Match's Mods: Homemade Integrating Sphere

No, you shine at the opposite wall. The baffle actually means an imperfection in the sphere geometry and thus is something bad, but direct light from source to detector is much worse.

What about making the first layers out of white paper?
I want such a sphere too and I thought that if we all get our olight i6s(I guess a lot of us have ordered this light) we can calibrate our spheres to each other…

I just found and read through this entire thread. And dangit! Now I’ve got to build one too! Wouldn’t ya know my son just burst a 20” rubber ball and we threw it away?

Subscribed.

Figured I’d give this thread a good ol’ necro bump.

Going to try my hand at making one of these spherical beauties in the coming weeks. No idea what I’ll use for the calibration… And I’ll need a light meter!

Funny, I just started at one just yesterday, before you dug this thread out :-) , only 25 cm (10ish inch) diameter because I do not have space for a bigger one, but using the 'match' method. No idea when it is ready, or if it is any good at all, but it has been a loooooong time since I last did paper mache :-)

Barium sulphate,out of date epoxy/polyester resin is quicker and stronger,after an initial papier mache layer,ping pong ball for light baffle,it's round so no direct reflections.Also,I might make a second light meter hole just to check the readings are the same.In theory,an IS puts out an even spread of light everywhere,that's the point!Always good to check.I stuck a layer of plastified foil(ground coffee packet)opposite the light source on the principle that ,for any excaping light through the sphere,throwers will lose more than flooders and hence underread.Opinions welcome

Considering that you can buy a 16” strofoam sphere for $40, I would not waste my time messing with paper mache, etc. The styrofoam spheres that are now available seem to all be made with a flat surface on the ends of the interior. I made a sanding jig by tracing the sphere inner radius onto a 3/4” thick board, cut the board, placed sandpaper over the edge, and sanded the flat spot to match the sphere radius.

Instead of barium sulphate, a light sanding with fine sandpaper on the interior of the styrofoam sphere will remove the sheen and perform pretty much as well as the sulphate/latex paint coating.

Epoxy/polyester can yellow with time. I would not use it.

How much to ship that foam to Canada, plus styrofoam does not reflect all the light striking it, i have styrofoam boxes that i can put a light in and see maybe 10% coming out the other side (maybe 3/4 thick), i would assume the paint will reflect more of the light, now mind you if the light penetration is linear with increasing brightness then a calibration should compensate when taking readings, if i had some calibrated lights of very varying brightness it would be a worthy experiment (unless you already did it and have data to share)

The reflectivity of the styrofoam is not much of an issue… it drops out when you calibrate the sphere.

With my sphere, I can do around 5000 lumens with the sensor mounted in the sphere. With the sensor mounted on the outside wall (using the styrofoam as a light attenuator) I can do over 600,000 lumens. The sphere is absorbing over 99% of the light.

Styrofoam is light… but bulky. Shipping to Canada should not be to bad. You might be able to find a local seller.

The reflectivity does matter in the type of IS we are building.

Not having build one myself I do not know how much it matters, but in this type of sphere the detector 'sees' the first reflection of the light. There are two scenario's in which that is not a problem: 1) the detector (luxmeter) is not angle-sensitive, i.e. it reads the same value for all incoming angles, or 2) the reflectivity is near 100%. If that is both not the case, the integrating sphere will be more or less sensitive to beam profiles: throwers will read different from flooders or bare leds. And that is precisely what we want to cancel out by using an IS.

At least my cheap luxmeter is not quite angle-insensitive, I found, so if I build an integrating sphere for it, I will pay attention to a good reflectivity of the inner surface.

Again, it may be a bit theoretical, and in practice may not matter that much, but in theory at least a not optimal reflectivity is not just a matter of a different correction factor.

Annnnd, now I have a desire to build IS. God dammit guys, you have a negative influence on me :bigsmile:
Luckily those fitness balls are way too expensive to destroy one…

I contemplated using one of those that had been torn. Patch it up and inflate it just enough to get a sphere. Probably would be far too big to be practical, and it got thrown away anyway. :frowning:

No, for practical purposes the reflectivity does not matter much at all. White styrofoam is pretty darn reflective. Whether the surface reflects 95% of the light or 90% does not matter much once you calibrate your sphere… been there, tested that… Even with a rather tight thrower, the detector is seeing reflections from a LOT of angles… unless you are using a focused laser as the light source.

What you want to eliminate is a specular/non-difuse reflective surface that can direct a direct reflection of the light to the sensor. That is where the barium sulphate/sanding the styrofoam surface helps.

To test how well your sphere integrates, aim the light at different parts of the sphere and see how the readings change.

No such thing as “too big” :smiley:

Thanks for the practical input, I am too going to build an IS (yay!), so this helps knowing how fixed I need to be on small details (not apparantly).

When I first got my sphere, I did not sand down the flat spots to make them spherical (I put the 4” diameter light port where one half of the sphere had the flat spot and the hole pretty much took up the whole flat area). The sphere still worked rather well with the flat spot on the other side of the sphere where the light hit it square on. There was maybe a 5% improvement in directional sensitivity after I sanded the flats down to a spherical radius.

My sphere sits on my kitchen floor and the room is lit by around 2500-10,000 lumens of room lighting. I have not opaqued the outside of the sphere. It works just fine and the room lighting does not appreciably affect the readings… unless you are trying to measure a small/dim light. I get around 1-2 lumens of room light leaking into the sphere. For dime lights I wait until night and turn off the room lights or just have my system subtract the background light level reading. (For color temperature testing of dim lights I use the darkened room).

heres a bump for this thread - I am currently on layer #4 of this giant glue ball of lumens :slight_smile:

For the US/UK users I see most people use Lux. 1 FC x 10.76 = 1 Lux. Other than way bigger numbers while using Lux, any reason not to use Foot Candles?

For whatever reason my 14.5” IS/Meter combo uses in LUX:10.40697674418605. The correction for 172/1790. After looking at the correction number for my (only) ANSI/NEMA FL-1 certified light of 172 lumens, I switched to FC on my meter. Basically same readings but now can use direct readout. I understand if I change meter or repaint sphere, everything changes.

I see that many others got correction numbers from 7.xxx to 12.xxx, so while fun, I would need to send IS, meter and light to BLF laboratories, INC? I like the suggestion to buy a few more calibration lights, so will continue to enjoy using the IS.

Going to build one of these! :smiley: Thanks Match!

Welcome to the ball-of-light measuring group. :slight_smile: