New DIY LED light

VegasWinner

Well-Known Member
Finished up my new light and hung it up tonight. It consists of ten Samsung Acuity hard strips, 56 LEDs in double rows on 22" x 3/4" strip boards. I mounted two strips to a 48" long 3/4" aluminum U-channel, for a total of five light bars. Framed them together with 3/4 aluminum angle, for an 18" x 48" light. Powered by a 24 Vdc 15 Amp power supply. currently running at 1 amp per strip for 240W total.

View attachment 3962384 View attachment 3962385

The light in action
View attachment 3962386 View attachment 3962387
They are bright. Nice work
 

Big Green Thumb

Well-Known Member
I was referring to the meanwell 24v constant voltage power supplies I found online. There are several more efficient meanwell options but they are all exponentially more expensive. Right now I have 3 24v light setups capable of 360 watts each. I am driving them with 3 Chinese $20 power supplies. I also have some 33v 48" strips being driven by a 36v version of Chinese power supply.
https://www.rollitup.org/t/can-700-watts-of-led-strips-grow-trees.946736/page-16
 

GrowLightResearch

Well-Known Member
I was referring to the meanwell 24v constant voltage power supplies
Me too. Yes, the HEP is more expensive than the HLG. Although IMHO, LEDs are best driven in series by CC drivers. Have you ever measured the current in the individual strips?

Your grow and lighting setup is very impressive. It appears you could drop the strips closer to the canopy if you had a PAR meter to use while adjusting the current. You could probably save quite a bit of electricity by doing so.
 

Dave455

Well-Known Member
Me too. Yes, the HEP is more expensive than the HLG. Although IMHO, LEDs are best driven in series by CC drivers. Have you ever measured the current in the individual strips?

Your grow and lighting setup is very impressive. It appears you could drop the strips closer to the canopy if you had a PAR meter to use while adjusting the current. You could probably save quite a bit of electricity by doing so.
Whats good par meter to use ?
 

GrowLightResearch

Well-Known Member
Whats good par meter to use ?
First off you do NOT want a Lux meter. Lux and lumens are of very limited or no value in horticulture.

________________________________________________________________


Li-Cor is a higher end that gives measurements for each individual wavelength. best of all the calibration is traceable to the U.S. National Institute of Standards and Technology (NIST). The bad news is it costs over $1000.
LI-250A Light Meter, LI-190R-BNC-2 Quantum Sensor, and 2003S Mounting and Leveling Fixture $1195.
LI-250A Light Meter.jpg


________________________________________________________________

A very popular meter is the Apogee MQ-200. It measures the the sum of the number of photons from 400nm to 700nm and it reads out in one number. It can store measurements that are downloadable. The sensor is not too big and is waterproof. $350
Apogee MQ-200 Quantum Meter.jpg


________________________________________________________________


The best value is the Hydrofarm LGBQM Quantum PAR Meter $169 ($105 at horticulturesource.com)

Hydrofarm LGBQM Quantum PAR Meter.jpg
________________________________________________________________


Reviews: http://ledgrowlightguide.com/best-par-meter-reviews
 

wietefras

Well-Known Member
First off you do NOT want a Lux meter. Lux and lumens are of very limited or no value in horticulture.
Depends on what you want to do. If you only want to check light distribution uniformity then a lux meter will do just fine.

Which means that for say 99% of the applications we would normally encounter, a lux meter is enough. Divide Lux by a factor of around 70 (or look up the actual value for your light source) to get an idea of the PAR value.

Cheap PAR meters (ie < $1000) are woefully inaccurate and plants don't care if you are a bit off anyway.
 

canadian1969

Well-Known Member
Depends on what you want to do. If you only want to check light distribution uniformity then a lux meter will do just fine.

Which means that for say 99% of the applications we would normally encounter, a lux meter is enough. Divide Lux by a factor of around 70 (or look up the actual value for your light source) to get an idea of the PAR value.

Cheap PAR meters (ie < $1000) are woefully inaccurate and plants don't care if you are a bit off anyway.
There are comparisons of the hydrofarm and Licors on youtube,

they are pretty close, good enough close imo. I use a Lux meter, its fine as you say, there are also conversion calculators online for Lux to PPFD which get you close estimates.
 

wietefras

Well-Known Member
If you had read the original post the purpose was to measure the quantum flux in order to adjust the height as close to the canopy as possible. That cannot be done with a Lux Meter.
The correct height is determined by the uniformity. Which you can check with a lux meter.

And if you want to know PPF you can estimate it by using a lumen to par conversion factor. That's basically what those cheaper PAR meters do anyway.
 

GrowLightResearch

Well-Known Member
The correct height is determined by the uniformity
Maybe, if you have a really shitty fixture. Using half a dozen or more evenly spaced Samsung 4' strips in a 4'x4' area is going to get very good uniformity if not powered in parallel. They can be lowered to inches above the canopy. How many inches requires a PAR meter or spectrometer.

And if you want to know PPF you can estimate it by using a lumen to par conversion factor.
Actually with that calculation he just posted you can. LOL
You can but an estimate is just that, only a worthless estimate. So no.

That's basically what those cheaper PAR meters do anyway.
No. Lux is estimated from ambient light sensors (e.g. photo diode) , not the other way around. They either use an optical filter before the sensor or have an on-chip Lux conversion table after the sensor.

A PAR meter cannot use the same circuitry as a Lux meter. A 550nm photodiode is all that is need to make a cheap Lux meter. Here is a schematic for a Lux circuit using less than $5 in parts. The key part being a 550nm Photodiode.

photodiodeLuxMeterSchmatic.jpg






Cheap PAR meters (ie < $1000) are woefully inaccurate and plants don't care if you are a bit off anyway.
Not true! You can get a very accurate spectrometer well under $500.

Here is a $220 Mini-Spectrometer Head with very accurate readings from 340nm to 650nm. 256 pixel CMOS linear image sensor with 15nm resolution. It needs to be mounted on a PCB with a connector for its serial port. Can be connected to a $10-15 Arduino.

HamamatsuMiniSpectrometer.jpg



There are also Ambient Light Sensors (ALS) using photodiode arrays for simple and fairly accurate measurements.

I am betting I can make a highly accurate PAR measurement tool for under $100.
______________________________________________________________________

Typical 550nm Photodiode SPD curve

photoDiodeSPD.jpg


_______________________________________________________________________________

Ambient Light Sensors

The above photodiode circuit can be reduced to a single chip, called an Ambient Light Sensor (ALS) costing less than $1. The above photodiode response is shaped to the human eye response curve either an optical filter, logic tables, or computational conversions.

AbmientLightSensor.jpg

__________________________________________________________________________________________

Ambient Light Sensor with Lux Optical Filter

The optical filter shapes the photodiode response to match human eye response.
AbmientLuxSensor.jpg


_________________________________________________________________________________________


There is NO way to use a Lux circuit to measure PAR.

A PAR circuit requires multiple photodiodes with a variety of wavelengths.

An inexpensive PAR circuit will use an ALS with an RGB or RGBW photodioode array.

spectrialPowerDistributionDatasheetWhiteRGB.jpg




_____________________________________________________________________________

Red, Blue, Green, and White sensors can make a very decent PAR meter.
Or an excellent Lux meter with better than 10% accuracy.

ALS
= Ambient Light Sensor

AbmientLightRGBWSensor.jpg


____________________________________________________________________________________

After that you are into spectrometer territory.


A spectrometer diffuses (with prism or diffusion grating) the light and measures from UV to IR and beyond.

spectrometerBlockDiagram.jpg




Where the detector is typically a Photodiode or CMOS Linear Array sensor like what is used in a scanner or copier with 128-2048 sensors.

The higher end spectrometers will use a CCD sensor.

There is a very inexpensive way to use a camera (e.g. web cam or mobile phone) to capture the diffused (prism) output. Then analyze the image and convert it to an SPD chart.

This is a Kickstart funded ($110,000 pledged) Public Lab project that sells a cardboard spectrometer that uses a piece of a DVD as the grated diffuser and a web cam as the CCD sensor. Sells for $47.


publicLabSpectometerSoftware.jpg






________________________________________________________________________________________


I am currently working on a design using a six color photodiode array using violet, blue, green, yellow, orange, and red. It will be a very inexpensive, but fairly accurate, for the measurement of PAR positioned somewhere between a PAR meter and spectrometer.

The cost would be less than a PAR meter. About $50-$100. Basically you use a PC or laptop to read the six color values from the sensors and use software to calculate the PAR. Each color is very easily calibrated.





spectrialDistribution6color.jpg

 
Last edited:

nfhiggs

Well-Known Member
You can but an estimate is just that, only a worthless estimate. So no.
An estimate that is accurate to within 10% is perfectly acceptable and useful. So YES it is. I've got a shit ton of test equipment that has accuracy specs of +/- 10% that the FAA says is perfectly acceptable for testing avionics with.

If my target is 1000 uMols of flux density at the canopy, the plant is not going to give a rats ass if its really 900 or 1100.
 

GrowLightResearch

Well-Known Member
An estimate that is accurate to within 10% is perfectly acceptable
If you think you can get within 10%, just dream on. You cannot. Your cheap Lux meter will not even be in the ball park. A top notch Lux meter, at best, will be within 10%. And there is NO conversion to a Lux meter reading to quantum PPFD.
 
Top