The Planted Tank Forum banner

Assembling a DIY PAR Meter

12K views 14 replies 7 participants last post by  jeffkrol 
#1 ·
I just posted, in the Lighting forum, a discussion about what it takes to make a PAR meter that will rival the Apogee Quantum PAR meter for accuracy, at a much lower cost. This is the details about actually making one.

A PAR meter consists of a sensor, a silicon photodiode, some filters to adjust the spectral response of the diode, a diffuser, and a readout module, which powers the sensor and displays the reading.

See http://www.plantedtank.net/forums/showthread.php?t=443073 for a discussion about the sensor requirements.

Digital readout modules can be DIY’ed, which is beyond my capabilities, or you can use the readout module from a cheap lux meter purchased from Ebay or Amazon vendors. Those lux meters use silicon photodiodes as their sensors, so we can substitute a different silicon photodiode and the readout module will work fine, if we do it right.

Most of the lux meters sold on those two sites are one of two designs: one is given a part number of LX1010B, which is manufactured at more than one factory in China, so there are small differences among the various ones being sold. But, all of them I have tried have been usable, even though they are not exactly alike. The other one is given a part number of HS1010A. This one is more difficult to adapt to a PAR meter, because it uses 5 push button switches to select various modes of operation and various ranges, instead of simple slide switches. I have yet to find a way to lock the switches into the appropriate positions needed for it to readout PAR.

The LX1010B lux meters cost from about $8 to over $100, apparently depending on whims of the seller. The ones I have purchased and modified have cost me from $8 to $20. There are some quality differences among them, but nothing of significance that I have found.

To use a LX1010B readout module, I just disassemble the sensor and cut off the electric cable at the diode, to get the maximum length of cable attached to the readout module. Before doing this, always install the battery and turn on the lux meter to see if it works. I have had a few that didn’t work - one had a segment of the display that wouldn’t light up, a few had the display show a constantly changing number instead of reading zero, one had a broken wire in the battery connector, and one had corrosion damage internally on the circuit board. Just remember, these are CHEAP!

The sensor for a PAR meter needs to be waterproof, so we can take readings in the tank, as well as out in the air. The sensor needs to be small so it can be easily moved around in the tank. It needs to be made so it can be held without our hand or arm getting in the way of the light and changing the reading. And, it should be made so one of the three range settings on the lux meter readout will give a number that equals the PAR we are reading.

PAR meters use a diffuser for two reasons: one is to give the diode a small light source that is at a fixed distance from the diode. The other is to make the sensor relatively insensitive to small deviations from the ideal pointing direction, so light off center to the sensor still gives about the same reading as perfectly aligned light. (That is why they are called “cosine diffusers” - the cosine of an angle near 90 degrees changes only slightly with small changes of the angle.) A small diameter piece of frosted plastic works fine for this purpose.

The housing for the parts of the sensor needs to be able to hold the diode, the filters, and the diffuser in specific positions, with the distance between the diffuser, and the alignment of the diode, diffuser and filters being easy to maintain. It also has to be easily made waterproof, with very little air trapped inside. Finally, the hard part is that is should be possible to adjust the distance between the diffuser and the diode after assembly, to adjust the sensitivity of the sensor and make the readout be accurate.

The best, cheapest, and easiest material I have found for making a PAR sensor is clear acrylic tubing, and for the diffuser, frosted acrylic rod. TAP Plastics has both types of acrylic at reasonable prices, and the tubing is available from at least one vendor on Ebay in small lots. Acrylic is also a good choice because it is easy and fast to glue it together with waterproof bonds, using cement that is viscous enough to form filets at the joints, and to bridge big gaps between pieces if necessary. It’s only disadvantages are its fragility, and the fact that there is no opaque black acrylic tubing available, which forces us to use black acrylic nail polish to make the housing opaque.



The current design I am using for PAR meter sensors is:

The most critical dimensions are the 7/32” long frosted 1/4” dia acrylic rod, for the diffuser, the 1/4” long 3/8” diameter tube that holds the diffuser, so there is a small gap between the end of the frosted rod and the end of the 3/8” diameter tube, and the 15/32” long 1/2” diameter tube, which controls the distance from the end of the diffuser to the photodiode. The top of the photodiode should be even with the low side of the bottom 1/2” diameter tube. One problem is that the 3/8” diameter tube, which should fit inside the 1/2” diameter tube, doesn’t do so. You have to use fine sandpaper to slightly reduce the outside diameter of the 3/8” tube to make it fit. If you try to force it, the 1/2” diameter tube just cracks - it is very brittle.

The two pieces of 1/2” diameter tube are cut 80 degrees off the tube axis, to form two 10 degree “ramps” so you can twist the two together to adjust the distance between the photodiode face and the diffuser for adjusting the sensitivity of the sensor.

It is easiest to assemble the diffuser part if you first install the 5/8” diameter sleeve on the top 1/2” diameter tube, with that sleeve overlapping the short side of the 1/2” diameter tube by about 3/32”, cementing it into place. Then, install the 3/8” diameter diffuser holding sleeve inside the top of the 1/2” diameter tube, cementing it in place. When the cement dries, paint the outside of the 1/2” diameter tube and the top of the assembly with two coats of black nail polish. When the nail polish dries, use a knife to scrape any nail polish out of the inside diameter of the 3/8” tube, and cement in the 1/4” diameter frosted acrylic diffuser. This prevents getting nail polish on top of the diffuser.

To assemble the bottom, photodiode holding assembly, first put the photodiode leads through the thin 3/8” diameter sleeve, bend them outward to hold the diode in place, then bend the outer part or each lead back towards the center to reduce the distance between them. Poke the end of a 3 foot long piece of electric cable, two conductor cable, through the hole in the side of the bottom 1/2” diameter housing, and work the end up and out the top of that housing by about a half inch or more. Then, strip the wires back about 1/4 inch, and tin the bare wires. Solder the diode leads, the “+” marked lead to the red wire, and the other lead to the black wire, making sure the diode is in the right position to be where it needs to be when you work the cable back into the housing to get the diode and sleeve in place into the housing. Once you get the knack for doing this it is easy.

Work the cable back so the diode fits where it needs to be, with the top face of the diode about even with the low side of the housing, and with the thin acrylic sleeve under the diode aligned at 90 degrees to the axis of the housing, so the diode faces up and isn’t tilted to one side. Put a big drop of acrylic cement on the cable to hole joint inside the bottom housing, and another drop so it runs down inside to cement the inside sleeve into place. You can make slight adjustments when the cement is almost dry. After the cement dries, cement the bottom of the lower housing onto a 3/4” x 5” strip of thin acrylic sheet, at one end. I squirt a big dollop of cement onto the sleeve, and hold the housing in place on the dollop of cement for 3 minutes before releasing it. This is now a good time to cement a stub of 3/8” dia or 1/2” dia acrylic tube at the other end of the strip, to act as a socket for a wand that can be used to hold the sensor where you want to measure the PAR. Add another big drop of acrylic cement on the outside joint between the electric cable and the hole in the housing to seal the cable to the housing.

Solder the sensor cable leads to the lux meter readout cable, keeping red to red and black to black. Use shrink tubing to insulate the connections. I use three layers, one over each wire joint, one over the whole cable, holding the ground wire, which I have left one inch longer than the two active wires, against the insulation of the readout cable, to help reinforce the joint mechanically, and one over the whole joint to finish it off. At this point turn on the readout to verify that it works - it will read something other than zero, and the reading will stay steady.

Cut out a 5/16” octagon of each of the filter materials, being very careful with the coating on the #1995 infra red blocking filter - don’t touch it or rub it against anything. Put the #1995 filter in place, with the coated side towards the diffuser - that is the reflective side, and you can identify it by the reflected color of that side. Lay the other two filter pieces on top of the #1995 filter, and install the filter retaining sleeve. It works best to use a piece of 3/8” diameter tube, slotted on one side, as a tool to shove the retaining sleeve in place.

Temporarily assemble the two halves together, and use black electricians plastic tape to hold them in place and block any light from entering the side of the assembly. Put the assembly into light of a known PAR and adjust the height of the top half as needed to get the right reading. You may need to add another filter, one of the Roscolux diffuser filter materials, to further reduce the reading. #114 reduces the reading about 10%, #162 reduces it less than 10%, and #163 reduces it about 20%.

When the assembly gives the right reading, carefully cement the two halves together, maintaining the length adjustment. Then paint the rest of the outside of the assembly with 2 coats of black nail polish.

I find that this assembly reads correctly on the middle range position on the readout. I cement the range selector switch into that position to avoid errors from selecting the wrong range.
 
See less See more
2
#2 ·
This is really cool, and very impressive. I think it's well within my comfort zone for DIY stuff, and would be fun to build. However, I'm really glad that I got one from you because you were able to calibrate it with a real PAR meter. That's the questionable part to me. Sure, we could pick a known light source that should read a certain PAR at a certain distance, but you're assuming very little or no variability in that light source, if you are shooting for relatively precise measurements.

That said, perhaps that amount of variability doesn't matter for the purposes of how we use it on our own tanks. I mean, if it was off by 10%, that would only be off by 10 PAR at 100 PAR reading. And for what we do, that's not a big difference in real world terms. It would seem that the more important thing would be if, say, you were measuring plant growth in your own tank or trying to change lighting due to algae, that adjusting the lights up or down to change our PAR reading by 25% or whatever, that you have a known starting point, and a way to measure from that point, and can get back to it, or adjust it further.

In other words, it doesn't matter so much that I am getting precisely 70 PAR at the substrate so much as if I want to lower my lighting, I can lower it by a certain measurable percentage, observe changes, and make further adjustments. So maybe it's not too important to get hung up on being incredibly precise, but rather simply trying to get as close as possible, and that's good enough and super useful for hobby use.

Thoughts?
 
#3 ·
I would trust my PAR meters to be accurate to +/-10%, but not much better, while the Apogee Quantum PAR meter is claimed to be accurate to +/-4%. But, it is very hard to measure PAR for an aquarium and do it well enough to trust that you are within +/-10%, no matter what PAR meter you use. Small errors in how you make the measurements make a noticeable difference in PAR readings. Those can be errors caused by light reflecting off your arm, by light reflecting off a close white wall, by small errors in how you point the PAR sensor, by errors in placement of the sensor, and by the fact that some types of lighting take up to 15 minutes to stabilize. (My PC light starts at about 50% of its maximum PAR, builds up to maximum reading in about 5 minutes, the drops back a bit over the next 5-10 minutes. If I take a reading right after turning the light on, then take other readings as I adjust things like height of the light, then go back to the first reading, I will get different readings, and by more than 10% different too.)

I think we get 90% of the benefit of a PAR meter with only +/-10% accuracy compared to 100% of the benefit, with +/-4%. And we pay a lot to get that added accuracy.
 
#5 ·
What would the possibility be of someone assembling and selling the sensor units which then could be attached to our own lux meter? Like most people I believe on this one, I would find acquiring the parts and putting everything together a little overwhelming. Yet making 10 would be litter harder than making one.
My guess they would sell like hotcakes at the right price. Sign me up.
 
#6 ·
Theoretically you could make just the sensors and let people splice them to their luxmeters. The problem with that is the difficulty in making all of the sensors exactly alike. That takes more accuracy in the dimensions than can be achieved with this design. Also, there is no guarantee that the several versions of that lux meter would all read the same with the same replacement sensor. http://www.plantedtank.net/forums/showthread.php?t=453225&highlight=par+meter is one approach that would stand a much better chance of being a true standardized sensor. If I could teach myself to do CAD drawings I would play around with this idea. I once knew how to use the very sophisticated CAD system Boeing uses for their commercial jet airplane design, but I have forgotten about all I learned about it.
 
#7 ·
I used to own a CNC shop for 20 plus years. I did most the CAD drawings myself. I also have a friend who teaches CAD/CAM and I have access to his equipment. I would consider this an interesting project to cut out a bunch of these. However I'm currently starting a new career as a teacher so I have to be careful with time.
 
#9 · (Edited)
I don't see a good way too make it a larger picture, but I can make it a pdf file, so I will PM you that. (If I can again figure out how to upload an attachment????) EDIT: I found the way to do this - the paperclip in the header.
View attachment NanoPARsensor7.pdf
If I wanted to rely on a "standard light" for adjusting the sensors to a known PAR the only one I can think of is a Finnex Fugeray or Ray2. I did measure the PAR for one of them and my readings agreed with the data on http://www.plantedtank.net/forums/showthread.php?t=189944 I suspect you could get one of the shortest ones and use it as a "standard".
 
#12 · (Edited)
TEchnically you can get data .. as to its "fit" to a PAR meter.. well that is a bit tricky..

Each photodiode has a sensitivity chart..
Like this:


Purpose of the filters ect is to 'cut" out unnecessary light gathering such as all photons > 700nm (cutoff def of PPFD) or less than 400nm
A ir/uv cut filter is about minimum.

After you make the output in range you now need to equalize the curve to avoid mostly under sampling..
Now that said, using LEd's most of the "cut" work is already done due to narrow bandwidths of most diodes.

Even "whites' produce mostly within the range w/ little outside tha PAR range (some red phosphors added to emit past the 700nm though ( IR diodes do though)..few will go much past 400nm except .
violets or true UV

NOTE: little >700 or <400 w/ this white LED but also note lots of blue which w/ a raw photodiode will be under-sampled..

But you still need to deal w/ the lack of blue sensitivity in most photo-diodes..Can mean little or a lot different than a true PAR meter..
don't know of anyone who quantified the error and it will vary by light and light type..

with magenta hort. lights (ratio of royal blue/deep red diodes most common)
you will under-sample blue the most BUT could possibly correct this a bit mathematically

Final point is it "may' be possible to get close but highly unlikely..
 
#14 ·
I think you would be lucky to get 30% accuracy out of it. The Graph jeff posted shows that that sensor is mostly sensitive to red light and much less sensitive to blue. This can cause a sensor to be strongly affected by the color temptation (Kelvin rating) of light. This can result in your proposed par meter giving a good par reading to a 2700K butb and and lower reading for 6000K bulb. In reality both bulb could have the same par output or the 6000K bulb might actually have a higher PAR reading.

The reason PAR sensors are so expensive is that they need basicallyy 5 filters to get good occuracy. A uv and infrared blocking filters and a red, green, and but filter to compelsate for the sensitivity for the photodiode used by the manufacture. Note manufactures will sell you just the señor without the digital meter. This is because you can use a simple volt meter to measure the output. You would then use a calculator to convert volts to par. based on the infomation provided by the manufacture. However using a volt meter will not save you a lot of money because most of the cost is in the filters.
 
#15 · (Edited)
you could buy a $20 Lux meter and remove the filter pack.. Well mostly a green colored filter...
you can also convert LUX to PAR .. Divide by 67-70

You will most likely be closer than w/ a home made using a RAW diode..
There are back illuminated photodiodes that have increased blue sensitivity.. but usually not very cheap..

Project very similar to what I was, at one time, contemplating..
https://hackaday.io/project/12125/logs


If you want to make a DIY PAr meter I'd suggest finding 1)blue enhnanced photodiodes 2)PTFE diffuser 3)B&H highpass low pass filter and cut to size (cheapest best)
Baader makes the best.. and one Krean company.. both difficult to obtain or too expensive.
https://www.optcorp.com/ba-fuvir-1-uv-ir-cut-filter-1-25inch.html
Surprised on the "accuracy" of the eek bay one above.. Usually cut off in the red is to far to the right ala old Apogee sensor.
https://www.amazon.com/Excelitas-Technologies-Sensors-VTB4051H-photodiode/dp/B01NBQW5J7

done a lot of spec readings and Excelitas has the cheapest and most linear sensor in the vis. range.
2.5 to 4.25 in the vis range..

Still working on a "simple" colored filter to equalize this..
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top