AS7262 Visible Spectral Sensor, viable for low-cost PAR sensor?

Hello.

I have been trying to find a way to measure PAR without a Quantum sensor, there are those whom say that you can use a LUX-meter and simply convert the measurement to PAR by multiplying it with some value, but the problem then is the spectrum of the lux-sensor.

I have found a interesting sensor called AS7262 which is described as a "6-channel visible spectral_ID device with electronic shutter and smart interface", I don't know much if anything at all about these things but look at the spectrum of this sensor:

I have added a box showing the PAR spectrum as it is defined.
Compared to all the other cheap solutions I have found this looks to be the best, although I haven't been able to work out what this sensor actually measures. It is starting to look to me as this sensor is meant to be used with a light source, a white LED that shines light onto an object and then this sensor measures which wavelengths that return from that object.
But even if that's the case then another sensor might be able to help out to tell the light intensity and then...

I'm not sure what I am getting at here but do you know if this idea is at all viable or am I barking up the wrong tree here?

Cheers

First I thought that if I knew the output from my COB LED then I might be able to calculate the PAR, but now I realize that this sensor appears to be simply a device that tells you how much of each color there are while saying nothing at all about the amount of light in those wavelengths, which does absolutely no good at all.

This is a bust idea, sorry for wasting your time.

That is strictly to show you what intensities across the visible spectrum a particular light source boasts. Within close to UV blue to Far Red. It could give you a CCT value too. But not par.

I have been looking at light sensor, this is a brand new idea of mine so I haven't yet thought it all through but for a kind of low cost you could perhaps use between 4 - 8 different sensors targeting wavelengths that combined would give values that could be used to calculate a PAR value that's more reliable than the converted value of a lux-meter with completely unknown spectrum characteristics?

Or do I really not understand how to acquire a PAR value at all?

Airwalker16 said:

That is strictly to show you what intensities across the visible spectrum a particular light source boasts. Within close to UV blue to Far Red. It could give you a CCT value too. But not par.

Click to expand...

All that really requires is calibration against a known source.

It's not going to be a true par meter on it's own since it's not senstive to the whole "square" PAR spd, but you could create a conversion factor based on the spd of the light and the spectrum of this sensor (sensors).

If you calibrate it against a known light source (on of those Tungsten test bulb perhaps) you should be able to get more accurate conversions than with a lux meter (with unknown sepctrum)

It all sounds like a lot of work and cost though.

I have a lux meter, I drilled a hole in the cap. I run 4 cobs, 2-3500K, 2-4000K

with COBS, the conversion is Lux divided by 70 for 3500K and 60 for 4,000K..

Since I use both, I divide by 65.

So, at 12 inches, I get 80,000Lux or 1230PAR aroundish..

Then, I take my meter, with the drilled hole in the center of the cap, 3/8in hole. Then i find a fixed area that i took the LUX measurement, and take scotch tape and tape over the hole, till you hit your LUX conversion number.

Might take 3-4 pieces. Before putting the tape over the hole, fingerprint the tape, to make a dull look to it,

Its not perfect PAR, but it can give you a general idea. Ive done this with a few of my lights with known PAR values, and its right on the mark.

Then, everyweek, throw a few bucks in a can till you can afford a real Meter.

InTheValley said:

with COBS, the conversion is Lux divided by 70 for 3500K and 60 for 4,000K..

Click to expand...

Higher CCTs tend towards HIGHER conversion factors, not lower. With Samsung 80 CRI, for example, 3000K is 69. 3500 K is 71.1, 4000K is 71.8, 5000K is 73.3. This is because they generally have more green, and lux is very green-centric - so you need to have a larger divisor to compensate.

InTheValley said:

Then, every week, throw a few bucks in a can till you can afford a real Meter.

Click to expand...

Word. You can get a Hydrofarm PAR meter for 100 bucks, and I just came across a wireless spectrometer for $399 from PASCO.

nfhiggs said:

Higher CCTs tend towards HIGHER conversion factors, not lower. With Samsung 80 CRI, for example, 3000K is 69. 3500 K is 71.1, 4000K is 71.8, 5000K is 73.3. This is because they generally have more green, and lux is very green-centric - so you need to have a larger divisor to compensate.

Click to expand...

Yeap, my goof, I think it was randomblame that posted these in another thread, cant remember. I just know, my conversion is 65, per His information.

Mine are actually 2x3500K90CRI and 2x4000K80CRI,

all i know, is these bitches rock, and grow huge nugs, for 225watts average im using.

InTheValley said:

Mine are actually 2x3500K90CRI and 2x4000K80CRI,

Click to expand...

Ahh. The CRI difference might be it.