That reference page is golden…doesn’t quite show everything but it puts you in the ballpark and most of what we use in lights is on there. That said, in cheapie lights you’ll often have mystery emitters and many made good effort to replicate various Cree models (sometimes very good fakes). Latticebright models were (are?) a popular cheapie emitter found in a great many lights. I think we have a thread on here somewhere about identifying fake Cree that had great Latticebright examples, if you search for that one. If you can’t figure out what you have from that reference page, people here will help if you give photos (or maybe the info is in the specs/reviews for the model of light), or you can look up the manufacturer datasheets and look toward the end for the spec drawings which give measurements and also a view of how the contact traces on the bottom of the emitter are laid out.
So visual familiarization and die size are best, but you can use a good multimeter for some basic tests. The diode-check function on many meters can be used to see if the emitter is good, although for our higher power and SMD emitters the function on the meter should deliver at least 2.5v or more. Most meters use an unhelpful diode-check that only has a half volt or perhaps up to 1.5v so that’s only good for bi-axial old school led’s (and for normal diodes of course). You can also check for open and reverse voltage/leakage with this test but we don’t see emitters fail like that often, so just the basic light-up test is generally enough. (You can also do this with a battery and a couple leads/wires that you hold by hand if your meter doesn’t give enough oomph on that function.). When doing the test, just connect things briefly to see it light up. If you hold them and allow the emitter to keep emitting, there’s a chance that it may overheat and suffer damage or death if it is not mounted to an mcpcb/heat sink. To do other performance tests you need to test while in the light itself and/or set up a test rig using a bench power supply and such.
That’s a neat site, love seeing the machines and processes…thanks!
I was told by a Samsung rep years ago that they did charge and discharge new cells off the line. I don’t have the email anymore but it was probably about 2014 or so. I don’t know if they’re so easy to get in touch with these days but I used to have great luck communicating with them, even got a storage/SSD systems engineer one time that provided a ton of fantastic information.
Maybe a quibble but a point to add is that every battery - no matter who makes it - must receive at least a forming charge, else there is no finished product (can’t do that at home). Probably not what you meant, though.
Thanks, Correllux! That answers many of my questions. Makes a lot of sense, for Lithium-ion cells to be at a lower charged state for safe transport. So they likely have an automated routine at the tail end of the production process prior to packaging.
I’d bought a bunch of new cells just recently and their charge levels were a bit mixed. The two 21700 cells, different makes and mAh capacities, were showing initially 3.4v and 3.5v. Some of the 18650’s were at 3.8v. That’s what triggered my curiosity. Anyway, I’d forgotten about Battery University and will check the links on best practices. I can go some months without putting some of my flashlights to any serious use, so it seems a sensible idea to be mindful of the battery charge level and adjust accordingly for longer battery life.
Yeah I meant when the can is first welded shut. That site seems to be more in the business of producing battery packs rather then trying to sell individual cells. I would assume all manufacturers go through a couple of cycles before packing and shipping. The reference to grade A, B and C used to be more common. It’s not something you hear about often now. Makes you wonder sometimes?
I also wonder about variances in production methods by cell type. 18650’s are a cash-cow cell, first made by Sony in 1991. The method of manufacture is mass batch and streamlined. What of 21700 cells? First made by Tesla in 2017. I imagine it’s licensed and made by a variety of manufacturers. I imagine the assembly process is very similar to the 18650. But outside of battery pack building done by Tesla, maybe 3rd party makers do things a little differently from the 18650? A little more manual process?
It is relatively easy to unsolder the two wires from the mcpcb and swap in a new mcpcb that has your new leds installed. This mainly comes down to whether you can buy the leds you want on a mcpcb that fits. This has the advantage of changing the led footprint such as swapping from 3535 to a 5050 led.
If you cannot get a loaded mcpcb then you are limited to reusing the stock one. This means your new leds have to be the same footprint and you have reflow solder it. This is much more complicated and usually requires a hot air station or a hot plate. It’s a more advanced technique.
So you have to do your research and look for the parts you need. In the US, Mountain Electronics is a good source of leds already reflowed onto mcpcb. He can mix and match whatever he has in stock. Otherwise you’ll have to find a preloaded mcpcb probably from overseas.
It’s definitely smarter and cheaper to buy a solder iron and learn the basics yourself. Then you can at least swap parts and fix wire connections, etc… Having to ship lights to people for simple swaps can get very expensive, be time consuming and you risk a package being lost or damaged in the mail.
I think it’s also worth investing in a test LED for trial and error reflowing. One you would not care for breaking in a cheap light. Also find some junk electronics and have at it with the soldering iron to learn. Just make sure it’s unplugged and there is no battery with what you are working with.
Will the Sofirn SP36 BLF ever get Anduril 2 from the factory? I am a bit of a newb and don’t want to attempt to flash anything myself unless there was a dedicated flashing port, etc. I have Anduril 2 in the BLF LT1 and the Wurrkos TS21 and I am looking to continue with that as opposed to trying to learn Anduril 1 at the same time. Thanks!
Only Sofirn can answer that. They probably buy the drivers in batches so they may have hundreds still on hand. If they get low they may order the next batch with Anduril 2. In order to do that they would need at least one of their software employees to rewrite the code for this older light instead of writing code for newer lights. Or they would have to pay Toykeeper a fee to modify the code. They may not want to spend the money on an older light. So it seems unlikely they would do that.
If the light is still selling well, Anduril 2 becomes highly asked for and they don’t have a new light being designed to replace the SP36 - then they might upgrade it.
My guess is it would be unlikely to happen. You might sending an email to Sofirn to see if they know.
As the year 2021 is ending, I wonder what will happen to those unsold 2021 flashlights. I think they will not be able to compete with newer, better-designed, and possibly cheaper flashlights of 2022.
And what has happened to those old stocks of 2020 and 2019?
Same as any other product. They will sit in a box until its sold. If sales slow down they will go on sale. Eventually it will get sold. This is why they are built in batch sizes based on how they are selling.
Yeah, my cut off range would be closer to 4 or 5 years. Even then, there might be a cool design that’s really old that can be modded. Or you might find a low powered light known for its nice color and tint and it fits your needs. Why not buy it? So there’s way more factors involved than just age.
How do you guy’s characterize beam profile? I understand terms like flood and thrower, but was wondering if there are some more quantitative definitions. Do you define angular beam spread by the contour of 50% max luminance, 50% encircled flux, or just what kinda looks good?
A more down-in-the-weeds geeky question: When posting beam shots do you use conventional JPEG imagery which converts detected luminance values to display levels using a nonlinear factor (gamma) that reflects human visual response or do you see some value in using linear RAW conversion (gamma=1.0) which produces output levels accurately reflecting the actual luminance detected? I’ve done beam shots both ways and the results differ markedly in appearance.
I’ve only been at this game of characterizing lights for a short while and don’t want to spend a lot of time figuring out issues which you guys settled years ago.
I don’t “characterise” them except descriptively. Eg, a beam can be “ringy” (rings or just a “bullseye” pattern), can have a nasty corona around the hotspot, can be “fried-eggy” (blue spill, yellow hotspot), have a “cloverleaf” corona/spill, etc.
Or just “Noice!” or “Eww.”.
Some people do go all-out with snapshots of the beam, put little ’x’es in various places and show the color temp at each spot, etc., but that’s beyond my ken, and my barbie.
Basically, the settings you see are the upper limits of measurement. They are multiples of 2s times 10 to some power. On the left are DC measurements, the right are AC measurements. So to left would be the 1000 Volts (’cause it isn’t rated for higher isolation/components), 200, 20, 2 (which they represent as 2000 millivolts) and lastly the 0.2 (200 millivolts). The same would go for the AC scales (only two are available for common voltages).
Below the DC volts are the resistance settings. Never to be used in a live circuit. Either blow the internal fuse, but this model would fry the insides. On the right are the Amps (current) drawn. Only DC measurements and you need to switch the black red lead to the ‘10 A’ hole*. Below that is the frequency and the diode (continuity) test. That last one may have an audible sound when the circuit is below some set value – usually less than 1 Ohm. Frequency is on live circuits (like household plugs and receptacles), the diode test is a resistance test, done on NOT live circuitry.
*edit. Strange they have the 10A receptacle in black. I stand corrected.
Never worked with series and parallel circuits I presume. You’ll need some beginner’s guide on how to make measurements of voltage and current.
I hope you didn’t pay much. If you are a beginner, most probably will make the common mistake of turning the switch when plugged into some live circuit and WILL burn your device. That is why these things are still sold, for beginners and not too expensive.
The V~ is your AC (alternating current) like in your house.
The V-… is your DC (direct current) like in a battery.
The different numbers are just the max ranges. You could measure the battery voltage at 1000 volts, but you won’t have enough accuracy.
The 2000m is 2000 millivolts (or 2 volts). Thats not enough for your battery so 20 volts is the best option.
The horseshoe is Omega or ohms. That measures resistance.
The A is DC amperage. You have to be careful using this. Or at least the 10 amp setting which requires you to move the red probe to that other port. Try not to use that at all and leave your probes in their current positions.
I would have recommended a different DMM. One that auto ranges and auto turns off. This one looks like to have to remember to turn it off manually. If not, you’ll run the battery down and that sucks.
And a loose driver causing flickering. Either thinner PCB (1.5mm versus 1.6mm) or something making partial contact on the driver rim whose retainer cavity isn’t threaded all the way down (intermittent contact when screwed down).
Did you get that fixed? A series of unfortunate events…
