[JT118]

It is not really a question of Weatherduino matching your gauge. if your gauge is in error then it is in error.

However one of the many useful attributes of weatherduino is that it is very configurable, and you can compensate for the gauge error, in the case of the rain gauge, almost infinitely, although the arithmetic becomes a bit harder.

You mention 653ml of water registered 18.2mm on the readout. The other figure you will need is how many tips it took to achieve this. If you are going to alter Weatherduino to compensate for Gauge error then you will need these three figures:

Readout figure: 18.2mm
Water introduced to gauge: 653ml
Number of tips to achieve the readout:??

If you have change the TX config file to 0.200 as follows;

*/
#define COLLECTOR_TYPE 0.200 // Please note: Values must be in metric units. Max. resolution: 3 decimal digits
// --------------------------------------------------------------------------------------

then it will have tipped 91 times to register 18.2mm. (18.2mm @ 0.200mm per tip = 91 tips.)
This would work out 7.18ml per bucket tip. (653 / 91 = 7.18)

As your gauge is 200 mm diam then:

i) 200 / 2 = 100
ii) 100 x 100 x π = 31428.6 mm2 cross sectional area.

Therefore: we have 31428mm2 area receiving 653ml of water this does equate to 20.78 mm of rain:

(653 / 31.4286 = 20.78 )

20.78 divided by 91 = 0.229 mm of rain represented on readout for each tip:

Therefore you need to change your TX config file to 0.229; as follows. (Don’t put in 2.2 as in your last post you will register 10 times the correct rainfall)

*/
#define COLLECTOR_TYPE 0.229 // Please note: Values must be in metric units. Max. resolution: 3 decimal digits
// --------------------------------------------------------------------------------------

Then as a check;

Get 628ml of water for our 200mm diameter OTA gauge (not 653ml for our 203.89 diam TB3) and introduce into gauge as described previously. You should get 100 tips and 20ml registering on the readout. (314ml should tip 50 times and register 10mm on readout).

628ml / bucket size 7.18 = 87.465 tips

87.465 x 0.229 = 20.03 mm which is as much accuracy as we can achieve with three decimal places available.

[JT118]

Good link with many doc and youtube video etc :

https://www.youtube.com/channel/UC1x0s3q...G7NK_613Zw

Watching this I am sure your gauge is a: OTA Tipping Bucket Rain Gauge
OSC971OB100A ( A = 0.2mm per tip rather than the 0.5mm per tip setup) not a hydrological services gauge. It is just possible, though unlikely, that you may have a 0.5mm rather than 0.2mm bucket. Hence the need to check using above methods. You are correct in your measurement of 200mm so you definitely need to use 628ml of water as your standard test and this should register 20mm on readout. You can see in your photos the rivet holes where the makers plate has been removed.

http://www.meteoew.nl/ota-rain-gauge.html

http://www.osc-tokyo-japan.sakura.ne.jp/...OB100.html

OSC971OB100.pdf Size: 235 KB  Downloads: 337

https://www.youtube.com/watch?v=4FAtUuULfHc

[werk_ag]

Hi Mark

...Do you think the WD software just isnt matched to the gauge..??

There is no need to set anything on the WD software. It gets the Collector Type from what is defined in the RX user config settings.

Can I put a decimal place in the TX config for the rain gauge, and make it 2.2 or something like that (instead of 2.0 or 3.0)..??? Or do you think I should crack the locktite and start mechanically adjusting the stop screws in the gauge..???
Cheers for now...
Mark R

Yes you can. Not in TX user config, but in RX user config settings.
Also don't forget to adjust Cumulus Rainfall multiplier according to this formula:

Multiplier = Collector Type / 0.2

For a 0.2mm collector (Davis default) multiplier should be 1.
Also keep in mind that Cumulus shows rain values rounded to one decimal place, however internally it uses more decimal places.
Just to illustrate: If we use an 0.25mm rain gauge (like mine), on first click, Cumulus shows 0.3mm, on second click, we get 0.5mm, on third click we get 0.8mm. The total is always correct.
This will also happen if you set a Collector Type of 0.22mm. On the first click Cumulus will show 0.2mm not 0.22mm, on second click 0.4mm, instead of 0.44mm, and so on... but as said the total rain value, is recorded on log files correctly.

[JT118]

Werk, As you say it is nothing to with the WD sketch but I do not see where this is in the RX sketch.

But Wizza must alter his TX config from;

*/
#define COLLECTOR_TYPE 0.200 // Please note: Values must be in metric units. Max. resolution: 3 decimal digits
// --------------------------------------------------------------------------------------

TO:

*/
#define COLLECTOR_TYPE 0.229 // Please note: Values must be in metric units. Max. resolution: 3 decimal digits
// --------------------------------------------------------------------------------------

If he does not he will always be reading less than 18mm when 20mm has fallen an over 10% error.

As you say cumulus will not necessarily reflect this on one tip but it will sort itself out after a number of tips.

"This will also happen if you set a Collector Type of 0.22mm. On the first click Cumulus will show 0.2mm not 0.22mm, on second click 0.4mm, instead of 0.44mm, and so on... but as said the total rain value, is recorded on log files correctly. "

Alternatively break the seals and do the considerable amount of adjustment to get exactly 6.28ml in each side of the bucket tip and 628ml of water to represent 20mm of rain on the readout. Then keep the TX config at 0.200. ( I would not do this unless something is obviously out which from the photos it does not appear to be, and it is very much easier through the software.)

[Wizza]

Hi Guys, I altered the collector_type to 0.229 in the TX config. I did run two very quick test based on 314ml of water and I got 42 and then 44 tips. Didn't have much time avail around family yesterday. It occurred to me that on the first test the syphon may have been empty and the buckets totally empty, thus 42 tips. The second test immediately after, the syphon would have been full and one bucket may have been part full, thus 44 tips. I re-ran the calculations, and agree that 0.229 seems to be the correct correction number. This would seem to indicate a bucket tip of about 7.1mL. This all said, while I could leave the calibration number at 0.229, maybe I should start to re-calibrate the gauge. Doesn't really phase me, and I think the tests are showing repeatedly that it is not calibrated. Who know, maybe the guy selling it made a gauge out of spare parts, and thus it looks all official but it is not calibrated. We'll never know. What we do know is that after installing a new reed switch, it seems to work, and work well, and looks like it needs some calibration. Its back to work for the week, and I'll look at put some time aside next weekend. Thanks for all the comments and assistance guys. Very much appreciated. This project is fast turning into a hobby, and something quite enjoyable. For me at least...
Cheers for now...
Mark R

[JT118]

I think it is one the best gauges you can get; and therefore worth calibrating correctly. It all depends on time and equipment. If you have both then it will be worthwhile. The above makes 314 / 42 = 7.476ml per tip. Need to adjust TX to 0.238 (10 / 42)

If you manage to get 314ml to register 50 tips + or - 1 tip (2.0%) and set the TX to 0.200 to get 10mm on readout repeatedly you will have done well.

Good luck; let us know how it goes when you have time.

[werk_ag]

Werk, As you say it is nothing to with the WD sketch but I do not see where this is in the RX sketch.

That's right. The Collector Type setting is in the TX user config options. Thank you for the correction.

[Wizza]

Thanks for the info JT. The following URL was quite informative, after doing a page translate:

http://www.meteoew.nl/ota-rain-gauge.html

I'm not sure the difference between the set screws under the buckets, vs the counter weight. Would seem that the counter weight is what you adjust to set the amount of water in each tip, and the set screws "balance" the left and right bucket quantities..???? Does that sound right..???

Mark R

[JT118]

Sounds right. The Kiwi video is quite good. I am honestly not sure of the exact process. But it might be an idea to start with getting each bucket to tip evenly and get it as near as you can to 6.28ml per tip. In theory if you get this exactly right and put 0.200 into the TX. Then 314 ml of water should result in 50 tips and 10mm on readout. I expect you will need to do small adjustments when you use 314 ml or 628ml as it will be difficult to measure one tip accurately using home / hobby equipment.

I had a quick look at the Dutch website. My kindle won't translate but if I read correctly he says 4 grams = 0.2mm . This can't be right, it is definitely 6.28ml (or grams) for a 0.2 bucket / gauge, but maybe it is my Dutch that is lacking. Looking closely at the makers plate, his is the 0.5mm bucket which means 20 tips will = 10mm . Each bucket tip will contain 15.7ml.

You could change TX to 0.238 (10mm / 42tips = 0.238). Given the numbers you are coming up with I wonder if your gauge has been calibrated as an imperial gauge, though Australia has been metric since the 60's I think. If it was, then I expect it would have been set to 0.254 or 1" = 100 tips (25.4 mm ) (0.01" per tip). I expect your gauge dates from the 1980's so it is possible.

The balance I think will alter the tip side to side. The set screws below the buckets will alter the volume. I expect you will need to do both simultaneously to get it right.

[JT118]

Thinking; driving into work this morning, that it is likely this gauge was calibrated to imperial units. It seems unlikely to me that anyone who put it together from a box of bits would bother to put seals on etc.

It seems to me to be more likely that the gauge had the makers plate removed when a changeover from imperial to metric occurred in some institution so that it wasn’t used and put into a store for the next 35 years.

A suggestion, before you break the seals; put 0.254 into the TX config file and then do a few runs of 314 and 628ml.

In theory if it has been calibrated to imperial units then each tip will be 5.08ml and represent 0.254mm of rain.

Therefore it should take 40 tips (39.878 actually ) to register 10 mm rain using 314ml of water.

If after a few runs of 314ml and 628ml you don’t get the required result with the tolerances specified ie: 3 to 4% then go ahead and break the seals and get the gauge back to metric.

One last thing, if you do break the seals, try to reseal once you have adjusted, the seal acts as both a thread locker as well as a visual aid to check if the instrument has been tampered with.