[danner]

The wires you are using are short enough to not need to be twisted at all. Maybe it's better not do it.
A reading of 12w/m2 in the dark is to high. If possible try to measure the voltage output of the SP-110 in the dark, it should be zero (testing things before having them installed its easier, however I know that sometimes isn't possible). Even with a 0V input, can happen that the output voltage of the OP drifts from 0 to 3 mV, but will never reach a value of 7mV (which is needed to produce a reading of 12W/m2).

OK

Simply it isn't calculated.
If you find a way to calculate it with an Arduino, without consuming a lot of memory, please let me know, I will be very happy to include it.
The WeatherDuino Pro2 emulates the Davis communication protocol, and some aspects of a Davis weather station like the data logger, but it isn't a Davis Weather Station.

OK that answers it. Thanks.

[werk_ag]

Well, good work!

Of course I also forgot to stop Cumulus MX before I yanked the sensor connector off the TX board. Now I have to edit all the ini files and logs to delete the solar readings at 10pm at night, LOL. Oh well.

This is one of the reasons why I'm still using Cumulus 1. The Editors!
Editing any file with the editors included in Cumulus 1 its very easy. The editors as they work in Cumulus 1, are the most claimed feature for CumulusMX, until they are available (if they ever will be), I will stay with Cumulus 1.

[danner]

Paper to evapotranspiration formula. FAO Penman-Monteith equation. http://www.fao.org/docrep/X0490E/x0490e06.htm

This is one of the reasons why I'm still using Cumulus 1. The Editors!
Editing any file with the editors included in Cumulus 1 its very easy. The editors as they work in Cumulus 1, are the most claimed feature for CumulusMX, until they are available (if they ever will be), I will stay with Cumulus 1.

Yeah it's a pain in the rear for sure. LOL. I have to wait until midnight so I can edit the log for today. Oh well, it's not the end of the world just some extra work.

So far so good. No fluctuations thus far that I've noticed. Hope it worked!

[danner]

This is the graph from last night. Sunset was around 7:30pm. The western sky was still lit until 8:15pm-ish but not at 10pm. LOL

[werk_ag]

Paper to evapotranspiration formula. FAO Penman-Monteith equation. http://www.fao.org/docrep/X0490E/x0490e06.htm

I know it. The problem is implementing that math in a microcontroller like an Arduino Mega. Even more difficult when the existing code already uses a lot of memory. Its so hard to implement, that Davis used an approximation, recurring to tables to avoid some calculations (of course the tables used are not published, at least I haven't found them anywhere).

Yeah it's a pain in the rear for sure. LOL. I have to wait until midnight so I can edit the log for today. Oh well, it's not the end of the world just some extra work.

So far so good. No fluctuations thus far that I've noticed. Hope it worked!

It seems you have removed the "clip anything below 6" which is good to test if the spurious readings are gone.
I will be following your webpage.
Whish you a good night. I will be back tomorrow.

[danner]

To answer a question I missed further back. TX1 is powered by a U1 sized battery, like my TX0 is, but with a small AC-DC trickle charger connected to the battery. At first I thought that maybe the trickle charger was "dirty" and making the readings fluctuate, but I disconnected the charger so it was just running on the battery and the readings still fluctuated.

It seems you have removed the "clip anything below 6" which is good to test if the spurious readings are gone.
I will be following your webpage.
Whish you a good night. I will be back tomorrow.

Yes, I edited out the clipping in the code.

Good night, sir. Thank you for the help and advice.

[danner]

Oops, just got a 1 w/m2 reading. Looks like I'm going to edit the code to clip 1.

Made a quick review of the professional weather station just to my north that also uses an apogee pyranometer but the more expensive one. Apparently that one also has the occasional blip during the night. That station shows a few of these at night. 0.676578 w/m²

https://www.wunderground.com/personal-we...008/mdaily

So it looks like the occasional low reading isn't abnormal, but 12 w/m2 is.

[danner]

At 4:15am I edited the code again and set the clipping to 4. Guess it made no difference switching the wires. Oh well, was worth a try and was free. LOL

[werk_ag]

... Guess it made no difference switching the wires.

By the contrary it seems that it did a big difference. According to your graphics, during the last night, you got some spurious readings but this time they are no more than 3W/m2 most of them are only 1W/m2 (none of 12W/m2).
I think I already have explained why these readings could occur, also we don't know if the output of SP-110 is always at 0 during the night time.
Clipping the readings below 4W/m2 doesn't seem a big deal, it doesn't have any effect in the sunshine hours calculation (that only starts when readings are over 50 to 100 W/m2).

You can also try disconnect the ADC1 cable, currently there is nothing using it.

[danner]

The big test is when the sun goes down. The last 2 sunsets it didn't want to stop reading something until well after the sun went down, the stars were out, and continues to read 12+ w/m2 for hours afterward. Will continue monitoring.