possibly be correct. For the time, you have to wait for better data. If I can find that regression I'll try and rework to find where I went wrong.

Usually when I get done with a new regression, I run it backwards with the input data to see how closely it approximates the measured speed. Somewhere along the line I didn't so that - or made a mistake in translating my exel sheet to the expression.

We have a new temp controlled wind tunnel over the last year and I need to combine the controls for temperature, wind speed and sensor testing, which all live on their own arduinos now, plus a pitot tube, so that the wind tunnel can work unsupervised to produce data.

It's in the works but I'm also running the shop with little labor from covid and workers heading back to college.
yoc1983
Posts: 4
Joined: Wed Aug 26, 2020 6:29 am

### Re: Wind sensor REV P - adding compensation

no worry friend.
where can i check for updates? when the new results will be ready,

i am using the sensor with the equation that comes in the Arduino code
when comparing to my anemometer the result are not equal but close.

generally i like this sensor it got some nice advantages:
- small can get in small places/ducts/holes
- working on heat and can measure very close to fan
compare to anemometer (reading are inflicted by EMF)
paul