Even though it is easy enough to DIY, this is a nice initiative! What is very convenient of course is that the calibration is done for you.
I can only really comment if the exact design is known, the exact design is crucial for function.
About a few concerns that are noted above:
-indeed a 15cm sphere will heat up a lot, no problem with measuring for a short period, but continuous measuring of high output lights will be a problem,
-a high hole-to-innersurfacearea (which will happen in a small sphere) will make the multiplier very dependent on the shape and reflectance of the flashlight in the entrance hole, in other words: without multiplier-correction for every measured flashlight the sphere will be more inaccurate, calibrated or not.
-with a small sphere like this you need to lower the amount of light reaching the luxmeter to get into a workable measuring range, in the sphere of the link above I did that by measuring through the wall of the sphere.
-do not bother with coatings of the inside of the sphere, I experimented with BaSO4 and I hardly saw anything change compared to sanded bare polystyrene.
-please carefully sand the inner surface of each sphere, it will help integration a great deal
I did some testing on my 12” D.I.Y. IS here and found that the biggest problem is the change in reflective area based on the size of the light placed in the entry hole. Lumen numbers seemed to be within reason if you calibrate with a light with a medium sized head. I’m not totally done tweaking my calibration yet either.
It was an easy build. I’d be scared of using the HS1010A though! I’d move up to the LX1330B. My HS1010 proved to be junk.
Garry,
I saw you sphere and am sure your problem is you have the light sensor aimed mostly at the hot spot. You want the detector to be on the side so it does not look at the hotspot. Measuring the hot-spot is not a measure of lumen. And as you mentioned, different sized hot-spots will play wacky with the results. Perhaps your HS1010 was not junk after all?
No the HS1010 was definitely junk (I posted a pic of both out in the sun with both reading drastically different - in the Courui Gearbest thread).
My goal with position of the sensor was to not get direct light from most any light placed in the hole. And my video showing testing of integration shows my readings staying pretty consistent no matter where I aimed my SK68 set to zoom (i.e. tight hotspot).
Even if somewhat limited, I think it’s a great idea. Just like the group buy on the Uni-T, it brings more of us inline for comparative benchmarking which can only be a good thing.
Do you know if there would be differentiation between floody lights and throwers? Even some smaller lights like a Triple or Quad Eagle Eye X6 can have a good bit of variation in the sphere as compared to, say, a de-domed XP-G2.
Different doesn’t mean junk for our use. And how do you know which one was right? There is no universal conversion from the lux on the screen to the lumen we want. That is discovered through calibration. So again, I don’t see any compelling reason why the HS1010 could not be as accurate for us as any other.
I got the Uni-T clamp meter to be on the same page with a bunch of folks here and to double check my Mastech.
I’d like to have a back-up to my light box for similar reasons, mine is large and works ok with bigger lights but the smaller lights kind of get swallowed up by it. (not that my smaller lights are not also high powered, my MecArmy PT16 is a tiny 16340 sized light but it makes 2186 lumens. )
So yeah, even having a second meter to double check would be nice for peace of mind.
I should post the legal details. Sorry but I am a to-the-point kind of person. So buckle up.
It would please me a lot if you were a happy. But I did not build a profit in. The price will cover my bills and part of my time. Therefore there will be no warranties, expressed or implied. I am just promising to ship the parts described. I am not branding this my own. It is yours. Their will be a lux-to-lumen app, but the BLF community will need to maintain it. That will be easy for anyone knowing a little bit of javascript, no special development software required.
The most expensive part is a $17 meter. If you buy this setup, and your meter fails, I won’t pay for it or any other part for you.
Gary’s sphere measures the light through the wall so where-ever the sensor is placed there’s no ‘aiming’ going on. Moreover, if the sphere is doing its thing (integrate) there will hardly be a hotspot leftover where the flashlight is aimed: everywhere in the sphere more or less the same light will be measured, except if the sensor is receives light directly from the light source. Gary checked that with the zoomie in spot mode.
If aiming at the hotspot (even with the sensor mounted through a hole) gives a significant different reading than aiming perpendicular to it, your integrating sphere does not integrate well and will be beamshape dependent.
This is a great idea! Thanks for trying to put something together.
I am interested, but have concerns about size and heat. I would like to be able to do lengthy runtime tests on all flashlight modes with a logging meter. I don’t have any really big lights now, but since I am a newbie, the depths of my depravity have not been fully explored yet.
Who knows what crazy flashlight I might get tomorrow? BTW, what’s this thing called a K70? Someone said I need one! :smiling_imp:
After I joined BLF, the first subject I delved into reading deeply was IS’s since I’d never even heard of such a beast before. That’s been some time ago and I don’t remember it all. What I do recall is that size matters, as does the placement of the light sensor, and that these things come into play when testing lights whose lens size and lumen output differ significantly from that of the light source used to calibrate the sphere, The general consensus was that within reason larger is better.
My initial reading of the OP here missed the size. While it will probably do fine with small lights, I am not sure it will serve even comparatively with larger lights. I’d like to suggest using the largest styro sphere which can be shipped economically. A quick search led to this 25.4cm one : Michael’s Arts nand Crafts Not sure if it’s a half ball or both parts but it has alignment pins molded in and looks like a good candidate. I’m sure there are other larger sizes which can be readily had.
Which brings me to another thought. As long as everyone has the exact same ball, all that would be needed is a calibrated meter and a template for placing identical openings in the exact same place as was done to calibrate the meter. That would make logistics easier as you’d only need to ship the meter and hole template, and the members could buy the ball themselves. While not ideal for calibration purposes it should get everyone pretty close together which I think is the salient point with this effort. If the ball could be sourced from a cheaper international seller that would be even better- I will look into that later today (and hope others do that too as I may miss something).
So while I am still greatly interested I must set a ~25.4cm ball as the smallest I would want. Don’t let that adversely affect your efforts as I think most important is that we all have access to something of a standard IS at a budget price which would make comparison with other members lights much easier and more relevant, hopefully adding interest to our addiction hobby and bringing even more joy to us. If we’re going to do this, let’s do it as well as we can.
With the ANSI measurement at 30sec, can’t you shine out of the sphere for 29 sec, and then take the measurement?
To me, the lengthy measurements isn’t really a necessity.
Put me on the interest list, I like the initiative!
Just saw your post Sawmaster. Nice sphere!
Since it needs to be sanded by hand from the inside, my gut
feeling tells me it needs calibration. Even when everybody is using the same grit sandpaper, fluctuations will be present I think? I guess even the direction of sanding can alter the measurement.
I’d be happy to go the extra $20 for the bigger sphere, if there is consensus it would greatly improve the measurement.
What is very convenient of course is that the calibration is done for you.
)