Cumulus Support

Hi Alan

1)Bruce is not my family name
2)i only moved to Oban 6 years ago and hail from geordie land, good old newcastle upon tyne, toon army country.(sorry for those out side the uk thats north east england)

as for the rg11 from the online info on the site it keeps saying about measuring the amount of ir light detected hence the shields to stop external ir light causeing problems.so i think that it must have some fixed light levels to determine the rain drop size. the more light is refracted the more rain on the surface. this is only a guess.

they claim
"The RG-11 detects the surface condition, caused by dirt, contaminants, aging, and other factors. It properly compensates the data, making the device virtually impervious to environmental factors."

so that is how i have come to that conclusion.the tip bucket size should not matter from what i can see. im only using the 0.001" or 0.01mm to test what areas cause the relay to operate and looking for clear on/off triggers and not the on/on/off that i was seeing before the cap / resistor mod. the test i did was both with and without the mod and without the mod i did see on/on/on/off as well as clean on/off.

maybe other rg11 owners could try testing the areas to compare results


bruce

Hello Alan,

I did not check the debug this time and will check on the next rain. This rain was very short lasting a few minutes around 5 am. I had shut off debug a week or so ago after last rain and confirming the double tips with said configuration. I posted my results assuming that it was always going to be showing the same on the Cumulus debug when it rained. (it did last week). I will add a graduated cylinder rain guage that I have somewhere around to the mix mentioned.

Bruce,

Wife here too is wondering what I am doing climbing a ladder to play with the rain guages every other day. That said I have to check on the test generic IP camera as the storms from last week knocked out the night illumination so I have to go back and remove it to check it.

Hi

just had a reply from Hydreon about how the rg11 works.it would seem that it does count the number and size of the drops but works this out from the changing levels of light. so i think having the tip bucket set more sensitive should give more accurate rates.interesting statment in the second paragraph "and not physically measuring the amount of rain. "

the full reply is below

Hello Bruce,

The RG-11 bounces beams of light within the lens. When drops hit the outside surface, it allows some of the beams to escape. The RG-11 detects the change in beam intensity, and determines the size of the rain drop that caused the change. The RG-11 Rain Gauge counts the number and size of the drops hitting the sensing surface. The algorithms within the RG-11 interpret this data and produce the same sort of output produced by a tipping bucket style rain gauge.

One condition that may affect the accuracy of the RG-11 is if small droplets get stuck on the lens. The can happen because of wind pushing droplets back up the surface of the rain gauge, or any other condition which would keep droplets on the lens for a longer than normal period of time. That said, the software is calculating a predicted amount of rain based on data and statistical analysis that we have gathered in the past, and not physically measuring the amount of rain. Even in perfect conditions, there may still be some variation between the RG-11 and an accurate tipping bucket. Hopefully this answers most of your questions.

Regards,
Anders Hendrickson


please feel free to comment of there answer im sure you will.


bruce

Hi
So far my funnel modified RG11 set at 0.2mm tip size appears to need a factor of around 0.3 in most conditions compared to the manual gauge. But it's early days yet. The un-modified one is seriously under reading in today's continuous moderate rain.
Regards
Alan

Hi
During the last three days when it has rained the results of the funnel modified RG11 against the Manual Gauge are:
RG11 using 0.3 multiplier and 0.2mm bucket size
1.0mm
8.8mm
20.8mm
Manual Gauge
0.6mm
11.1mm
19.0mm
Not exact but close.
The Cumulus RG is seriously under-reading by a factor of about 4 in today's rain so I wonder if the debug log will tell me if there are any contact losses.
Regards
Alan

so far we have not had any rain.
im away on holiday for two weeks starting tomorrow and was wondering should i leave the debug log running? in two weeks how big would it be and could it still be read using notepad? hopefully over the next two weeks oban will have some rain so i can get some results to compare.

bruce

The debug log component crashes if you leave it running, the time overflows. I can't remember how long it takes to overflow, though.

thanks Steve ill just switch off the debug and see what happends over the 2 weeks.

We had about 5 minutes of rain (light sprinkle) this morning (literally). Not enough to get any measurements.

I've only turned on debug when I am thinking its going to rain; then shutting it off after a rain. That said very little rain to get any measurements.

I had a bit of rain today July 13, 2012. Here are my results

Cumulus via Davis Vantage Vue Tipping bucket - 0.01 in (note that the Davis instrument is not adjacent to other 3)
Cumulus via RG-11 - 0.02 in
Dallas Tipping bucket - 0.00 in
RG-11 via 1-wire Dual counter - 0.011 in

Hi
Results for July so far. Note the point that the manual site RG11 was modified. I was away for a few days so the manual gauge was only read when I returned so I have put in the 3 day total for the others. Regards
Alan

I had a bit of rain today July 19, 2012. We are in a drought period here in the midwest. These counts are from midnight.

Here are my results

Cumulus via Davis Vantage Vue Tipping bucket - .28 in
Cumulus via RG-11 - 1.54 in
Dallas Tipping bucket - 0.59 in
RG-11 via 1-wire Dual counter - 0.788 in

The new beta logs though are more informative showing the double tips and the counter increments. I guess maybe a 2 for 1 count might work? It doesn't make sense that its a debouce causing the double tips as the 1-wire counter is not incrementing in the same fashion.

Cumulus counts: 1.54/2=0.77 inches 1-wire dual counter counts-0.788 which is closer and the two sensors are about 20 inches apart.

2296.3734 : 9:16:53 PM WU Response Code: 200 Response = success
2296.6688 : 9:16:56 PM RG11 1 state change: On
2296.6688 : Adding 0.010 Total = 0.140
2296.6688 : 9:16:56 PM RG11 1 state change: On
2296.6688 : Adding 0.010 Total = 0.150
2296.6797 : 9:16:56 PM RG11 1 state change: Off
2296.8734 : 9:16:58 PM Updating Wunderground

I'll have to save more buffer / watch the 1-wire counter some more. Probably do maybe run owfs/1-wire 9097 on one of the seagate dockstars I have sitting around to get some better counting. Been playing with these very small USB dual port 9097s using same technology imports.

19 09:46:58 | Port=COM4 Address=B900080013564B10 Family=DS1920 Temperature=21.3125 |
19 09:46:59 | Port=COM4 Address=F900000009EA8F1D Family=DS2423 Counter=170713/43222 |
19 09:47:01 | Port=COM4 Address=B900080013564B10 Family=DS1920 Temperature=21.3125 |
19 09:47:03 | Port=COM4 Address=B900080013564B10 Family=DS1920 Temperature=21.3125 |
19 09:47:04 | Port=COM4 Address=F900000009EA8F1D Family=DS2423 Counter=170713/43222 |
19 09:47:06 | Port=COM4 Address=B900080013564B10 Family=DS1920 Temperature=21.3125 |
19 09:47:09 | Port=COM4 Address=B900080013564B10 Family=DS1920 Temperature=21.3125 |
19 09:47:10 | Port=COM4 Address=F900000009EA8F1D Family=DS2423 Counter=170713/43222 |
19 09:47:12 | Port=COM4 Address=B900080013564B10 Family=DS1920 Temperature=21.3125 |
19 09:47:13 | Port=COM4 Address=F900000009EA8F1D Family=DS2423 Counter=170713/43222 |
19 09:47:16 | Port=COM4 Address=B900080013564B10 Family=DS1920 Temperature=21.3125 |
19 09:47:17 | Port=COM4 Address=F900000009EA8F1D Family=DS2423 Counter=170713/43222 |
19 09:47:19 | Port=COM4 Address=B900080013564B10 Family=DS1920 Temperature=21.3125 |
19 09:47:21 | Port=COM4 Address=B900080013564B10 Family=DS1920 Temperature=21.375 |
19 09:47:22 | Port=COM4 Address=F900000009EA8F1D Family=DS2423 Counter=170713/43222 |
19 09:47:24 | Port=COM4 Address=B900080013564B10 Family=DS1920 Temperature=21.375 |

Running most current beta of Cumulus. I still see the double tips in the logs. Numbers "appear" to be twice that of what the 1-wire dual counter is reading. That said though it looks that the Dallas tipping bucket is under reporting a bit. I have a graduated rain cylinder (plastic) that I can put next to the two RG-11's, Dallas Tipping bucket that I might do in the next days or so. I am also swapping out Davis Vantage Vue for another Davis Vantage Vue that I have; Davis Vantage Vue #2.
The Davis numbers don't make sense to me so #2 will be mounted closer to test area and connecting two Davis consoles shortly as one will read two Davis weather stations one at a time.

I'm beginning to like the idea of using the RG-11 just as a rain sensor "on / off" switch and continuing to utilize the rain bucket method of measuring rain. It does work much better than the sponge rain on/off Rainbird sensor. (that too is connected).

Hi
Results for the first 12 days of August. Somewhat mixed - it's hard to draw any conclusion.
The manual site RG11 still has the funnel modification. The Cumulus attached RG11 has the bucket size set at 0.36 mm. Regards
Alan

I haven't had enough rain here to get any kind of benchmark at all.

That said I do see that the Davis Vantage Vue bucket counts have coincided with the Dallas tipping bucket counts.

The RG-11 connected to the 1-wire counter sometimes is almost synced and other times totally off.
(wind?)
I have not modified the RG-11 connected to Cumulus hardware yet and still see the 2 for 1 counts. (maybe just a calculation change factor would work for me?)

That said I am not using any sort of funneling and have the devices all within 1-2 feet of each other.

The above noted I have the RG-11's numbers going to different DB's. That said will broadcast xAP wise the RG-11's and the Davis Vantage Vue over to one DB such that I could graph the real time accumulation and put it on one graph.(chartdirector). The attached is just the Dallas tipping bucket using the HB 1-wire counter.

I do have one touchscreen display showing all four counts live on one screen via flash.

Steve, I am currently using the cumulus real time txt file to populate a small mysql DB. Is there an easy method to get the RG-11 data to my DB via the real time txt file? Am I going about this wrong though?

Just noticed that we had a bit of rain here yesterday and the RG-11 (unmodified) connected to Cumulus and the Davis Vantage Vue matched rain counts at .22 inches. I am current running 1043.