[JT118]

Here are a few graphs and numbers, no real statistical analysis done, I shall do that when I have time. As you can see on the spreadsheet the FO anemometer consistently reads 20% faster than the Inspeed instrument. This is a considerable discrepancy but a fairly consistent difference. Some maths in the software may help to reduce this huge error.

The gust speeds have an even bigger discrepancy, 80 to 180% of Inspeed readings suggesting that the FO instrument registers all sorts of numbers in short bursts.

Working on the assumption that the Inspeed instruments are better calibrated it would appear from these figures that the FO instruments are very unstable, and certainly will not yield any data of scientific value. The rain gauge falls into the same category. At best, it would appear, the FO instruments offer a representation of the weather conditions rather than a moderately accurate set of data.

   
   

FO VS INSPEED_1.xlsx Size: 388,68 KB  Downloads: 148

[werk_ag]

I spotted this difference, and although I have not quantified it, I was surprised because it seemed very significant. This led me to suspect that the wind factor used by me for the FO anemometer could be wrong (I used the value specified in the Argent Systems datasheet for the FO anemometers). Then I remembered a discussion that I read some time ago, in the Cumulus forum, which according to some calculations the real wind factor for FO anemometers should be lower than the one I used, but only 5%, which does not justify this difference of nearly 20%.

I have also checked whether the cause of this difference is not in the readings of the Inspeed Instruments, but I can't find any fault. Both are using the same routines to read the pulses. (obviously the formula to calculate the wind speed and gust, uses a variable derived from the Wind Factor and pulses per revolution, for each one).

[JT118]

I

Fell free to create the new thread .

Hi
Would it be possible to use a wind sensor such as the WSD1 from Environmental Measurements Ltd
This uses a continuously rotating potentiometer which could be connected as a voltage divider circuit.
http://www.emltd.net/products/wind/wind-...on-sensors
Thanks

I spotted this difference, and although I have not quantified it, I was surprised because it seemed very significant. This led me to suspect that the wind factor used by me for the FO anemometer could be wrong (I used the value specified in the Argent Systems datasheet for the FO anemometers). Then I remembered a discussion that I read some time ago, in the Cumulus forum, which according to some calculations the real wind factor for FO anemometers should be lower than the one I used, but only 5%, which does not justify this difference of nearly 20%.

I have also checked whether the cause of this difference is not in the readings of the Inspeed Instruments, but I can't find any fault. Both are using the same routines to read the pulses. (obviously the formula to calculate the wind speed and gust, uses a variable derived from the Wind Factor and pulses per revolution, for each one).

I suppose the main problem with the FO instruments is that they are manufactured to a price rather than to a quality. I suspect that if we set up a scientific experiment with 7 or more replicates, we would find considerable variability between the 7 FO instruments. It would be interesting to alter the software maths by 18% and see what the results yield.

[werk_ag]

I suspect that if we set up a scientific experiment with 7 or more replicates, we would find considerable variability between the 7 FO instruments.

I'm not so sure about that. The FO anemometers are simple devices, with just a reed switch inside, there isn't many things to cause they perform differently among them.

It would be interesting to alter the software maths by 18% and see what the results yield.

Yes it would be. When I can, I will install my FO anemometer again and let it run for a week or so.

[AllyCat]

Hi,

Yes, I also found the above comparison surprising. The majority of users (including me) have suspected that the anemometer actually under-reads. However, the "anemometer factor" (calculated from the physical dimensions) is very much at the lower end of the expected range, so indeed it might be expected to over-read.

But most users (including me), particularly of FO stations, are not able to locate their wind sensors "correctly" (i.e. 10 metres above ground with no nearby obstructions), so perhaps FO have built in an "error" to give more realilistic values from a typical installation ?

Cheers, Alan.

[JT118]

I have a collection of anemometers,
3 x Didcot
3 X Davis
3 x NRG #40C
and 5 X FO/Maplin. for UK residents these are very cheap

http://www.ebay.co.uk/itm/Pack-of-5-Mapl...1ea1e4c2f7

£7.99 for a pack of five! Type "wind sensors" into ebay.co.uk

Unfortunately I do not have the time nor the money to set these up to conduct a proper scientific experiment. Measuring wind speed is notoriously difficult and one would have to find a calibrated gauge to run as a base value to compare the others against. My NRG's were calibrated but some time ago and the calibration certificate expired about 5 years ago and it outputs a sine wave making it difficult to use on Weatherduino. See attached.

NRG #40C Calibration Cert..pdf Size: 273,06 KB  Downloads: 229

Werk suggested a Schmitt Trigger and I am looking into this but it is a bit beyond my current electronics expertise.

I accept that the FO instruments are Simple but there are many other variables to consider, bearing drag, cup accuracy, and many others, the trouble with measuring tiny parts of huge things is that tiny errors quickly multiply. See the bearing arrangement for the NRG sensor.

http://www.windandsun.co.uk/media/277060...-sheet.pdf

[uncle_bob]

5 X FO/Maplin. for UK residents these are very cheap

Wow, that is cheap. Shame they don't ship further afield.

[werk_ag]

Yes, really cheap. I bought them for around 15€ each (anemometer and vane) from a store in Spain.

[johng]

RE: Wind sensor alternatives to FineOffset. Any ideas?
Is there any interest in support for Lacrosse wind sensors (I happen to have some). I notice that John Geek from New Zealand has fully reverse engineered the interface and protocol.
https://www.john.geek.nz/2011/07/la-cros...-protocol/

[werk_ag]

RE: Wind sensor alternatives to FineOffset. Any ideas?
Is there any interest in support for Lacrosse wind sensors (I happen to have some). I notice that John Geek from New Zealand has fully reverse engineered the interface and protocol.
https://www.john.geek.nz/2011/07/la-cros...-protocol/

Adding support for devices which have its own RF transmitter system it's difficult, each one use it's own frequency, or modulation mode, which may not be compatible with WeatherDuino Pro2 receiver.