Cumulus Support

Today I was watching the anemometer rotating away and was pondering, "How many times have you rotated" thinking of bearing wear and subsequent friction due to that wear and then creating light wind errors.

Then I thought, I'm going to ask the question.

I know the wind run because cumulus tells me. All I need to know is the value (Distance) of run of wind for 1 revolution of the anemometer (For Fine Offset). I Thought of pi*d for circumference but where do I take the diameter? Ummmm!

Gina? I know you will have the answer.

I would figure from the centre of the cup to the cente of the anemeter would be the Radius (and not use the outer measurement of the cup).

I was even thinking (as initially, my rain gauge was 9% under), is the speed reading correct? Might jump in the car with the anemeter a good 1/2 meter out the window (out of the wind stream made by the car), and do a couple of runs at steady speed in oposite directions, and using combination of speedo and GPS to work out if it is reading accurately. Just a thought

That is certainly one recommended way of calibrating an anemometer. Take a pal to read the instruments and concentrate on the driving (or the other way round). A wind-less day is best. Nowhere round here that would be safe enough for that. You have vast empty spaces, haven't you? Good luck.

You can work out the rotation speed of the anemometer as follows:

The resolution of the measurement is 0.34 m/s i.e the reported wind speed is always a multiple of 0.34 m/s. This correponds to one count within the 2 s measuring period of the microprocessor. I had initially thought that the anemometer gave one count per rev but I now realise that it actually gives two - thus 1 count in 2s is actually 1/2 a revolution in 2 s which is 0.25 rev/s.Thus 1 m/s windspeed gives 1/0.34 x 0.25 = 0.74 rev / s. That should allow you to work out how many times the anemometer goes round in a year. If you assume an average wind speed of 3 m/s (about 7 mph) it works out to about 70 million times!

Incidently the cup speed (at the centre of the cup) is about 3 times slower than wind speed.

brassing wrote:You can work out the rotation speed of the anemometer as follows:

The resolution of the measurement is 0.34 m/s i.e the reported wind speed is always a multiple of 0.34 m/s. This correponds to one count within the 2 s measuring period of the microprocessor. I had initially thought that the anemometer gave one count per rev but I now realise that it actually gives two - thus 1 count in 2s is actually 1/2 a revolution in 2 s which is 0.25 rev/s.Thus 1 m/s windspeed gives 1/0.34 x 0.25 = 0.74 rev / s. That should allow you to work out how many times the anemometer goes round in a year. If you assume an average wind speed of 3 m/s (about 7 mph) it works out to about 70 million times!

Incidently the cup speed (at the centre of the cup) is about 3 times slower than wind speed.

Not to be contrary, but how did you arrive at the conclusion that there are 2 pulses per revolution? If you open the sensor you will clearly see 1 magnet and 1 reed switch in an arrangement that gives 1 pulse per revolution. (Go to the second sticky thread at the top of this group)

Where does the "2s measuring interval" information come from?

Also, do you have any math to support the ratio of cup speed to wind speed? Several of us have been searching for the formula for a few weeks now. This would be a useful number if it had precision to a couple decimal places, but "about 3" is like saying days in the week are "about 10" - correct to 1 digit or precision, but not too useful.

Thread originated here: https://cumulus.hosiene.co.uk/viewtopic.php?f=13&t=4138
(Could do with being combined)

Number of 'pulses' required to register a minimal reading - Gina promised to investigate

I can confirm that the anemometer gives one pulse per revolution. It's quite a wide pulse as can be seen in my waveform diagrams in another thread. I haven't yet measured the exact relationship of revs/sec to wind speed but I'm planning some tests this afternoon and I'll probably measure it then. As I recall from previous measurements, when the anemometer was rotating at around 9 seconds per rev, the system showed about 0.3 m/s gust speed and zero average. It had to go up to more than one rev per 6 seconds to show 1 unit of 0.3 m/s average speed. The average speed is always less than the gust speed even with a constant air flow (except when the gust is zero, of course).

Charlie wrote:

brassing wrote:You can work out the rotation speed of the anemometer as follows:

The resolution of the measurement is 0.34 m/s i.e the reported wind speed is always a multiple of 0.34 m/s. This correponds to one count within the 2 s measuring period of the microprocessor. I had initially thought that the anemometer gave one count per rev but I now realise that it actually gives two - thus 1 count in 2s is actually 1/2 a revolution in 2 s which is 0.25 rev/s.Thus 1 m/s windspeed gives 1/0.34 x 0.25 = 0.74 rev / s. That should allow you to work out how many times the anemometer goes round in a year. If you assume an average wind speed of 3 m/s (about 7 mph) it works out to about 70 million times!

Incidently the cup speed (at the centre of the cup) is about 3 times slower than wind speed.

Not to be contrary, but how did you arrive at the conclusion that there are 2 pulses per revolution? If you open the sensor you will clearly see 1 magnet and 1 reed switch in an arrangement that gives 1 pulse per revolution. (Go to the second sticky thread at the top of this group)

Where does the "2s measuring interval" information come from?

Also, do you have any math to support the ratio of cup speed to wind speed? Several of us have been searching for the formula for a few weeks now. This would be a useful number if it had precision to a couple decimal places, but "about 3" is like saying days in the week are "about 10" - correct to 1 digit or precision, but not too useful.

The 2 pulses per rev comes from a post by DanielF on the Fine Offset anemometer repair topic of the homebuilt forum. He posts a pdf file which shows photos of the inside of the sensor. It shows the reed switch horizontal and centrally placed. The magnet in the rotor rotates round it. My guess is that the magnet operates the switch when it is on a line at right angles to the length of the switch which occurs twice per rev, but I don't actually know how reed switches work or what direction of magnetic feild is needed to operate them. Anyway DanielF is of the opinion the switch operates twice per rev and I think he is right because it solves a problem with the anemometer factor (the ratio of wind speed to cup speed) being too low otherwise. See the thread on Fine Offset anemometer factor.

The 2 s measurement interval comes from the signatrol website (http://www.signatrol.com/product/35_wea ... cification). I emailed them and they say the info comes directly from Fine Offset.

The factor of "about 3" is the anemometer factor. This is discussed on the Fine Offset anemometer factor thread mentioned above. If there are 2 pulses per rev it comes out as 3.12 for the Fine Offset anemometer. I have a data sheet for a Vector Instruments R30 rotor. This gives a calibration which leads to an anemometer factor of 3.77 which is in the same ballpark as the 3.12 for the Fine Offset. More details in my post on the Anemometer calculations thread on the Homebuilt forum.

The anemometer factor clearly depends on the design of the cups and the geometry of the rotor so you can't give a value which applies to all anemometers. Wikipedia states the anemometer factor varies between 2 and just over 3, but there is no citation so I don't place too much reliance on that info.

Gina wrote:I can confirm that the anemometer gives one pulse per revolution. It's quite a wide pulse as can be seen in my waveform diagrams in another thread. I haven't yet measured the exact relationship of revs/sec to wind speed but I'm planning some tests this afternoon and I'll probably measure it then. As I recall from previous measurements, when the anemometer was rotating at around 9 seconds per rev, the system showed about 0.3 m/s gust speed and zero average. It had to go up to more than one rev per 6 seconds to show 1 unit of 0.3 m/s average speed. The average speed is always less than the gust speed even with a constant air flow (except when the gust is zero, of course).

I hadn't seen this when I made my last post. If you have actually measured the number of pulses per rev that supersedes all the guess work. I assume you manually rotated the anemometer over one rev and observed the contact closures?

If it is 1 count per rev that would appear to put the anemometer factor back to 1.56 which I find difficult to believe - why should it be over a factor of 2 different to the Vector Instruments rotor which is very similar?

I am just wondering if the Fine Offset electronics triggers on the leading and trailing edge so giving 2 counts per rev. That would take the anemometer factor back to 3.12 which I find much easier to believe.



Post a reply

beteljuice wrote:Thread originated here: https://cumulus.hosiene.co.uk/viewtopic.php?f=13&t=4138
(Could do with being combined)

Number of 'pulses' required to register a minimal reading - Gina promised to investigate

I agree they should be combined. How is that done?

Topic locked -- please continue this discussion in this thread....https://cumulus.hosiene.co.uk/viewtopic.php?f=13&t=4138