Sunshine Recorder

[Olejul]

Hi Apenwith and thank's your response.

I am getting some what keen on your project and intend to make a sensor similar to yours.
I will include an ambient light LDR pointing to north parallel to earth axis.
My plan is then as follows:
- Include a two channel A/D converter to measure resistance in the ambient LDR and the sun detector LDR.
- Get access to the humidity data from my new weather station (probably Le Crosse WS-2355 - not purchased yet) via a Cumulus file.
- Include software which can calculate the sun angle over horizon versus time and which can also calculate the theoretical sun radiation as function of the angle and humidity.

I then intend to record these data over a period of time for post analyze to see if it is possible to establish the optimal criteria of sun On/Off taking the above mentioned data into account.

It is quite surprising to surf on the internet for sun recorders - Cambell-Stokes (heavy, clumsy and useless for electronic processing) gives virually a million hits while McCleod principle is more less limited to your entries in the Sandysoft forum.

Do we almost have a revolutionary system available?

Best regards

Ole

[Olejul]

Hi
I digged a bit into some theoretical aspects - I have now created a small VB application which can calculate the sun angle above the horizon versus day of year and time of day.
The graph 'Sun Elevation' shows the result at my latitude 55° N for Dec-21, March 21 and June 21
The second graph 'Sun Radiation kW/m2' shows the theoretical sun radiation as a function of sun angle over the horizon - I assume that equals available light.
As seen quite a difference between summer and winter, and I think also it somehow illustrates the difficulties you have experienced at dawn/dusk.
Humidity is not included as a factor and I have not been able to find any documentation on that subject yet.
By the way - I did order the smal USB A/D converter you have referred to in one of the earlier post.
Only $29 - expect freight from US to exeed that.
I hope I will the hardware up and go soon so I can start my recordings.
I will keep you posted.

[apenwith]

Hi Olejul
That is exactly what the project needs - programming skills which you obviously have. I'm not sure if you need to take account of the radiation to any great extent as the characteristics of LDR's show a big difference in resistance when the sun's spot is reflected and when not but some degree of seasonal compensation may be needed but I suspect it does not need great accuracy. It occurs to me that the reddening of the sun at low altitudes may be more significant in the dawn dusk times. It might be worth investigating the LDR characteristics in relation to light wavelength. But if you have a resistance threshold set by date and time logged and compare the actual reading I think the problem would be solved. I did have a few occasions when VERY bright cloud reflected light was recorded and that may be where the ambient LDR is useful. I think Mcleod was never successful because the trace on his paper would have been quite small and difficult to interpret with any accuracy. Why it hasn't been taken up with a more modern sensor I don't know - I believe it does have real potential. Maybe we'll see an Apenwith-Olejul recorder as the standard !! Good luck
Regards
Alan

[Olejul]

Hi Allan.
My idea with the curves is to establish a dynamic threshold level dependent on season and sun angle.
That would compensate for the radiation differences dawn/dusk and summer/winter.
You have a very good point about the sun color at dawn/dusk.
The sun spectrum seems to fairly the same magnetude in the visible part, but due to humidity/polution etc. it becomes red when the light has to pass trough a much deaper air space.
I just investigated earlier to day and found out that nearly all LDR's peak sensistivity is at 540 nM = green. I will investigate for an anternative like a photo diode - they should be available with peak at say 630 nM = red.
Just for the hack of it I tried one of these red glasbulb for the christmas tree - it had a very bright reflection of the sun while the surroundings looked more dim compared to using an electrical bulb (very subjecttive observation).
There is still some experimentation ahead so it will take a while before Apenwith-OleJul Corporation enters the stock marked internationally.

Best regards

Ole

[apenwith]

Hi Ole
I tried the metalised Christmas bauble and it was too reflective - bright clouds easily started the timer.
The white bulb proved best for me
Regards
Alan

[Olejul]

Hi Allan.
The christmas bulb is back in the box untill proper christmas time.
The below is the result of pure desktop work and it should cover all latitudes - even the poles who ever want to record sunshine hours such a place (there will however be an iterruption caused by the mounting mast).
My goal was to achive at least 90 degress aperture angle to cover all cases.
I have found a photo diode which I think will be suitable - it an Osram type SFH-2302. Its peak sensitivity is at 800nM (infrared) and it has a 80% sensitivity at 640nM (red).
I will purchase 10 pcs (least quantity) - price is only about £ 4 for 10 pcs.
If it works I will post you a few.

Best regards

Ole

[apenwith]

Hi Ole
Very nice drawing - the only problem with mounting the diode without a tube to shield it is that it will be able to 'see' all light incident on it. The tube screens off all but the diameter of the bulb. Hope that makes sense
Alan

[Olejul]

Hi Alan.
Thank's your remark - I am aware of that issue.
The photo diode by itself have a fairly narrow sensitivity beam, but I have already made a revision of my construction to keep stray light as low as possible.
Btw - I just received my A/D converter and in e few days I will get hold of the photo diodes and start experimenting.
I will keep you posted of the progress.
Best regards
Ole

[Olejul]

Hi Alan.

Just wanted to inform you that I expect to have my gear up and running to morrow.
Actually I made 3 prototypes with different reflector solutions (pictures will follow).
By using my new USB A/D converter, my software will generate a file with date/time, sun elevation/azimuth and the photo transistor current from all 3 units plus ambient light.
System will record data every second from dawn to set (criteria is elevation > -10 degress).
My goal is to find the most optimal threshold level as a function of sun elevation and ambient light by analyzing the recordings in Excel.
I will keep you posted with the progress.

Best regards

Ole

[Olejul]

Hi Allan.
First a picture of my prototypes.

The graphs shows the result of a 4 minute periode where some smaller clouds moved fast in front of the sun.
The comparison graph shows the relative reflectiveness and as seen there is not a big difference between them.
The 'gold' bulb seems though to have the highest contrast which leads me to believe a real mirror glasbulb would be the optimal.
The second graph "Reflected Light / Ambient light shows the relative dynamic range and my decission criteria for sun-is-shining is simply when reflected light is higher than ambient light.

The test setup was rather primitive - just on a window shelf looking through double glass which could need a little cleaning.
I further need to shield off the ambient light photo transistor to ensure it never receives direct sun light.
I will leave the system recording to night and hope for a sun dawn with clear sky.
Up to now I think it looks promising.

Best regards
Ole

[apenwith]

Hi Ole
I've just written a reply and my satellite broadband failed ! I hope this doesn't appear twice.
Well done - great work. The peaks seem so high that I wonder if the ambient comparator is essential. I have a feeling that it may be of more use to give a baseline for software to use to compensate for seasons and maybe the dawn/dusk problem I found. I couldn't quite see where the ambient diodes were positioned so I'm not sure what they see. Anyway I'm very pleased that someone with more technical skill than me has progressed the theory forward so much - well done again.
Regards
Alan

[Olejul]

Hi Alan.
Good to have you back on air - we can't exist properly without internet to-day.
I am just about to have my third prototype version finished - the attached drawing shows how.
The reflector mirrow is now much smaller in height as only about 40% of the surface is in use.
The process was as follows:
- Red christmas bulb out of the box again.
- Fabricate the holder made of 12mm MDF plate for outside use
- Remove the outer red paint using cellulose thinner (the internal side is silver coated)
- Glue the bulb to the holder using two component epoxy glue
- After hardening carefully break of the surplus glass using a small cutter
- Glue a water tight lid on top using silicone

I have carefully designed the tube for the phototransistor so it only can see the bulb and all items are painted with matt black lacquer to eliminate stray light best possible.
To compensate for the dawn/dusk reduced light intensity I will increase the phototransistor sensitivity by a factor of 10-20 until sun is say 5-10 degress above the horizon.
The graphs actually shows that some sun light is reaching the ambient light sensor - I belive my modified design will help on that issue.
As soon as I have some meausurements and some conclusions you will be the first to know.

I am just hoping the have a nice dawn with clear skies and maybe a little haze to check out the increased sensitivity - dusk is not useable untill i get the unit up on a mast approx 10 meters above ground level.
Best regards

Ole

[Olejul]

Hi again.
I attached a wrong file - now in jpg

[Olejul]

Hi Alan,

Attached some pictures of my present prototype which I expect to mount together with my new La Crosse WS-2355 aneometer during the week-end.

The gain interface seems to be working OK.
This moorning i could sense the light increase from about 1/2 hour before sun rise.

Today is a very dull grey day with some fine rain - sun seems to be none exsistent
Best regards
Ole

[apenwith]

Hi Ole
I AM IMPRESSED - looks very professional and sounds as though you have hit the right solution.
Look forward to the test results. I hope you get to see a bit more of the sun than I have over the last few days.
Regards
Alan