WTS: Calibration lights for DIY integrating spheres / lumen tubes - 67 € -

Thanks for doing this maukka. Looking forward to checking my current calibration (best guess with a zebra light).
Well at least there will be 20 or more BLF members with calibrated spheres now.
I have already seen that rated lumen’s is not the same as actual OTF lumen’s. I did a led swap in a Sunwayman V10R ti light yesterday.
I swapped out a Nichia 219c D240 for a AEDe 90 cri Samsung LD351D and gained 100 lumen’s with out changing anything else. I would have never known that before. Having it calibrated will make it even better for testing. Thanks again maukka. :+1:

Do you mean rated lumens and actual lumens?

OTF just seperates the raw led measurements from the led plus reflector, lens, driver, etc… losses.

Manufacturers ratings are OTF, but they are usually very “optimistic”.

I swapped the xpl out of an On The Road U18 and put in a sst40. It went from 690 to 845! I’m guessing it’s due to differences in forward voltage and overall efficiency.

BTW, the U18 is rated at 1090 lm even though it only draws 2.3A. :person_facepalming: It’s still bright, though. :+1:

Yeah rated (printed on the box) versus real life actual out the front lumen’s. I think a lot of these company’s just assume if they put x led in and drive it with y current then they get z lumen’s and that’s what they put on the box. Never taking into account any losses. On a small light driven conservatively 20% losses isn’t gonna make much difference but when you start driving a led harder the losses are huge. 20% of 300 lumen’s is a 60 lumen loss but 20% of 10,000 lumen’s is 2,000 lumen’s. Changing wire size can even make a huge difference. We have all seen things these company’s do to cut corners or if there there x bin leds run out, they just use y bin. Or if drivers start burning up they just reduce current in they next batch.
So yeah that was my point what’s advertised or printed on the box will more than likely never be what’s actually coming out the front of the light. Even if you accounted for all the losses and knew exactly what a x bin led made at y current, you would still have cree’s margin of bin’s. You would have to rate the lumen’s on the low side of the bin to not be lying, because out of the thousand’s of lights they make their still playing the bin lottery. There are a few exceptions to this though that get close, probably because of the bin lottery.

And so goes the world of ’Chinese Lumens’ in the realm of reality. It’s usually like comparing Grapefruits to Grapes. :smiley: :smiley: :smiley:

Don’t know how many of these yo are making available, but I would like to have a set if available whenever.
Thanks for all you do.
Keith

You think certain companies skip the OTF part? Here is why that doesn’t make sense to me. The manufacturers don’t actually know what the raw leds will do since Cree’s (or whoever) data sheets only go up to a certain level. If you think the manufacturers are looking online at independent emitter tests from BLF and TLF, etc… that drive them beyond spec, I think you give them too much credit. :wink:

It really is completely unknown how some manufacturers get their rated specs. Some may even have homemade lumen tubes that are calibrated by guesses based on other companies specs. :stuck_out_tongue: lol

Some are just pulled out of thin air. A xml on a aluminum star driven at 2 amps, yeah 1500 lumen’s sounds good. Wait x brand is at 1700, just make it 1800 lumen’s, Hurry up get them listed on ebay and AE. :smiley: Just making fun, I have no idea how they come up with their lumen rating.
.

It’s crazy at some of the inflated exaggerations we see, but I already know that there bogus. What I didn’t realize is how some of these good lights we all love seem to be off as well, I mean a expected a 1000 lumen light to be at 900 or 850 or so. Just didn’t expect it to be 650. I have tested several of my recently bought lights, which I expected to be down on lumens from the rating. Some are reasonably close to what I expected but some where a lot lower. My tube hasn’t been calibrated yet besides my guess with some known good lights, so i don’t expect it to be accurate yet but I don’t think it’s out to the extent to which I’m seeing on some lights. These calibration lights can help shed a lot of light on these rated flashlights. Instead of one or two people being able to accurately check lumen’s the BLF community will have several that can get a reasonable reading.

You got the extra diffuser discs from TA and dropped your readings down by 32, right? So you should be “calibrated” to about plus or minus 5. I call it calibrated. I’m hoping Maukka’s test light can narrow it down to an even smaller variance. Maybe plus or minus 2%.

No, mine is self built tube using a zebra light and a few other lights to calibrate that should have been close. I just don’t know how close yet.

Ah, okay. I thought you had a TA Tube. :wink:

I have enough lights coming in for at least 40 sets. It may take some weeks though.

About ±2% is what I would rate my lights at considering all the variables such as (ambient) temperature, battery differences, repeatability of measurements, calibration accuracy. Never saw a difference of more than 0.5% between measuring sessions of the same light.

All the S2+ were measured with a Samsung 30Q on the very stable mid mode (2/3) and the BLF348 with an Eneloop AAA.

Btw, the output difference between the highest and lowest performing Convoy S2+ from a similar set of 20 was 12%.

I will send CRI measurement data as a text file to all who ordered a set. It includes the spectral (SPD) information and you can open the file in Babelcolor CT&A software to visualize data (CQS, CRI2012) that is not shown in the CRI measurement screenshot I also sent. The software is not free, but it can be used for this purpose when a measurement instrument is not needed.

I also recommend browsing through the help file it comes with. It covers many aspects of color rendering. You can also download the PDF separately here.

That’s interesting. Lets assume all that is from the emitter and not other factors. Each of Cree’s bin numbers go up in 7% steps, but they don’t say their tolerance is plus or minus 3.5. They go with a much wider plus or minus 7 which overlaps half of the bin on either side of it.

I think they give this wider tolerance just to account for how other people measure it and the measuring conditions, temperature, etc… It saves their butt as 12% is just barely within their 14% tolerance. :smiley:

There is other varying parameters too, but someone has to measure them to know how large they are. Two examples that are simple to test if someone takes the trouble to do that:
-take 10 pieces of 7135 chips and measure the current they put out.
-take 10 S2+ reflectors and measure the relative efficiency.
This can all add to the variance in output of these S2+ lights

Sets 1-9 shipped. Got some more BLF348s today so more will follow soon. The new batch has a 219C but still seems to measure above R9080 albeit somewhat greener than the 219B as expected. Doesn’t affect calibration.

Received Maukka’s email with the test reports. Getting a S2+ with 6252k and practically perfect non tinted DUV of –0.0002 so I will have a reference of what pure CW looks like. The other is 5053k and DUV of –0.0045 so I can see how rosy/magenta that might look. Excited to receive them! :smiley:

maukka I see in the OP that you mentioned the possibility of tracking for an extra 30-40 euros. Will the options with tracking be significantly faster than the 7-12 days option?

If you just need the package as fast as possible, an express letter would be the fastest. It doesn’t have tracking but it has an estimated shipping time of 4-6 days. The cost is 14 euros more so total 78 euros.

A normal package with tracking to USA costs 34 euros more (98 euros total) and takes 6-8 days.

EMS would be 45 euros more (109 euros total) and takes 4 days.

I also got my email and had the same thought process. My numbers are different than yours of course, but I’m really looking forward to seeing a beam with my own eyes and comparing my subjective ideas of what that beam looks like to the objective numbers representing it. And that’s not to even mention how useful these will be for calibrating our homemade measuring devices. This is awesome, and I can’t thank Maukka enough for doing this for us.