Hi, I have been experimenting with the Wind sensor Rev P for sometime now.
I noticed that the temperature output (Pin Temp) from the sensor appears to always read higher than ambient, it calculates out at 33.69 deg C (AD (3.3V) for A2 = 328). I have installed a BME280 and the corresponding ambient temperature reading is 23.52 deg C. I am not sure if the Pin Temp is supposed to be showing ambient temperature?
I have not as yet used your temperature compensation formula as outlined in your regression for the Rev P wind sensor notes as I assume this might produce errors if the pin Temp should be outputting ambient temperature. I would appreciate your comments as i cant really find out that much about the wind sensor Rev P.
I note from several websites that the wind sensor REV P is suggested to be direction specific, stating that air flow should be directly at or within +/- 30 deg of the wind sensor hot wire element is this correct? I do seem however to obtain more realistic values for wind speed whilst holding the sensor in to the air flow at 90 deg to the hot wire element.
Wind sensor Rev P Temperature ambient or working?
Re: Wind sensor Rev P Temperature ambient or working?
MendipMac,
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 get the same result, hmm
Does the sensor respond when you heat it up (not over 70C)?
I'm wondering about the voltage regulator on the board. These are put together by a PnP machine but occasionally we do have to fix parts by hand. See if you can some part markings with a magnifying glass and report back, and also maybe check some voltages between ground and the temp sensor leads. It will probably be easier/safer on a flat surface (table).
Other things?
wrong pin
bad wire (you laugh)
bad solder joint on pin
Is this a PI? They have kind of funny ADCs but the error is usually at the high end - I doubt it would affect things that much.
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 get the same result, hmm
Does the sensor respond when you heat it up (not over 70C)?
I'm wondering about the voltage regulator on the board. These are put together by a PnP machine but occasionally we do have to fix parts by hand. See if you can some part markings with a magnifying glass and report back, and also maybe check some voltages between ground and the temp sensor leads. It will probably be easier/safer on a flat surface (table).
Other things?
wrong pin
bad wire (you laugh)
bad solder joint on pin
Is this a PI? They have kind of funny ADCs but the error is usually at the high end - I doubt it would affect things that much.
Re: Wind sensor Rev P Temperature ambient or working?
Hi Paul thanks for the reply, yes I was aware of the formula thanks.
I checked the voltage between temp pin and ground and it compares very favourable with the calculated value see below:-
RT 303 voltsout 0.98 29.56 C I checked between ground and temp pin and got 0.98V
T 487 voltsout 1.57 59.97 C (heated with hair dryer) so yes the temperature does respond.
As for markings specifically which components are you referring to and where are they on the sensor?
The problem was there when I was solderless breadboarding the circuit and developing the sketch and hasn't changed since building the completed project on a veroboard. All pins are correct.
I am using a clone Pro-mini 3.3v which seems to be OK.
I am using a POLOLU-2117-UW Converter to step up the 9v battery voltage to achieve the 12 volts.
I checked the voltage between temp pin and ground and it compares very favourable with the calculated value see below:-
RT 303 voltsout 0.98 29.56 C I checked between ground and temp pin and got 0.98V
T 487 voltsout 1.57 59.97 C (heated with hair dryer) so yes the temperature does respond.
As for markings specifically which components are you referring to and where are they on the sensor?
The problem was there when I was solderless breadboarding the circuit and developing the sketch and hasn't changed since building the completed project on a veroboard. All pins are correct.
I am using a clone Pro-mini 3.3v which seems to be OK.
I am using a POLOLU-2117-UW Converter to step up the 9v battery voltage to achieve the 12 volts.
Re: Wind sensor Rev P Temperature ambient or working?
MendipMac,
I'm not a betting man but I would bet you're going to find that that boost circuit is dragging 3.3V down, so do check that. That is the reference for your ADC so you'll probably need to alter the math. The setup you have is not going to be viable long term, a 9V only provides about 700 mAH and the wind sensor is going to eat it fairly quickly (it draws about 35 mA). Do this instead with your current setup. Wire the 9V straight to the wind sensor - the wind sensor's regulator will be in dropout but the .5 V is unlikely to matter.
For a portable setup it is viable for some hours but you'll need to check when the battery starts to fade. You could use a voltage divider and printout the battery voltage on another analog pin.
With the battery wired straight to your arduino will be free to go on its way (although the board draws around 20 mA or so anyway). If you know how to use your meter on current setting (read about it first) check some currents and report back.
I have a feeling that that is going to fix the temp problem though.
Paul
I'm not a betting man but I would bet you're going to find that that boost circuit is dragging 3.3V down, so do check that. That is the reference for your ADC so you'll probably need to alter the math. The setup you have is not going to be viable long term, a 9V only provides about 700 mAH and the wind sensor is going to eat it fairly quickly (it draws about 35 mA). Do this instead with your current setup. Wire the 9V straight to the wind sensor - the wind sensor's regulator will be in dropout but the .5 V is unlikely to matter.
For a portable setup it is viable for some hours but you'll need to check when the battery starts to fade. You could use a voltage divider and printout the battery voltage on another analog pin.
With the battery wired straight to your arduino will be free to go on its way (although the board draws around 20 mA or so anyway). If you know how to use your meter on current setting (read about it first) check some currents and report back.
I have a feeling that that is going to fix the temp problem though.
Paul
Re: Wind sensor Rev P Temperature ambient or working?
Hi Paul I did as suggested and wired the 9v battery direct to the the wind sensor, no change I am afraid - the voltage between ground and Temp pin was 0.88v (~28deg C). If the boost circuit was dragging down the 3.3V wouldn't the output (wind) values also be an issue?
I was using a 9V battery as I wanted to keep the sensor box small for portable use, and after evaluating it I was going to build it into a 4 x 1.5AA 6Vbattery pack and use the voltage step up to 12 V. I was going to use it in a logger to log wind speeds every hour.
I found I think its IC3 at TP1 got very hot to touch is that normal?
Unfortunately I had an issue whilst trying to measure the current draw (Yes I did know how to connect DVM in series thanks ) but it seems the output pin got shorted with the 9v power supply (trust me to try and do this without using a solder-less board!) I now don't have any voltage from either out or Temp pins. I think I have fried it!
Is it worth sending the sensor back to The JeeLabs for repair or what do you advise?
I was using a 9V battery as I wanted to keep the sensor box small for portable use, and after evaluating it I was going to build it into a 4 x 1.5AA 6Vbattery pack and use the voltage step up to 12 V. I was going to use it in a logger to log wind speeds every hour.
I found I think its IC3 at TP1 got very hot to touch is that normal?
Unfortunately I had an issue whilst trying to measure the current draw (Yes I did know how to connect DVM in series thanks ) but it seems the output pin got shorted with the 9v power supply (trust me to try and do this without using a solder-less board!) I now don't have any voltage from either out or Temp pins. I think I have fried it!
Is it worth sending the sensor back to The JeeLabs for repair or what do you advise?