Hydrofarm par meter

Yodaweed said:

Lux meters measure what the human eye can see, for what the plant uses you need a par meter, those lux meters are useless for growing crops. Not to mention when using a lux meter it only detects specific nm of light and since most LEDs use a larger spectrum it makes them even more useless.

Click to expand...

Most par meters use a silicon photodiode that is tuned to human vision and then filtered...I would say Li-cor might have the best, but their entry meter is $3k or so.....far above what the apogee and Scout cost, which are both filtered to achieve PAR readings.....most lux meters are going to clip the ends of blue and the ends of red in most cases...especially the cheapies....

Abiqua said:

http://rollitup.org/Journal/Entry/arduino-2-1-build-a-diy-lux-meter-for-12.31448/

Click to expand...

How's the weather in Antartica like this part of the year? I always have wanted to spend a summer out there...

Abiqua said:

Most par meters use a silicon photodiode that is tuned to human vision and then filtered...I would say Li-cor might have the best, but their entry meter is $3k or so.....far above what the apogee and Scout cost, which are both filtered to achieve PAR readings.....most lux meters are going to clip the ends of blue and the ends of red in most cases...especially the cheapies....

Click to expand...

I just need the right formula to calculate the PAR from the flux lol

Ironically, PAR meters cannot accurately measure your canopy PAR levels, just a rough guess that cannot be used to compare LED to HPS or even from one LED to another. Lux meters are just about useless when it comes to comparing lights. So for most of us, might as well just get a decent lux meter like RS did and work with that to check light distribution, lumen depreciation, temp droop, reflectors etc.

We can estimate average PPFD by knowing the dissipation, efficiency and canopy size. From time to time we get that data backed up by spectroradiometer. Stardustsailor has done that with Vero29 and CXA3070 and confirmed their output is well above their claimed minimums. And on top of that we should be seeing some impressive gpw results to back it up over time as well.

That said GGs experience is that his PAR meter readings were very close to calibrated spectroradiometer readings. Apogee claims error for LEDs is approx 10% although I am skeptical of that claim. For me, I couldnt justify the expense ($155 for the Apogee PAR sensor) or see the usefulness of a reading from a meter that cannot see above 655nm when our warm whites have decent output in that range

Looks like if you had a warm white there is a lot of error in the deep red range and some in the blue range. If you measure a cool white there is a lot of error in the blue range and some in the deep red range.

Agreed. In other words, lux meter and PAR meter are basically same things with slightly different software. A quality one (either of them) can get you accurate results, toys like the one this thread is about can be used "to check light distribution, lumen depreciation, temp droop, reflectors etc" at best.

Ya the short version, I figure a $13 lux meter is about as useful as the Apogee PAR sensor because either way we are guessing. And since we are guessing might as well just estimate PPFD mathematically. OR, $4-$5Kfor a calibrated spetroradiometer to get the real numbers, solid data points. My 2 cents anyway, but as always to each their own

alesh said:

Agreed. In other words, lux meter and PAR meter are basically same things with slightly different software. A quality one (either of them) can get you accurate results, toys like the one this thread is about can be used "to check light distribution, lumen depreciation, temp droop, reflectors etc" at best.

Click to expand...

I would say the only difference between the hydrofarm and the Apogee is quality of filter and quality of diode...a PAR meter is essentially a lux meter without the the UV and IR cut....I agree the lux meters are skewed because of their UV readings but still most sensors I have researched, have the same dropoff in the Red to Deep Red section..

This person was able to get a $5 dollar photodiode with a uv/ir cut to get within the tolerances exhibited between The Scout and the Apogee.....
http://www.plantedtank.net/forums/showthread.php?t=120109

This person, used an even lower tech lux meter within similar characteristics as the above $5 diode, and might be on to even better results...
http://www.plantedtank.net/forums/showthread.php?t=823705

Lastly, I have recently come across a guy named Forrest G Mims...haven't read much, but he developed a cheap LED photometer in the 70's and still works with them to this day...he has some plans for using an LED photometer as a Par meter, by using the led's as INPUT's instead of outputs, but a spectrometer is needed somehow to check accuracy of your test diodes, that is the only big hurdle I see at the moment.....

SupraSPL said:

Ya the short version, I figure a $13 lux meter is about as useful as the Apogee PAR sensor because either way we are guessing. And since we are guessing might as well just estimate PPFD mathematically. OR, $4-$5Kfor a calibrated spetroradiometer to get the real numbers, solid data points. My 2 cents anyway, but as always to each their own

Click to expand...

Exactly!!!!!!

The Li-cor is what the diehard should be after imho.....but its legitimately $3k on a good day

SupraSPL said:

Ironically, PAR meters cannot accurately measure your canopy PAR levels

Click to expand...

Is it point source that is problem correct...as in if you had a "web" of sensors and averaged the readings would that essentially be PPFD?

I have a pfd from a canopy related study, where they deployed an array of 30 photodiodes throughout a single tree canopy to measure the DLI, same concept?

My complaint is in terms of the sensor's non-flat response to each wavelength. Software would be required to adjust for the sensors imperfections in order to simulate a flat response to every wavelength from 400-700nm.

Once spectral response was flat, the next step would be to analyze the spectral distribution of the light in question and convert its curve into umol/s using the @alesh method (automated software ideally).

Then, it would need to be calibrated to accurately measure PAR W (magnitude). Now we have PAR W and umol/s/W of the curve. Final step would be to extrapolate that into a sq meter for PPFD (and you could take multiple readings to get a better picture of the spread into the canopy).

That is what it would take to get a real PAR reading. I am no spectroradiometer expert, but I believe they do all of the above.



From Konica Minolta:

"The exact CIE V λ curve and CIE colour matching curves are stored in the software and are used to process the data from the measured spectral energy distribution of the light source under test. Hence, the measurement error associated with photometers and filter colorimeters is avoided in spectroradiometers. "

Maybe you guys can clear up a question I have.
When you look at a graph like these - http://www.amaretechnologies.com/ppfd-chart-ss.350-uv-open-reflector.html
You see that the numbers fall off fast. I have read many places that 800? is the peak and best intensity?
If that is so then this light would basically cover a 1ft sq.
Obviously even with hps you can not maintain 800 all the way across the canopy so at what level does plant production become noticeably diminished?

SupraSPL said:

OR, $4-$5Kfor a calibrated spetroradiometer to get the real numbers, solid data points.

Click to expand...

I've been wondering for awhile why we can't use this $30 PublicLabs DIY spectrometer. You connect it to their SpetralWorkbench. Calibrate it to a CFL (some consistent mercury vapor signature present in all CFL bulbs, I believe is their theory.). We could then "calibrate" our graphs relative to each other by taking a graph of noon-time sunlight. It would be a process.

The workbench lets you export the data to CSV for a spreadsheet which can be graphed/smoothed. We could all use the same spreadsheet which would do the relative adjustment using the two exported datum. (A spectral reading of an LED light, and a reading of noon-time sunlight taken with a spectrometer that hasn't changed or been CFL calibrated since that sunlight measurement.).

It would be some work. I'm not volunteering to do it. Just throwing out an idea. I have this spectrometer and would be willing to participate. (I can definitely document some gotchas so we'd all use it the same way.).

SupraSPL said:

Ya the short version, I figure a $13 lux meter is about as useful as the Apogee PAR sensor because either way we are guessing. And since we are guessing might as well just estimate PPFD mathematically. OR, $4-$5Kfor a calibrated spetroradiometer to get the real numbers, solid data points. My 2 cents anyway, but as always to each their own

Click to expand...

So an even shorter version is that my voltage divider with a photocell and a resistor attached to a multimeter or arduino analog in is just as useful as the apogee PAR sensor!! lol.

Good at giving you relative differences in intensity when comparing the same SPD, but impossible to make comparisons of lights with different SPD.

Some of them have different absorbance bands and resistances, but for the most part, the relationship between intensity and R looks like this.

linear plot

log plot


Or use the iphone thing! lol. That's probably also just as "useful". Keep in mind, I don't even use the meter I have laying around anymore. It works.. at finding hot spots....

4 digit 7-seg decoder
2x4 binary decoder
8bit serial in, parallel out shift register



555 timer on the right doing nothing.

churchhaze said:

Some of them have different absorbance bands and resistances, but for the most part, the relationship between intensity and R looks like this.

linear plot

log plot

Or use the iphone thing! lol. That's probably also just as "useful". Keep in mind, I don't even use the meter I have laying around anymore. It works.. at finding hot spots....

Click to expand...

I have a lux app on my phone but want to cross reference it with the meter I just bought

Edit: my typing sucks lol

OneHitDone said:

Maybe you guys can clear up a question I have.
When you look at a graph like these - http://www.amaretechnologies.com/ppfd-chart-ss.350-uv-open-reflector.html
You see that the numbers fall off fast. I have read many places that 800? is the peak and best intensity?
If that is so then this light would basically cover a 1ft sq.
Obviously even with hps you can not maintain 800 all the way across the canopy so at what level does plant production become noticeably diminished?

Click to expand...

I like the sharp cutoff of the 90 degree setup because you have more control of where you want to put the light and less dependence on the location and reflectivity of the walls (diffusion vs airflow). That is especially true in a small canopy where you might not benefit from overlap. From the 90 degree chart I think you would get good results somewhere between 2X2 and 3X3 with the 90 degree lens version.

On the other hand, the 120 degree setup with a shorter distance to canopy might make even more sense and light the canopy from more angles.

SupraSPL said:

Ironically, PAR meters cannot accurately measure your canopy PAR levels, just a rough guess that cannot be used to compare LED to HPS or even from one LED to another. Lux meters are just about useless when it comes to comparing lights. So for most of us, might as well just get a decent lux meter like RS did and work with that to check light distribution, lumen depreciation, temp droop, reflectors etc.

We can estimate average PPFD by knowing the dissipation, efficiency and canopy size. From time to time we get that data backed up by spectroradiometer. Stardustsailor has done that with Vero29 and CXA3070 and confirmed their output is well above their claimed minimums. And on top of that we should be seeing some impressive gpw results to back it up over time as well.

That said GGs experience is that his PAR meter readings were very close to calibrated spectroradiometer readings. Apogee claims error for LEDs is approx 10% although I am skeptical of that claim. For me, I couldnt justify the expense ($155 for the Apogee PAR sensor) or see the usefulness of a reading from a meter that cannot see above 655nm when our warm whites have decent output in that range

Looks like if you had a warm white there is a lot of error in the deep red range and some in the blue range. If you measure a cool white there is a lot of error in the blue range and some in the deep red range.
View attachment 3402241

Click to expand...

That 3000K 90CRI is killer...

SupraSPL said:

My complaint is in terms of the sensor's non-flat response to each wavelength. Software would be required to adjust for the sensors imperfections in order to simulate a flat response to every wavelength from 400-700nm.

Once spectral response was flat, the next step would be to analyze the spectral distribution of the light in question and convert its curve into umol/s using the @alesh method (automated software ideally).

Then, it would need to be calibrated to accurately measure PAR W (magnitude). Now we have PAR W and umol/s/W of the curve. Final step would be to extrapolate that into a sq meter for PPFD (and you could take multiple readings to get a better picture of the spread into the canopy).

That is what it would take to get a real PAR reading. I am no spectroradiometer expert, but I believe they do all of the above.

View attachment 3402279

From Konica Minolta:

"The exact CIE V λ curve and CIE colour matching curves are stored in the software and are used to process the data from the measured spectral energy distribution of the light source under test. Hence, the measurement error associated with photometers and filter colorimeters is avoided in spectroradiometers. "

Click to expand...

Apogee corrects/estimates for the lack of their sensor doing it. Though the responses are low at area's of the curve...they are there, and can be/are corrected for. If you give apogee the SPD of the spectrum you are measuring...they will calibrate it to that perfectly. The sensor may not be ideal...but the meter as a whole unit...is actually very accurate. Put it up to a licore anyday...I have done it on at least 10 occasions over the past 2+yrs with LED's, induction, hps, mh, cmh, T5, and none have had issues with consistency or accuracy between them within the margin of error. I use the apogee in my room daily...it is very consistent.

One very important occasion was the illumitex measuring...basically pure 660nm and 450nm...the so called weak area's of apogees sensor. In this I got to use the Apogee(stock), Apogee(AT calibrated), Li-Core, Spectroradiometer. All within margins of error or better. If I remember correctly in was ~4%+/- between the hand meters. And the spectrometer was higher than all the hand meters. So I can tell you that it is not over correcting in most cases.
I used the same gang of meters to test the drone(cxa 3K's) on that same day...with the same error between them pretty much.

My focus when using many meters for testing is not to test meters against each other...it's to know what light my canopy is getting...but next time I will focus on the meters themselves and record the difference in detail to really clear this up. But I have actual experience with all the meters in question, and they are all very reliable, accurate, and useful.

Spectroradiometers will not tell you the PPF unless you are using it with an integrated sphere. Otherwise...it's just a fancy/accurate par meter(ppfd, one point of the space...not all of the space).
Estimating ppfd based on components and conditions is no more accurate than an apogee...probably much less considering the variables left out in that method. Just because the source emits it...doesn't mean it's making it to the canopy. Which is the whole reason for these meters indoors...lab vs field measurements. Otherwise we could just go off bulb/light specs, which do come from spectroradiometers and spheres.

Good link for errors and apogee
http://www.apogeeinstruments.com/content/Quantum-Sensors-LEDs-Downing-College-September-2012.pdf

Good data, incredible to me that the results weren't all over the map. The way I see it there is a big difference between guessing and estimating. If the meter cannot see in the ranges I highlighted the readings should be useless for quantitative data, but if we are estimating using the full range of imperfect data Id call that an estimation. I assume the Apogee response chart that I posted is after their correction.

I agree, don't want to mislead anyone to thinking spectroradiometer will measure PPF, only PPFD.