UF-1504, 1503, 1505 - multiple LED's tested for throw (just what you have been waiting for!!!)

I just noticed that i didn’t answer your “max lux performance and lets say LG18650D1” question correctly.

The thing is when you run it fully DD like this, its only the first few seconds that is at max output, then it drops like a rock.
So a few minutes maybe 2-5min is most “fun” then it its visually shorter & shorter, when i goes under the Vf it drops a lot of output quickly.

I am only talking about high drain 18650 batteries though, you could get a 5000mha+ 26650 high capacity battery & it would last much longer, with much longer usable output. the best way & only way really to regulate a light like this is to use different batteries to do it, because the batteries is the only thing that “drives” the light.

I am not sure what theory you are referring to n10sivern.

But the only thing i have done is i have cut out the potential voltage drops from the extra resistance, from the driver, driver board ground and led - wire.

You are thinking about the circuit is to restrictive, every connection in the circuit is an intersection, and every intersection you can widen & or remove will lead to a faster journey for the electrons & less loss as heated wires.

And to continue you analogy, that is like taking a short cut over a higher speed lane, because of the less potential resistance & therefore less voltage drops on the way & therefore a total of more potential wattage that goes in to the led.

It wont get brighter than a battery hooked right up to a well heatsinked led, but it will get closer because i have cut out some of resistance in the circuit :wink: now one of the biggest limiter left is the switch for example & the unsoldered switch ground.

The less resistance there are in the circuit the more of the voltage & current the led will see, and therefore the more wattage it can consume, we are limited by the high Vf & the starting voltage of the 4.2v batteries we use.

I was referring to soldering the negative to the pill. I had a brain fart for a second and didn’t think that resistance in a circuit is cumulative but it is, so that is one piece of resistance removed.

No worries :slight_smile: it always nice (for me) to be questioned in a friendly environment, it makes me solidify & express my own thoughts about it more clearly even for my self.

Now I want your results even more... Maybe you could borrow light meter or go to some professional photographer that will take measurements for you?

Cajampa, just get a HS1010 lux meter. They are like $12-13 at fasttech.

I started a thread about that. Figured some things out, did some initial tests. I would not bet my life on the accuracy, and you will have to calibrate them, but it looks like it is usable.

When there's time and motivation, I will continue testing on my smartphone luxmeter.


I have been reading up on those today, it seems the general consensus is that they read a little low about 5-10% compared to better meters.

I found the HS1010 on ebay for $10.18

And what is according to the listing pictures at least is the HS1010A for $13.20
The listing description says HS1010, but they have sold 324 of it so far, so if the picture was wrong it should have been fixed by now on account of all the disputes.

What i wonder is if it is worth the added cost to get the reader on a wire instead of it being on the meter it self? Is it even possible to read the measurement while you are targeting it if it stays on the meter, it seems like it would wash out the display.

the wire is very useful. The problem with these meters is not only that the calibration is not correct (within one brand and type it can already vary), but the main fault is that the wavelength response is not correct. That said, if you stick to measuring neutral an cool white flashlights, that fault stays within a few percent.

Thanks :slight_smile: djozz

That is good to know, for throwers i aim at CW & NW anyway because of thinner phosphors, and the potential of collars, that i hope to play with soon.

It’s nice to know that I’m not the only one here who blew a fuse on their DMM. Those 11 amp fuses aren’t cheap either.

I wish i had a fuse in my DMM’s 10A measuring circuit, but no when it blew/shorted or whatever happened the shunt just burnt off :frowning:


Thank you for taking the time to do all this testing. I was sort of waiting to see everyone’s results before ordering any components. The numbers for the XM-L2 are impressive. Would you say it has a more usable beam than the XP-G2? I was originally going to go with the XP-G2, but 400 kcd on the XML is fine in my book.

It’s bigger. I didn’t actually do any beam shots with these except for the lux meter.

I thought maybe I should point you guys to something that can explain goofy lux results. Color temperature alone can fool a meter one way or another. With those cheaper lux meters, you’re really only making a comparison well when you have two tints exactly the same. Even when you do, that’s not saying the result is nearly accurate to what a lux value is. LEDs are one of the most difficult sources of light to properly read and measure in lux. Because of LEDs varying spectral emissions, and the sensor having points where it cannot detect all wavelengths, when a 1-sensor meter is used be wary of results. If a 1-sensor meter is used, it needs to be calibrated using a filter that helps the meter achieve a reading within the photopic vision spectrum curve of humans. If it is just a cheap meter, it will look exactly the same, but it will not have the proper filters inside to read LED light well and emulate the photopic curve. As you may recall, photopic means vision during bright, or daylight conditions. The CIE 1931 standard is based on the photopic curve. Scotopic is the vision used at night. Mesopic vision is basically using both those visions combined, in dim situations.

That basically explains what type of curve lux meters should be sensitized for—something that’s not so simple to test with 1 sensor. What the meters are actually picking up and reading may somewhat shock you. Some say, to calibrate a meter with a 60W incandescent because the value is known. Unfortunately, I found out that it cannot be done like that. An incandescent light, using heat to release light via black body radiation, has an extremely broad and solid spectral curve spanning from blue and going well into the IR region. An LED has typically only 2 peaks in its spectral signature (1 near 450nm and one in the 540-580nm area). The two light sources are very far apart in color spectra.

This article I found explains a pretty good reason for large variance with the 3 types of lux meters (cheap single sensor/calibrated-filtered single sensor/multi-sensor): http://www.olino.org/us/articles/2009/12/07/measuring-illuminance-correctly

Now I just have to buy a multi-sensor unit. J)

Luminosity function from wiki, black is photopic (day) vision, green is scotopic (night):

I did some tests on luxmeters that I posted here on BLF last year. The tests were all within my limited hobby possibilities at home, but perhaps nice to read too? The conclusions were quite like the tests in your link.


Your test is the reason I own the CT1330B.


So I put an S4 2B in a 1405 with a 5 amp driver, dedomed, with a collar and measured 470kcd. I’m confident now that the difference between our numbers is due to the meters. I’m willing to put together a light with a battery and charger and pass it along to a few select members so we can get some baseline numbers so we can compare our results accurately. I have a 1504 with an XP-G2 and no collar that would be a good candidate.

Each person would be responsible for shipping it on to the next.

I have been thinking about this idea of yours KKW, and what if we would take it to the next level :wink:

What if BLF had a light with a CW, NW & WW led in different pills for example a 1504, and we all sent it around to calibrate our lightmeters. Then we should be able to devise a BLF lux & basically be able to use very cheap meters but calibrated to the official BLF conversion factor, that is this light.
Someone could even sell “pre calibrated” already checked conversion factor lightmeters.

I am not sure how the maths would work, but i think it should be possible.

It’s been talked about before, I think the problem is usually too much interest. I like the idea of multiple pills.

I’d probably limit it to about 6 people with a goal of about a month. I hope n10sivern is interested, and I’d like djozz to be in on it, but shipping to Amsterdam might be cost and time prohibitive. I think anyone with multiple meters and a history of doing testing here would be my first choice for candidates beyond that.

I’ll start a new thread to organize this.