Difference in rainfall on console vs. Cumulus via Envoy

[Vossenberg]

I am experiencing a strange difference in rainfall. The rainfall recorded by the console is different from the rainfall recorded by Cumulus.

For example today I have 0.2 mm recorded on my Vantage Vue console. Cumulus shows 0.3 mm. I have setup the envoy to use the 0.2 rain sensor.
Where could the difference come from, and how is it possible to show the same values as on the Vantage Vue console?

Thank you in advance for any suggestion,

Kind Regards,

Pascal

[steve]

That does sound strange, considering the resolution of the rain gauge is 0.2 mm (I assume it is the same for the Vue?).

Please zip up the diags folder and attach it.

[mikkimii]

Have you changed the value of the calibration different from 1,000? Configuration->Calibration->Rainfall.

[Vossenberg]

Steve,

Hereby the Diags, the tipping cup is indeed 0.2mm.

I have not changed any calibration settings, the rainfall calibration in Cumulus is at 1,000

Thank you in advance for analysis of the files.

Pascal

[steve]

Here's where the rain was recorded:

21-1-2013 12:40:00.923 : Writing today.ini, LastUpdateTime = 21-1-2013 12:40:00 raindaystart = 22,1000003814697 rain counter = 22,1000003814697
21-1-2013 12:46:48.790 : New monthly record, month = 1: 2013-01-21 12:46 22,400 "Highest monthly rainfall" 2013-01-16 00:00 22,102
21-1-2013 12:50:01.008 : Writing today.ini, LastUpdateTime = 21-1-2013 12:50:00 raindaystart = 22,1000003814697 rain counter = 22,3999996185303

The counter increased by 0.3 (Cumulus uses the annual total). Are you sure you have the Envoy set to 0.2 mm and not 0.01 inches? The latter would result in a tip size of 0.254 mm, which the Davis DLL would probably round to 0.3 mm.

Which build of Cumulus are you now using? You should be using the 'official' 1058 with the older DLL. One thing that the newer DLL does is that it sets the units in the console (unconditionally) and there seems to be a bug. I noticed that after I'd tried the newer DLL, my console had been set to degrees F, even though I use C. The rain collector type may also have been set incorrectly by the new DLL. Actually, I should check whether that's happened on my console.

[steve]

My counter is still going up by 0.2 mm each time, so I don't think the DLL can be blamed for what you're seeing.

[Vossenberg]

Steve,

I am using the 1058 official version at the moment. I have checked the setting and it is actually at 0.2mm.
The attached image shows the setting with in the background the monthly report.

Where the rain today shown is 0.2mm. Strange that it actually reports another number via the DLL.

[mcrossley]

Isn't it the 22.1 starting figure that is also wrong - that is not a multiple of 0.2

[steve]

The image you've shown is Weatherlink talking to the Envoy, rather than the Vue, yes? What does it say your annual rainfall total is? Adding up the totals on that display gives 17.6 mm, which is 88 tips. If those were being recorded as 0.01 inch tips, that would be 88*0.54 = 22.4 mm (rounded) which is what the DLL says your annual total is. So something thinks your rain gauge is the 0.01 inch version.

[steve]

Isn't it the 22.1 starting figure that is wrong - that is not a multiple of 0.2

See the post I've just made. It's recording 0.01 inch tips rather than 0.2 mm.

[Vossenberg]

The image you've shown is Weatherlink talking to the Envoy, rather than the Vue, yes? What does it say your annual rainfall total is? Adding up the totals on that display gives 17.6 mm, which is 88 tips. If those were being recorded as 0.01 inch tips, that would be 88*0.54 = 22.4 mm (rounded) which is what the DLL says your annual total is. So something thinks your rain gauge is the 0.01 inch version.

Indeed the 17.6 is the right number the console shows this value for this year. The conclusion that somewhere it starts counting as 0.01 inch must be correct. The question is where does it come from, and how to correct it?

Weatherlink is communicating with the Wireless Envoy, (As well as cumulus).

[steve]

Looking at the serial protocol, the annual total is transmitted as the number of tips. So it's the DLL which converts that into an actual amount. So the DLL thinks you have a 0.01 inch rain gauge. The DLL determines that by asking the console (or the Envoy in your case). Have you tried re-setting the rain gauge type using Weatherlink? That's how you normally do it with an Envoy, I assume? With a display console you can set it on the console itself.

Perhaps this is another problem with the 3.0 firmware?

[steve]

Note that if you do manage to reset the rain gauge type to the correct setting, it may not be immediately obvious in Cumulus, as it won't apply retrospectively. You'll need to look in the diags file to see if the counter has changed to the correct value.

[Vossenberg]

I have done another walkthrough setup for the station, when I choose Vantage Vue as station type in the start.

When at the set rain calibration number on you station option, it complained that the station detected was a vantage pro instead of a vantage vue.
(Probably the Envoy identifies itself as a Vantage Pro)

rain_calibration.JPG

Therefore I went back and set the station to Vantage Pro. Now it is possible to actually set the tipe of rain gauge.

rain_calibration2.JPG

As shown in the second picture it now asks if I want to set the rain cal number for a .2mm collector. I have accepted this and continued.

Let's see if it will actually report the same now. Now we need some rain to test actually

[steve]

If you look in the latest diag file, it will show the rainfall counter, which should now be 17.6.