Hi Brian,
You fixed this in 10.19a but its broken again in 10.19k. 
Screen shot below. Only happens with the % display I am using.
Thanks.
Interesting āhistogramā look your solar data has.
Mine is more continuous. See attached .jpg . This happens to be version 10.19L.
Has always been this way here.
Wonder where the histogram look is coming from?
Using hobby board 1 wire electronics.
Brian
Henderson, NC
Today is the first day without clouds in a while so solar should be a smooth line as it increasesā¦
Attached is 3hour chart. Sunrise was at 6:54 this morning. This is the using the Davis Solar/UV sensors
check this out:
i.e could be a daylight saving issue
i.e could be a daylight saving issue
But you had it fixed in 10.19a and nothing in my setup has changed since then - except installing 10.19k. I use the straight % reading from raw/max volts. Daylight saving is not ticked and should not be a factor in this raw calculation.
That histogram look appeared when I changed the graphs to 6 hour, the solar reading is updated every minute as I understand. Iām also using hobby-boards solar sensor.
Not seeing spike in 10.19L with 1w hobby board unit
Also see no dropouts in solar plot. Great!
Did notice on 12 hour graphs the sunrise label is about 20 minutes late compared to WD computed sunrise time.
The sunset label is fine.
On 24 hour graphs the sunrise label is more than an hour off (if it shows up at all) while sunset is OK.
Regards,
Brian A
Henderson, NC
Are you using % or wm/2 display ? I believe it only happens with the % display.
hi have a 1 wire solar here (i brought a new one)
so i will do some testing, once i solder it up
Thanks Brian. You certainly did appear to have this fixed in a 10.19a version that you asked me try.
Also see no dropouts in solar plot. Great!
Unfortunately I still have these too, as done another 1-wire HB solar sensor user (I checked his graphs today).
i have checked the code, and i did not change anything after 10.19a
so i thought
but i will look again
Not sure iif the query was directed here.
Seeing no spikes or dropouts with % solar display on 10.19L
Attached is a plot of the solar % from yesterday.
Generally run 6 hour display and previous version dropouts were easy to spot.
Note: I am using a 1 wire hobby board but not the photodiode detector. Instead I have an array of 3 solar cells in a 3 sided vertical āskyscraperā configuration (North direction has no cell). Gave up trying to orient the photocell so it didnāt show some anomolous peaks/valleys.
Brian A
Henderson, NC
OOps,
Last post should read "gave up trying to orient the PHOTODIODE so it wouldnāt show anomolous peaks and valleys
so it wouldn't show anomolous peaks and valleys
Unfortunately these are not anomolous (so far as the sensor goes). Iāve watched the raw volts coming off the sensor and when the graph spikes up the raw volts have stayed at the correct level, so I can only presume a calculation problem. Iāve updated to 10.19n last night and within the next 3-4 hours will know if there are any dropouts, then Iāll look for the spike at around sunset.
The dropouts are taking the graph to zero, something that even heavy cloud on a rainy day does not do.
So at this stage Iām just going to hang tight until Brian gets his HB solar sensor kit built and can do some more tests. 
hi have a 1 wire solar here (i brought a new one) so i will do some testing, once i solder it up
Hi Brian,
Any luck with this as Iām still getting these solar jumps at the end of the day.
Johns āmilk bottle opaque filterā has fixed the drop-outs, but these jumps are something different agin.
Many thanks.
41 south, check that you have ticked, daylight saving , in the solar sensor setupā¦
Brian,
can you also do some testing on the hours?
Metservice still reports more sunshine hours here than what I am measuring, the difference is more than 4 - 5 hours a day.
Thank you very much.
41 south, check that you have ticked, daylight saving , in the solar sensor setup.....
Yes that is ticked - set to 60 mins.
set the 60 to zero
that is the problem me thinks (maybe)
that adjustment is only needed for fine tuning your time zoneā¦
OK will do that tonight and see what happens.
set the 60 to zero
#-o
Yahoooooooooooo
=D> :salute: =D>
Looking good Brian, I think you hit the nail on the head.
Maybe I ended up with this ticked once before with no off-set time and thats why I thought it had been āfixedā.
This spike has been bugging the heck out of me so its really really good to see it gone.
Thanks !!!