My Light Sphere Build (Pic Heavy)

So, my dad came to visit and we decided to do a light sphere build. Herewith some readings I took. My multiplier is estimated at 2.5:

Edit: The last 2 lights are also DD.

Edit: The Convoy S2+ 18350 does not run a dedomed XP-L V6 but rather a dedomed XM-L2 U4 1C. Sorry about the error. Nonetheless, I believe the performance of the 2 LEDs are on par.

We made use of a 40cm [15.8”] diameter styrofoam ball. 2 holes were made, one at the top for the flashlight entrance and another at around 90 degree angle for the light meter.

We got the light opening perfectly flattened out. Further down you’ll see the wooden plates made as light openings. These plates seal perfectly on the main light opening.

Both holes have been honed out to create a smooth ‘transition’ surface. Some humps inside the ball was also smoothed out. No special coating was used in the inside, left as is. I also did not sand down the inside surface.

We’ve inserted as small as possible baffle to ensure no direct path from the light opening to the light meter sensor.

We made a set of wooden plates, sprayed matt white, with different size holes. This is not the full set, only some of them.

One of the plates lying on top of the main light opening. They seal perfectly on the flattened styrofoam surface.

We ended up fitting the lid with hinges to nicely close up the flashlight hole.

So, there you go! Credit to my Dad for helping out with this AWESOME build! I’m very satisfied with the outcome. At last, now I can take my own lumen readings… :wink:

Very nice. It looks good functional and attractive. No need to hide it away if guests visit. Congrats.

Congratulations! H)

Very good job :open_mouth: everything looks very prolix, it would be very useful for me :wink:

Looks awesome man!


Very nice and useful build! I like the attractive looks as well :-)

I can't help a few remarks (sorry about that. ) :

*please do sand the inside to a matt finish, it really improves the integration (insensitivity to light direction/beam profile)

*a square baffle by definition is not as small as possible for round holes (but will do the job well enough )

*the thin paper baffle looks like it may possibly bend to another position over time, and this may influence your multiplier a bit.

*using white plates as spacer around different sizes of flashlights, while using an unaltered multiplier, assumes that the avarage reflectivity of flashlights is about equal to white painted wood. This may be correct but I know of no one who actually did measure that.

And if it does not work out as an integrating sphere, you can always use it as a chamber pot… :party:

I like.

Sweet. Thanks for sharing. Can I borrow your dad for awhile?

great build!

Edit wants to know which diameter is necessary for a sphere to measure up to 10k+ Lumens?

i found 40 and 50cm ones at Amazon

My 50cm ball together with a luxmeter that measures up to 200klux has a limit of 9300 lumen (a few other factors are involved such as hole sizes but they have a more minor effect) , but when it was still uncoated, the reflectivity was higher so the limit was lower. A smaller ball will lower the limit only further so certainly go for the 50cm.

Thanks for all the comments… :slight_smile:

Hi djozz

No worries, thanks for your remarks. You have lots more experience in this type of thing :slight_smile:

From what I’ve read, I could not get a conclusive answer if I should do it or not? I’ll take your advice. Will report back as soon as its done. I’m curious to see how its going to affect the figures.


I’m not worried about this. Its white carton and the way I’ve attached it will prevent this from happening. I’ve cut a ‘supporting pillar / foot’ on each side of the baffle base, each bent 90 degrees in opposite directions. The center part of the baffle base slides into a groove that has been cut into the styrofoam. All glued with styrofoam-safe glue. The baffle will not shift / move / change orientation.

I would have liked to build a much bigger sphere where this effect would have probably been minimized. To try and find a multiplier for each plate / hole size is obviously not easy. For now, and whats practical to build / implement by me, I’ll just accept the, possible small, percentage of inaccuracy.

Appreciate the comments! Thanks.

With the sensor mounted inside my sphere, I can go to around 5000 lumens. With the sensor mounted to the outside of the sphere (using the sphere wall as a filter) I can go to around 500,000 lumens.

so you have 2 calibration sets?

I currently have the same issue… my limit is around 5100 lumens. I’m thinking of maybe just fitting some kind of a filter (piece of film) over the light meter hole inside the sphere? What do you think, will this work? Its going to be more difficult for me to shift the meter further outwards. Don’t want to tamper with the mounting mechanism now.

Indeed very nice build. Two questions:
How did you get multiplier estimate? The lumen numbers sounds high to me.
For the the C8 test, you got >1600 lm. Is this possible? (Cree spec for XP-L V5 at 3A is 1058 lm. V6 maybe higher but not by that much.) What is the tail current?

I’ve estimated my factor by measuring a handful of standard ‘name brand’ lights. Its basically all the ‘Stock’ lights in the list above & 1 or 2 additional ones thats not listed…

I don’t have the C8 mentioned above anymore, so cannot measure the tail amps. I’ve measured another C8 XP-L Hi V3 2B now and got 1625 lumens with turn-on and 1555 at 30s. Tail amps were measured at 5.0 amps after 30s with my fluke 374 clamp meter.

I don’t know about official tests here on the XP-L Hi V3, but Djozz did some testing on the XP-L Hi U5 - around 1600 lumens. According to he’s test, I believe my figures above might be in line. I’m definitely not saying Djozz’s figures are 100% accurate, this is the only figures I can bench against here.

But, I still have a good feeling about my figures measured above, and my current sphere multiplier. I believe its accurate to around 10% up or down. I can live with that.

You’ll see that some of my stock light measurements are below the advertised lumens, e.g. Maxtoch 2X. Its rated at 1300 lumens, I measured 1215. Same with the Thrunite T10. Its rated at 169 lumens, I measured 147. So, I did not just pick the highest multiplier possible, I did try to find a happy medium between my stock light measurements.

I should probably sign up for the light currently travelling the world to get our spheres in comparable state :).

O.K. May be you are correct. I don’t have many of these flashlights anyway so I probably shouldn’t make the judgement.
For C8, if your current is 5A, 1600 lm could be possible but I don’t have a driver can produce that high current anyway. BTW, what driver do you use to generate such a high current?
For the Convoy s2+ 18350. Is it one 18350 one? Again can you find out what is the tail current it took for the measurement if you still have it?
I am trying to get my IS calibrated so I would like to have any information I can have. Thanks.

The C8 is driven by this.

Not sure what the question is here? This is the S2+ ‘Short’ running an Efest Purple 18350 battery.

I’ve realized an error in my OP list. The Convoy S2+ 18350 does not run a dedomed XP-L V6 but rather a dedomed XM-L2 U4 1C. Sorry about this. The tail amps measures at 3.5 amps after 30s.