Building a "light box" for measuring lumen output

I used gloss paint to seal the cardboard. I used 4 coats to finish the tin. I taped a piece of cardboard over the light port just to keep dust out - it gets folded out of the way when I'm using the box. For datalogging I installed a small solar panel whose current output in microamps I measure in my runtime plots.

The box. Amazon packaging is excellent for providing self-adhesive flaps.

Light port cover folded back.

Meter port and cover. There is a small solar cell inside the box which I use for runtimes, the leads go to a datalogging meter.

Cover folded back. Match - this is modding! - it isn't proper modding if duct tape isn't involved :)

The baffles probably aren't strictly necessary but they should mean that any mis-positioning of the light shouldn't affect the results.

The box contained a RAID controller and is about 18cm deep, 30cm wide and 45cm long. Essentially it was the best quality, thickest cardboard box I had in the house at the time.

A larger box probably gives more consistent results than a smaller one, but is much less convenient for those who have things like wives that care about tidiness. I'd reckon that baffles are more important in smaller boxes than larger ones.

LOL! You've got me there.... :bigsmile: :bigsmile:

Budgeteer, coincidently enough I'm in the process of building my own "lightbox" of sorts. I'm taking a different approach however (I know...there's a shocker...). Too many projects, too little time...but give me a few days and I should have a post up on it.


You need at least one between the light and the light meter....

I've been trying to plan one myself. I've got my box, I'm just trying to figure out how to put baffles to diffuse the light enough so that the beam shape doesn't affect the readings.

Easy enough with two. You just attach one to cover more than half of the width (and the full height) of the box to one side and the other of the same size as the first to the other side of the box. Then only reflected light can get to the meter. Regardless of beam width, it still has to be reflected at least once. I put mine about 50mm either side of the centre of the box.

Not if the box is bigger than the beam angle of the widest angle beam you intend to measure - the baffles just make sure. In a sphere you only need one - but if you only measure narrow beams it isn't essential - it is just essential if the edges of the beam can hit the meter port without being reflected. I don't know why, but the meter port is always at 90o to the input port. There is almost certainly a good reason, I just don't know what it is.

With a large enough sphere, a baffle would not be necessary.

There aren't any in the typical colour head of a photographic enlarger which definitely requires integration. But then the beam can be controlled prior to entry to the integrator. The sphere in Wallner heads for Leitz enlargers is less than 150mm in diameter and has no baffle. And Leitz do optics better than anyone else. I used to sell the things. At a price that would have bought two new cars.

m.... The guys on CPF, who did them after lab spheres, put the light directly across the meter, I think. So even though they have 24 inch spheres, they still need a baffle.

If the direct beam contacts the meter, the beam has not been integrated. The lab spheres have baffles built into them or the results would be invalid.

The box I'm using is small giving consistent measures only in candle mode or bare stars for the reflection issues. I set the light meter to give direct reading in lumen , to check emitters is great...The sensor stay in one side of the box near the window at an angle of 90 with the beam which go through some kind of tunnel long half the deep...

From what I've seen and read (thank you Don) the things I'm focusing on are:

1. DO NOT let the beam shine on the sensor

2. Diffusion

I think I have a pretty good idea of how I'm going to build mine. I've been thinking about it for months and it only took a few posts here to help me build it in my mind. :) I suppose one could fiddle around forever trying to get it perfect but seeing as thats imposible, I'm just gonna wing it!

EDIT: Here my first try at the design. I'd appreciate input if you have any ideas for improvement.

Budgeteer, I wish you good luck with your light box.
When you make a measurement on a light you measure of course the lux level (L) in the box and need to know the calibration factor K in:
lumens = K * L
It is in fact quite easy to find the factor. You need some good parallel light, for instance a thrower with a uniform hot spot and another luxmeter.
Calculate the area (A) in square meter of your light input port in the box (easy if you made the hole round, square or rectangular).
Tilt the light box so that the thrower can light directly into the opening from a distance of 5 to 8 meter. Align it so that the hole lies in the uniform part of the hot spot.
Note the box level (L) and measure the intensity (Lin) just in front of the hole with another luxmeter. Repeat the two readings as the thrower are likely to sag in its output.
And you're done...
You have just put (Lin * A) lumens into the box and got a reading of (L), so
Lin * A = K * L
You find your calibration factor as
K = Lin * A / L

The accuracy depends on the second luxmeter and the calculation of the opening. Also cut the other light in the room and beware of reflections coming into the box, push the box to the edge of the table or what is supporting it.

Example: In my own light box the input hole is round and 48 mm across. The area is A = PI * 0.024 * 0.024 = 0.00181 sq.meters.
When I put 110500 lux into it (using the sun as thrower) I read 20000 lux in the box so my K is 110500 * 0.00181 / 20000 = 0.0100 so when I set the build-in luxmeter in position x100 I have readings directly in lumens. To obtain that I removed the opal diffuser on the meter head and fiddled with the position and size of the baffle in the box.
I'm now in the process of making a metal integrating sphere which is nearly finished but the paintwork is stalling me...

My method maybe is not that precise but is quite simple.A rectangular hole in one side of the box where the sensor can slide holded with a rubber band, an ITP A3 as light source. When the meter in X1 scale show 84 lux glue a cardboard top to keep the position, done.

Wow thanks for the tips. I will try the harder approach hoping that my lumen's readings are as close to real lumens as possible. Have to borrow another luxmeter. I have a friend that is a professional photographer that should have a decent one.

I was bored and tried a 5min work lightmeter setup. Made it really cheasy with plenty duct tape and got out the following:

C3 SS nimh 120lm

XTAR D01 XP-G R5 314lm

KD XML @3.33A 810lm and sag to 660 in about 1min.

I do not know how off are those results but the outcome seems very plausible. Parhaps about 5-10% on the lower side if anything.

"Calibration" was made with a "trustable" known exact lumen flashlight (right...). The rest was basic mathematics. I found out i need to take the lux reading and multiply them by 4.1 factor.

Very cheesy made lightmeter with a white styrofoam fish transport box (properly sealed). The size is very annoying but is easily forgotten when you start to feel the smell. Cleaned with vinegar but did not help much.

I can safely say there readings are okayish perahps a bit fishy. :P

It's very easy to watch output sag this way.

I loved this description! I think it has the impact that you were probably going for.

Try bleach - that should shift the smell.

Or sodium hydrogen carbonate (bicarbonate - it can be applied as a paste with water for destinking.)

Failing those, try 100 volume hydrogen peroxide - or 20 volume if you can't get 100 volume. You can make a paste with 100 volume peroxide and bicarbonate (Wear gloves and eye protection)

Ahhh the joys of not being a native english speaker... It felt wrong when i wrote it but could not be arsed to rephrase the sentence. :/

Don:

I'll ditch the box and do a proper one without the smell. Correct sized this time!

I’m interested in building my own lightbox. I have access to a quality cardboard boxes and multiple kinds of paper for the interior surface.

The most interesting part for me is making it data-logging capable. I have a DMM with datalog features and a luxmeter.
Will it be possible to use some kind of photosensor or similar, being conected to the multimeter? I mean, datalog the V measurements for then convert that values to relative lumen?
Maybe a small solar panel? :quest:

Thanks in advance!

I think this is what you want:

This is my cheap lx1010bs luxmeter, you can see where I found the linear lux/voltage output.

I always set the luxmeter to the most sensitive state, that means: the voltage wil vary between 0-1V depending on the lux value. And absolutely linear!

On high lux values the luxmeter shows overload on the display, but this isn't problem, there is no effect on the voltage output.

Direct measuring from the photosensor isn't good, because the sensor does not provide linear lux/voltage value.