Measure sunlight

Here is a different question and thank you Bruce for you quick and informative answer. If I used a CdS photoresistor I would get a linear output for light input. The only problem is the devices can be as much as 3.66 times  different. One device can be 3k ohm at 10 lux and another of the same part number can be 11k ohm at 10 lux. Any Idea how I could add a sliding series resistor to make them all say 11k at 10 lux? the resistor would have to add 8k resistance  to the 3k at 10 lux to reach the 11k value set for all devices and to the same device at 110000 lux would have to add 125.33 ohms to the base 47 ohms at this lux value to make all devices equal to 172.33 ohms that an 11k device would translate to at 110000 lux of intensity. Again your help is most appreciated.

Alvin,

I don't know nearly as much about CdS photoresistors. I believe they are highly nonlinear and not well calibrated. They have a dark resistance that must be carefully calibrated for any consistency at low light levels. I believe they are seldom used for precision measurements.

Regards, Bruce.

CdS photoresistors are linear in light to resistance measurements and don't have big temperature problems. Their big problem is the the large manufacturing resistance range at say 10 lux. The reason I suggested CdS is because the photodiode will have nonlinear response to light and temperature due to the fact that it is a diode. I will have a heater in my device to melt snow from the box's clear top which may help stabilize the diode's response to temperature but the top of the box will be clear which will cause it to act as a greenhouse accumulating heat. There could potentially be a temperature swing of 22 degrees F to 159 degrees F. Any ideas? Thanks. Alvin...

Alvin,

It appears that you are describing the nonlinearity and temperature effects of a photodiode when used in photovoltaic mode. The circuit I provided operates in photoconductive mode and response is quite linear and not greatly affected by temperature.

See the link I provided earlier for a more detailed explanation.

Regards, Bruce.

Alvin,

I came across your question and I had a few follow-up questions regarding your project:

Have you considered using an Ambient Light Sensor (ALS) with a photopic response and simple i2c interface?

Is there a benefit to the discrete solution over an ALS?

Would an ALS that could measure up to 130kLux be satisfactory?

Thank you!

Jason Bridgmon

I have looked at ambient light sensors with I2C interfaces but did not find one that would go as high as 130k lux.  One problem is the domed top. I need a cosine diffuser but could probably install one on top of the domed top. I was considering using a light filter to reduce the 130k lux done to the 10k lux of the intersil ALS but if you have an ALS that would go to 130k lux that would be very interesting. This device would be out in the weather with rain and sunlight (measuring sunlight) but I guess I could create a housing for it. Note that I have to work from 30 F to 130 F (100 degree F temperature swing) and I don't know how the device would react electrically to that. Please get back to me with your thoughts. Thank you for your input it is greatly appreciated.

Jason - You never got back to me. What is the device you are suggesting?

Alvin,

My apologies!  We are working on a new device here at TI - the OPT3001.  It is an ALS and the preliminary data sheet is available if you would like it.  I can email it to you if you send me an email at bridgmon_jason@ti.com and we can discuss your application further if you would like.  I look forward to hearing from you!

- Jason Bridgmon

The TI OPT3001 is now sampling on the web.  Visit http://www.ti.com/product/OPT3001 for more information.

Jason Bridgmon

my measurement of sunlight was 146900 lux so I do not think your device will go high enough to measure full brightness. Also I probably need infrared as the device needs to register on cloudy days. The plant's pores open up on cloudy days and dry out fast as much energy is still delivered on cloudy days. Will your device integrate over time? can I lengthen the distance the I2C will operate (10 or 20 feet) by lowering clock rate to 1000 or 100 hz? Thank you for remembering me. I am using a photo diode with an OPA340 op amp and you LM331(?) voltage to frequency chip to provide a count that will essentially integrate over time.

Alvin,

The OPT3001 does only register up to 83klux, so it will not be enough of a full scale range to accommodate 147klux.  Also, the IR rejection of our device is 99% or better, which is absolutely excellent for matching the human eye but apparently not so good for plants!  The OPT3001 has two selectable sampling periods (both under a second) which increase stability and accuracy of the measurement.

Regarding the I2C interface, I've never tried running an I2C cable that far away and slowing it down to compensate for the line resistance and drop and other transmission effects.  My answer is "it will probably work" but there are some things to watch out for.  The first thing that comes to mind is that there is often a timeout for serial transmission protocols, and going too slow may cause the operation to be reset and interrupted when partially complete.  The other thing to keep in mind are the VIL/VIH/VOL/VOH levels needed.  Since long transmissions are lossy, it may be difficult to drive the line high or low enough to signal the other device.

USB2.0 communication can extend 5M without buffering, and that's running at 480Mbps, so anything is possible; whether I2C is robust enough to handle the low bandwidth at length remains to be seen.

Are the photodiode and OPA340 and LM331 working out well for you  or are you still looking for a different solution?

Regards,

Jason Bridgmon

Some time ago I was looking for a photo sensor to use in light studies for agriculture purposes. You suggested the OPT3001. It fits the bill as far as light frequency that I want to measure. Unfortunately I need two of them so it would have to be able to modify its I2C address. If this is not possible do you have a similar version that just contains the visual spectrum sensor and a transimpedance amplifier? Also, It would have to measure up to 1120 watts per square meter of sunlight which turns out to be 0.112 watts per square centimeter. To do this the sunlight would have to be filtered through a light reducing filter but I do not know how much to reduce it to get it in line with the maximum reading of the device. Any help you could give me here would be greatly appreciated. 

We are interested in doing a similar thing, measuring sunlight, but thinking the OPT3002 might be a better choice because if its wider spectral response.

Any other thoughts on this?

S. Dunbar, AFA Colorado