The Legendary BLF Integrating Sphere starts here! (Delivered)

For the “thru-wall” readings is it going to be necessary to center the light on a specific spot? If so, then that spot should be accurately marked permanently somehow. Consistency between all users is what makes this project different and so valuable to us.

It would also do well to include usage instructions (could be posted on BLF somewhere) so that everyone knows how to position the light being tested correctly, to allow 30 seconds runtime before getting the reading, etc.

Phil

Yea I plan to mark it. There will be a video too.

According to this lumens-to-lux calculator, for a 5.125”R sphere (area = 2.29 sf), it would take over 40,000 lumens to exceed the LX1330B’s 200K lux range.

Granted, that’s a theoretical calculation.

In our model, we’d have an adjustment factor that could be higher or lower depending on how the light is collected within the sphere. However, I’d suggest that it is most likely that the sphere could accommodate a beam of 10K to 20K lumens without fear of overtaxing the light meter. The higher meter range might be less accurate than the lower range, but it would probably be more accurate than using the through-wall method.

With that in mind, I’d suggest that the 3” diameter opening still be considered for discussion.

The problem with a 3” head inside the sphere is the beam angle crawls up the wall so far. The meter starts seeing the outter hot spot. Through-wall is totally solving that I see.

When you cut out the hole would it make sense to save the plug that is removed? That way you can reinstall it and have a solid sphere for the shine-through-the-wall readings?

Yep, that’s one reason for the tapered hole. The plug will push right back in.

Okay, so then do I understand correctly that the sphere will be set up to do “direct” readings through a hole for most lights, up to a certain size (which may or may not be 2”) but that it’ll also be capable of measuring larger lights through the styrofoam wall? I would assume then that there would be two different calibrations, cause the “through-wall” reading should be quite a bit lower than a “through-hole” reading.

So what then do we expect to be the maximum size this sphere could be used for in some way? Would I be able to read, for instance, a 4” diameter light with it?

My largest is nearly 5” and it looks good.

This is great, I tried to get the guys at CPF to do this years ago, when there was a member there who had access to a sphere at work (MrGMan - GMan?). This is the best way to go about this, have someone build a light source, have the lumens certified at different levels, and calibrate other spheres. What I suggested then was everyone chip in and build a light source, either a large bodied flashlight with CC regulation or, more preferably, a heat sinked emitter. The latter would have required users to have a lab power supply, or shipping one around to the different participants. The former would have required shipping the light, charger and batteries around. Not to mention everyone would have been responsible for building their own sphere. This way is much simpler for people who don’t have the time or equipment to build their own.

I don’t know about here, and I haven’t read past page one yet, but over there we got plenty of naysayers. Don’t listen, it’s a great idea, and while they may not be true IS’s, you can obtain readings which are repeatable, reliable and very close to an actual sphere or light output.

I built a light box, not even a sphere, when I first started in all of this and the readings were very surprising, right in line with manufacturers specs. And all that was before I built an actual sphere and bought an NIST light meter.

I have some PowerPoint presentations that might be interesting, they’re over the accuracy of off the shelf light meters and how modern LED’s affect the readings. They’re from a couple years back, and I’m not sure how I’d go about converting them and posting here, but I’m willing to give it a shot if you guys are interested?

Again, Great Idea!

-Michael

BTW: Here’s the old thread where I built my first light box, and the readings matched the manufacturer ratings closer than expected for a cheap meter and simple box.

Interesting Light Box Numbers…

ETA: Made it to page 2. Those saying the sphere needs to be bigger are correct, especially for the types of lights built around here. I believe mine is 16”, but all my stuff is packed away in boxes in storage. The balls can be expensive, but I’ll look for the link where I purchased mine, IIRC I paid around $36-38 shipped. Also, don’t forget to account for the thickness of the styrofoam in your calculations, for instance, a 16” ball may be 1” thick, thereby reducing the inside diameter.

Hole size on mine is around 2-1/4 or 2-1/2 inches IIRC, it would accomadate most lights but not some of the bigger multi-emitter lights or even some large reflector lights I have, one of the Lowes Taskforce lights that was posted as a deal here a couple of years ago, or some of my (comparatively) small handheld spotlights . Maybe I’ll start over and build another sphere, I’ve put new emitters/drivers in some of those and some are in que to be modded again with new LED’s, it’d be nice to be able measure those.

And to be inline with ANSI there is a small waiting period before taking the measurements. But unless it’s uncomfortable to hold your hand in front of the light for a short period I don’t think it’ll be an issue. With hotwires definitely so, not so much with LEDs. I’ve got some that you can feel heat with your hand in front of the reflector, but nothing hot enough to melt styrofoam. I would think the greater danger here would be in an extended runtime test, where the light itself became too hot to touch or hold, and if it was in contact with the edge of the opening might cause a problem. The meter I have is an Extech 407026 , it’s a data logging meter (if you have the cable and software) and when I’ve done extended runtime readings I’ve used a small PC type fan to keep the temperature under control, both for the light and sphere.

Great input Texlite, the more expertise, the better! I’d love to see your powerpoints about the subject. Direct posting is nice, but if it is too much work, you could drop them somewhere and post a link?

Found the place I bought my styro sphere from, it is indeed 16”, and miraculously, I was also in the ballpark for price~$38. The bad news is they’re no longer on the site, the link is dead. I’ll post the link as it has the model number, maybe someone can dig up another source. Barring that, they can be made relatively easily out of a cheap bouncy-ball and fiberglass or papier-mâché.

Here’s the link: http://www.smoothfoam.com/mm5/merchant.mvc?Screen=PROD&Store_Code=PCO&Product_Code=10073

-Michael

Great stuff happening! :slight_smile:

I’m a little worried about through wall, but won’t use it for my bigger lights if it proves problematic. I DO have lights that will melt the styrofoam, in short order. Even the 5000 lumen M6 is capable of that.

I have a UCLp floor plate on the way for my PVC Trap box, courtesy of Chris (Thank You Chris! :slight_smile: ) so I’ll report on my findings on how that impacts the readings.

That is a very nice powerpoint about the problems you encounter when a luxmeter is used for led lighting! Thanks for sharing, I can recommend everyone to read it.

Our biggest problem right now is a source of rated lights to perform the calibration with. I PM’d manxbuggy1 and rdrfronty yesterday, no word there yet.
I received a Lux meter today and did learn my BLF A6 is ultra sensitive to battery charge, so getting some constant-current lights here for calibration is an absolute must.

I just got in a 1330B, my 2nd, and after doing a few experiments of room light and PVC pipe light box of various lumens levels, I’ve found the new 1330B reads 4.35% lower than my old meter. The new meter was bought off of eBay for$20 shipped - from the eBay user ivandogone mentioned earlier in the thread.

I emailed him about the issue and questioned about his sourcing, but of course he did not answer the direct question, and instead, apologized and offered a refund for shipping it back.

My thinking is the new meter is off. I didn’t compare straight throw (lux to kcd) yet, but expect the new meter to be reading lower. His 1330B meters are the cheapest deal on them by far. They go between $30 to $50 everywhere else on eBay. I paid $35 for my 1st one. I really was thinking these may be factory rejects of some sort that are out of calibration. It certainly looks brand factory new though.

I always thought the sensor and calibration of these 1330B meters was a cut above the cheaper meters out there. I think manxbuggy1 reported way back his 1330B was about dead-on to his more expensive ExTech meter - might have been the LT300 featured in the PowerPoint linked above (I viewed it - nice!). He mainly wanted the ExTech for it’s more advanced features.

Not sure if Dale got together with manxbuggy1, but would have interesting comparing their 1330B’s. I’m pretty sure lights I worked on for manxbuggy1 in the past compared very close between his meter and my meter, and I thought the same was true for Dale and manxbuggy1 as well. That’s why I always thought #’s Dale post and #’s I post are very comparable in throw and lumens (same meter in the same range, same PVC light box).

You are talking lux though. Since I will have 10 of them I can calibrate lux to the average of the 10 meters. Our problem is sphere-design specific. We need a flashlight that is definitely 10 lumen so I can stick it in the sphere and say, look, that reads 60 lux. And another that is definitely 1000 lumen so I can do the same and say, look, that reads 10000 lux. Then graph all the data into a equation for the app.

We need rated flashlights. Not meters.

To clear things up, light meters measure lux. Lux is brightness at a POINT. Lumen is a measure of total light, so you need an integrating sphere to take it all into account and get TOTAL light.

+1, totally agree with djozz that’s it’s more important to have a “calibration” flashlight with consistent output, amp regulated/limited, like with a single 7135 for example, low power where heat is eliminated as a variable. A light that doesn’t drop output over time, not dependent on the cell used.
These meters/spheres scale well. If you have any doubts and want something in higher lumens, a multi-LED light like a SRK style with an amp regulated/limited driver might be a good choice. Keep the amps somewhat limited again to maybe 3-5A total, but driving 3-5 LED’s to get higher lumens. The BLF SRK 7135 based driver would be perfect for this.

I don’t really trust many of these so-called brand name, ANSI/NEMA rated lights - too many variables, and it’s been well proven one brand can tend to be high while another brand tends to be low. manxbuggy1 or rdrfronty first told me of this - probably has posts here on BLF describing it a while back. These guys take one light early on, maybe a prototype or pre-production unit, and send it out to be tested. Then things go into production, and who knows what changes. They don’t commit to say what specific tint or bin they use, so how can they assure keeping the tint and bin consistent from batch to batch? Each light doesn’t get individually tested, and you never see them update the specs unless it’s a full model upgrade.

NiteCore is a good example. I bought 2 identical NiteCore EA41’s at the same time, same model, and one was like a full 15-20% lower than the other. Nothing seemed to be wrong with the lower performing one, though who knows, maybe it was a defective driver or something, but had no other symptons accept lower output.

Good luck with this, if you get something working I would be interested in a purchase. I will reserve all my comments save one… If it doesn’t measure at least 5,000 lumens, I can’t use it.

What does this mean? Either I’m completely dense (a definite possibility) or I still don’t understand how you intend to calibrate these meter/sphere combos. Will my specific sphere be calibrated using my specific meter so that I’ll have my very own special calibration equation that just works with my pair, or will they all be calibrated together to an average value, so that some of our spheres read higher/lower than others just due to the averaging?