NOTE: A third version of these readings (this being the second) has more recently been completed (20/7/2013) with more lights added and a with the light and lux meter repositioned to effectively eliminate the throw bias, which was still present to some extent in this version. You can see these results here: V3 Comparative Lumen OTF Table.
This is a comparative table of lumen readings taken using ceiling bounce in a white room with a photometer in a set position. These readings are estimates, and there may be a slight bias in favour of more focused/throwy lights, but not enough to make a significant difference unless the one being measured is almost full flood.
The columns, left to right are: Torch ; Mode ; Lumen reading ; Corrected lumen reading (108%)
The ‘Corrected’ column is because having completed the list I thought it was under-estimating actual readings. Any suggestions on a more accurate % for the correction are welcome, but remember that any % added to/taken from one will have to be added to/taken from all of them.
More lights may be added in time, but I have done most of the ones I have.
Also, I apologise for the poor quality image format of the table. I had tried organising the information into a normal post but whenever I preview or post something all formatting disappears and everything gets pushed up together again in a big mess. So the only way I could think of doing it was to save the spreadsheet as a .png, upload and then link to it. If anyone has any better ideas (which I’m sure someone has) then please let me know Now sorted, thanks SashiX
Ok, I’ve just updated it to correct the Mini AA confusion and include OTF values on eneloop. I also added in readings for the Preon P0, Jetbeam BA20, Zebralight H501w and a couple more P60 drop-ins. You can see the effect of focused vs full-flood in those last two readings, an R2 XP-E reads 206, compared to a fully diffused R5 XP-G at the same drive current reading 200. Subjectively the output is similar for the whole room though so maybe the two layers of diffusion film on the L2 are eating up some output.
SashiX; I would do but I don’t really know what I’m doing tbh. What I’ve done so far is copy the table from the spreadsheet into word as an image, then ‘save picture as’ before uploading to imgur and putting the link in here. I think it is transparent as on imgur the background is very dark grey, the same as the surroundings. But I’m not sure how I could change that tbh.
I didn't know that the Nitecore Sens AA was able to run on 14500 batteries. After seeing your post I double-checked the Nitecore user manual and sure-enough, 14500 @ 3.2V are acceptable (good news).
Another thing I didn't previously know is that there are (at least) three different voltages of 14500 batteries (3.2V, 3.6V and 3.7V). It looks like it might be a problem to put 3.6V or 3.7V 14500 batteries in a Nitecore Sens AA. With my luck, I'd get my batteries confused and do exactly that.
Very nice. I have several of the the same lights and meter them using a DSLR and come closer to your unadjusted numbers, so your numbers look pretty good to me. Are you “calibrating” to some base light or scale?
FWIW, one of best ways to equalize the floody and throwy beams would be to move the lights closer to the ceiling so that you can match the complete beam sizes between the lights. Actually, for my measurements, I bounce horizontally between two walls in a narrow hallway (bounce off one wall, meter off the other wall). The closer distances improves brightness which seems to give more accurate and consistent readings, especially with the low and sub-lumen levels.
I actually run mine on 3.7v (li-co) 14500. It does overdrive the LED when it’s near full voltage, pushing it to around 2amps before dropping quickly to 1.7 and then 1.4 ish for most of the run, so 3.2v ones would probably be safer for long-time use. I just wasn’t impressed with the output and tint on an eneloop so I thought I might as well run it on li-ion until the LED burns out before replacing it with an XP-G2 or Nichia 219 That said, it doesn’t show any sign of degradation in output yet despite a few hours of very hot runtime now. The automatic dimming is an especially useful feature for keeping heat down when used like this too.
The first column was a simple x5 of the reading on the photometer, which actually seemed to be reasonably accurate, if a little conservative with some base lights which had ANSI/well-documented measurements from others. Then I just thought that adding 8% seemed to match them all up a bit better to what others have measured, but then it’s all relative anyway so I thought I’d leave both columns in for those who like being more conservative with the figures.
I had considered raising and lowering them, but then I thought it was probably safer to try and keep possible extraneous variables at a minimum and just measure them all at the same height (raising them up would mean my arm getting in the way of some of the light from the one side wall as well as adding a more subjective aspect to the process).
OK, tried that myself and Word doesn’t save it well :~ The solution is very simple: select the Excel part you need, right button -> “Copy”. Open Paint -> “Paste” et voila Now you can save it in PNG
So you are saying the BC40 really does have 900 lumens? Because someone else did a review and said 948lms. I also have one and have always thought it was brighter than 830lms. I know it would be hard to tell the difference between 830 and 900lm, but it is much brighter than other lights I have that pull over 3A.
I think that other guy may have been me last time around. I’d actually forgotten about that but I had done some readings in my bathroom of where I lived last year. These reading seemed a bit fairer and less affected by flood/throw so I re-posted a new table (also including anything new I now have). I might delete that old post and link to to here now…
900 OTF for the BC40 does sound a lot, but I would definitely agree that once you’ve actually seen the output of it it doesn’t seem unreasonable. I think the key factor is that Jetbeam used copper PCBs for the BA/BC series, which seems to have increased efficiency at higher currents, so especially on the BC40. If you look at Match’s tests for the XM-L on a normal star and then on copper it makes the 900 OTF for the BC40 seem much more believable too
I think that’s the reason that I have decided not to buy budget lights anymore. Today I just put a new driver in a C8, it pulls 3.1a from my Sanyo’s, but my BC 40 still blows it away! That says a lot really and the tint is so much nicer in my BC40.
So the question is, do you go out and buy a budget torch for $20, then change the LED, add that to a sinkpad, change the driver and the realise it is still not as bright as a BC40!
I think the BC40 may appear even brighter than it is due to the deep reflector giving it a relatively narrow but bright spill. Though if the difference is still there in a ceiling bounce then we can’t argue with it.
I would point out though that my BC40 pulls 1.8A from two 18650s at 3.8v. Thats 3.8x2=7.6 x1.8=13.68 /100x85(assuming efficient buck driver)=11.628 /3.25(rough LED vf)= 3.58A to the LED.
Couple that with the copper PCB and the high transmission glass lense and there’s some way to go yet to match it with a single 18650 on a budget
Now you can save it in PNG 