Over the last 2 years I have been watching a bad trend of a 4-8 deg error every day at 830am before the fan started spinning and at night time had a serous lag in temp changes. Acurite's solution was changing the top head to include twin solar panels. But I kept hearing reviews of the same problem still being present. I felt it needed to go a step beyond and come up with a superior solution.
I started looking at this 50x50mm panel that would fit the stock head. It had some promises with more l crystalline silicon surface area and higher voltage. When it came time the higher voltage was helpfully but only half the short circuit current. It came out to be 10% less efficient so I decided not to pro sue the unit further. If any one was interested. 50x50 unit stock is 49x49 
So I came to the realization that the panel needed to be external mounted so I came up a interesting idea. I wanted the unit to run 24-7 and at the
stock motors true higher voltage 2.5v rating. I found this unit vary
interesting. It uses a laptop Li-ion 3.7V/1800mAh cell to store its energy. It setup to run at a higher voltage and only turning on at night. . The solar light panel
I received the unit but not to my shock, one of the lights was busted. Not that I really cared any way but the panel was a vary decent quality. Fist step I did some preliminary test running a stock fan with a cheap step down module.
I dialed my fan to 2.5v at 50ma and after 2 days still going strong I called it good. The math works out to hours 36 hours per charge in darkness. But it exceeded that so I was pleased with the prelim viability test. Next step was I needed to get the charge controller running during the day. After probing the stock pcb I decided to ditch the stock unit for a better unit with crazy fast shipping!
Only 5 days to ship from china! I did some testing and found the led indicators and under-over volt protection worked better then expected. I did a 4 day run with the panel on my deck with poor lighting and found the battery indicator in the morning had only dropped to 75%. After a few hours it was back to full 100%. I decided it was time for a full beta test,
Next week I will post some results but so far it has done a excellent job at producing way more accurate temperature readings during the day and at night. I will do a build write up manual next week on this and replacing the fan motors and temp sensor.
Fell free to check out my station for your self.
