I’ve built a home made solar sensor for my VP2 and now just need to calibrate it. I got most of the information form an old topic on another forum and have modified for components readily available in the UK.
Can anyone post their solar readings (W/m2) from an original Davis solar sensor along with sky conditions and perception of how bright the day is. I know it’s a bit hit and miss, but absoulute accuracy is not a major problem as I would just like to know hours of sunshine and sky conditions for my WxSim forecasts. At the moment it’s very dull and my station is showing 132 W/m2. I have put a preset pot on the board to I can adjust the gain, so with a bit of playing around, I should get it somewhere near.
That’s not so easy. Max solar radiation depends on your latitude, your altitude, time of day and day of year. WD can plot a graph with the max solar expected for your location.
As Luis says it will be very dependant on location for the expected max however I do keep a log which contains date, time, sunrise, sunset, hours of daylight, expected max, actual w/m
Thanks for your responses and sorry for the late reply. I’ve now seen where WD graphs my predicted maximum solar readings so hopefully I’ll be able to get a fair calibration on the next sunny day (hopefully tomorrow or Saturday). I still have a problem with the sensor having an offset bias, this is giving a night reading of 56 W/m2 8O . I’ve got the day off work tomorrow so plan to have a good play.
Thanks again
Rob
I’ve built several Davis style solar sensors, my first rough calibration is to physically point the sensor directly at a clear sky, middle of the day, sun (may be a challenge where you are at the moment) and call that 1000 watts/m2.
I’m using a photo diode and an op-amp configured as an inverting amplifier. I’m going to tackle the bias offset first, then calibrate later. Once I’m happy I’ll take some pics and start a new post in the correct section.
Cheers for now.
Are you using the classic photodiode amp configuration (below)? If so, and you have offset, then you may not have a good choice of op amp. Happy to help if you care to provide details.
No I’m usig a single opamp as shown below. I’ve thought the opamp in question (LF358) may have been to blame so a bought another (CA3140) with offset nul pins. Even this will not go to ground and was giving about 16W/m2 in complete darkness. By putting a pot between the output and ground I can get 0 W/m2 but at about 10 ohms resistance, which is clearly useless as the chip cannot produce enough current for the required voltage drop. One of the main difficulties I’m having is finding an opamp which will work on a single rail at 3v. Your diagram is using a 5v rail, does this mean you have a separate power supply?
Those amps won’t do it, you need a different one, TLV2472 is what I used, single supply, low power, works down to 2.7 volts, offset is good enough for that circuit.
As pictured that circuit works fine off the VP(2) ISS, 5 volts was used to test, I didn’t have a 3.3 supply.
I have a 50K pot as a feedback resistor and get values up to about 1034 W/m2 . If I short or put the non inverting input to ground I get 56W/m2! Sunlight values are around 200 W/m2 in tests
If I short or put the non inverting input to ground I get 56W/m2!
Assume you mean the inverting input? You need an opamp with lower input currents, the TLV2472 is in the pico amps vs nano amps for the LM358N, so 1000 times better for this application. Look for any single supply CMOS parts that are available to you.
Thanks again Niko. I’ve found a supplier (RS Components but it will have to be mail order, so I’ll shelve this project until I get supply. I need to order a minimum of 5 so there should be some spares if anyone in the UK is interested.
This thread has me intrigued. I would like to build the sensor as per your circuit above. Can you give me more specific details regarding component values, eg specifically for R1 and R2, and C1 and C2. Also which specific TLV2472 did you use, there are quite a few variants available from RS Components, in Brisbane, Australia?
I have built many 1-wire sensors including my current 1-wire solar sensor, and now that I have recently acquired a VP2, would like to have a go at a homebrew solar sensor for it.
must be the chip.