Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5

Anemometer Query
#11

Hi,

Sorry me again,

I think that have managed to work out the circumferance of this anemometer.

Center of cup =  413.12mm

Outer edge of cup  =  565.5mm
Reply
#12

Hi,

Sorry, that doesn't help very much, an exact figure for "wind run per revolution (or per pulse)" is essential.

That involves the "Anemometer Factor", for example from Wikipedia:

"... the ratio of the speed of the wind and that of the cups, the anemometer factor, depends on the dimensions of the cups and arms, and may have a value between two and a little over three."

In practice, the only method is to calibrate the anemometer in a Wind Tunnel and/or beside a known precision reference grade anemometer.  A "poor man's solution" might be to attach it to a moving vehicle with a GPS, in still air, but that is a very difficult measurement to make accurately.

However, that particular design does look quite "familiar", so perhaps somebody can identify the Original Manufacturer or some test data?

Cheers,  Alan.
Reply
#13

(18-11-2017, 04:40)tyntop Wrote:  Hi,

Werk_AG can you look at attachment in my last post, as there is a section in it "calculation formula " would that work out the wind run that you are looking for, like i said

Pulse type (high level 5V, low level 0V):
W = 0; (F = 0)
W = 0.1 + 0.0875 × F (F ≠ 0)
(W: wind speed indication m / s; F: pulse signal frequency)


Based on this formula, if the unit produces 16 clicks per revolution, it would be 3m/s per revolution (seems quite high).
I'm not sure about my deduction, so if someone (maybe AllyCat) could correct me, that will be good.

Tyntop, with a multimeter you can check the number of click per revolution. If they are 8, the results of the formula will seem more realistic (1.5m/s per revolution).
Every manufacturer should provide the "wind run per revolution (or per pulse)" of their product.

Reply
#14

Hi,

Ah, I hadn't looked at the data sheet, and it does make a few "strange" statements*, but the formula should tell us what we need.  The number of pulses per revolution is not important as such, although it's useful to tell us if the other figures look "sensible".

Personally, I worked backwards from a typical windspeed of 5 m/s (or around 11mph, 18 km/hr in units I'm more familiar with).  For 5 m/s the formula says that the frequency F (which is the same as pulses per second) will be 4.9 / 0.0875 = 56 Hz, which is unusually high for an anemometer.  So it clearly is generating quite a lot of pulses per revolution.  If we assume 16 pulses/revolution then the rotational speed at 5 m/s would be 56 / 16 = 3.5 revs/sec.

I can't actually see the diameter/circumference (for the centre of the cups) in the diagram of #10, but 413mm could be "about right" (I estimated slightly less from the figures shown).  So 3.5 revs/sec would give 1.445 metres wind run per second and an Anemometer Factor of (5.0 / 1.445) = 3.46.  That's slightly higher than the values suggested by Wikipedia (the Fine Offset anemometer is also towards the higher end) and this particular anemometer doe have quite short "arms", so I think these figures are all "believable".

Therefore 16 pulses/rev may well be correct, but of course a more usual number of 1 or 2 could be used by changing the scaling or calibration factors as appropriate.

*A few of the parameters may have "lost something in translation" (from Chinese?).  For example "Mechanical life: more than 20 million times" may look impressive,  but I make that only ~4 days of 56 Hz pulses  (5m/s windspeed) or ~66 days if  "times" is interpreted as "revolutions".   And "Electrical life: more than 70,000 times" must mean something very different.  

Cheers,  Alan.
Reply
#15

Hello Alan,

Excellent explanation. I always learn with your writings. Thank you.
Reply
#16

Hi Tyntop,

Based on the info posted by AllyCat, if you want to try the anemometer with the WeatherDuino, configure the custum anemometer this way:


Code:
// --- Custom type Anemometer ---
#if (WSpeedSensor_Type == 99)
#define ClickPerRev  16
#define FORM_FACTOR  1.445
#endif

If it will work (I believe it will) give a big thanks to Alan.

Reply
#17

Hi,

Thanks for your kind words.  Re-reading my text, the second sentence in my third paragraph is not strictly correct and perhaps I should really have said:

"So 3.5 revs/sec would give 1.445 metres of equivalent linear movement of the centre of each cup.  With the known Wind Run of 5.0 metres per second, the Anemometer Factor would be (5.0 / 1.445) = 3.46."

Cheers,  Alan.

PS: Sometime I must look to see exactly how the Wind Pulses are counted.  At a "maximum" windpeed of 100 mph (160 km/hr) this particular anemometer would deliver 560 Hz (or pulses per second).  That's not too high for most "electronic" (hardware) counters, but the 2 ms repetition period might be an issue for some interrupt-driven software.
Reply
#18

(20-11-2017, 10:56)AllyCat Wrote:  PS: Sometime I must look to see exactly how the Wind Pulses are counted.  At a "maximum" windpeed of 100 mph (160 km/hr) this particular anemometer would deliver 560 Hz (or pulses per second).  That's not too high for most "electronic" (hardware) counters, but the 2 ms repetition period might be an issue for some interrupt-driven software.

Hello Alan,

This is a pertinent thought. The wind pulses are counted using one of the two hardware interrupts available on the Nano. The 560Hz is not a problem for an hardware interrupt, however as you know, also important is the time needed to execute the code inside the ISR, and how all the software performs when there is an interrupt at each 1.78ms.
Before I bought my Inspeed anemometer, which produces 8 pulses per revolution, I did some tests by using a square have generator (simulating an amemometer), connected to the TX unit running the standard software. Honestly, I don't remember que maximum frequency I was able to achive before the readings start to become completly wrong, but I remember that I was happy with the results, as maximum frequency corresponded to wind speeds that not occour easily.

Any tests you do or thoughts you may have on this matter, obviusly are welcome. Perhaps a good reason to complete your TX unit, if you haven't finished it yet.  Wink

I think that there is a user, which is using a Vector Instruments model A100LK, which produces 10 pulses per revolution. Never got any feedback about any problem with this anemometer.
Now is the time for tyntop to give feedback to us.

Reply




Users browsing this thread: 1 Guest(s)