light-wolff,
Correct. At the solder points on the MCPCB.
TomE,
I'm just glad others are chiming in. Especially since you also have very similar results with the U3 bin and the higher VF. The more and more tests I do, and the more information I get here the more I suspect that I was shipped U3s by mistake.
I think if RMM gets results from his stock that reflect my original test results I will just end up order new emitters from him.
Think'n that could be correct - performs like U3's...
I've only found out very recent how low the U3's are compared to the U4's - had no idea. Really very disappointing for me because I just bought recently qty 10 of U3 3D's from IOS, and still have a bunch of U2 3D's thinking the U3's would be better. Like the tint but but being 1.0 to 1.2A lower than U4's is an additional penalty for that tint.
Slightly off topic, but ate you guys finding U4 bin to have a higher brightness than v6 when using a FET driver, all else equal?
Barring the evidence we apparently have, why and how could there be such a significant difference in the U3’s? After all, isn’t the U3 just a bin they assign after manufacturing? Were the U3’s all from a single batch that somehow got messed up with a higher Vf and Cree just deemed them good enough for rated spec?
Yes - I'm finding the U4's brighter than V6 XP-L's. Umm - diff in U3's? Well, newer T5's, U2's etc. are also low amps and high Vf like the U3's. U4's on the other hand seem to be different.
The bin assignment was theoretically that: pull the LED's off the line, test them, throw them into the appropriate output bin. But, we don't really see this - within a given time period, what we are allowed to buy, released to our somewhat gray market, is very specific tints in very specific bins. A bin though is a wide range, so maybe the CREE manufacturing process is so precise now, you don't get diverse bins spreads.
Cant really say why and how for sure. I know that around the T6/U2 bin something was different with the bond wires. Driving those bins to 6 amp was no problem. When we hit U3 I not only noticed the spike in VF, I also found that the bond wires would fry at lower current.
This is just an observation. I use about 400-600ish XML2s per year so its hard to judge exactly what is going on for say a reel to reel basis and such.
The bins are all bout efficiency - lumens/watt. I think CREE made a design change to get higher efficiency in the U4's and may explain why the Vf is lower.
Any discussion of measurement methods here is a moot point in my opinion. As long as the results consistently relate to one another (apparently they do) all the questions like “did you take the readings from your DMM or the PSU” and “did you put the DMM right on the LED” become irrelevant. While in our hobby a difference of 0.1v due to measurement method can be very relevant, that’s not the case here. As long as the method remains consistent the results compare properly to one another. VOB’s measurements on one set of emitters are repeatably lower than on another set. At the end of the day it sure sounds like a bad batch for whatever reason since nobody else has chimed in to say “oh I had that problem too!”.
RE the binning vs manufacturing methods conversation… as mentioned above, the waters are really too murky from our perspective to draw strong conclusions but it stands to reason that a lot of what we see is based on how few LEDs actually flow through the grey market. When we notice something like “U3’s have a high Vf” what is possibly more accurate is that the U3 bin required a higher Vf when it was a new cool-white bin. That’s happened before with other LEDs, bins, and manufacturers too. After whatever advances led to the U4 dropped the Vf back down this probably also applied to any new U3 production. Without a date code on some modern U3’s it’s impossible to say. In other cases Cree has come up with a new technique and then applied it where they could. I could be way off base here, and I’m definitely not speaking based on personal experience with either bin (as many of you can guess, hehh).
Here’s my 2 cents. Built a Convoy C-8, with all the tweaks, FET driver, and VTC 5 cell. With a hot de-domed U3-1A (from Richard I believe) pulled around 4.3a. Removed the U3 and replaced with a hot de-domed U4-1C, from Richard, and amp draw went to 5.1-5.2a. Lux increased 20-25kcd. I did this a few weeks ago and forgot the exact numbers, but I’m close. Not overly thrilled with the 1C tint. Awaiting delivery of some U4-1A from James@ 3Tronics. If the 1A has the same output as the 1C without the yellow tint, I will be very happy! BTW my current measurements are done with my Fluke 336 clamp meter.
Hi, my last build is a Convoy M1 + FET 1 driver (from bangood) Noctigon Xm-l2 U3 1A (from IOS) and i´ve got 4,3 amp
The battery is Pananasonic/sanyo NCR18650GA 3500mha and ths springs are in stock form (not bypass)
How much amps can i got with a better battery (less resistance) and springs mod?
Hope this help.
Alberto.
Usualy 4.7- 5amps, FET+1 driver from BG isnt a good option though, it has tons of issues: the best battery for a DD fet driver atm is Sонъ VTC5A or if you want best runtimes Samsung 30Q
What´s the issues with the Bangood fet +1 driver?
Anyone know a better springs (Beryllium copper) or similar for flashlight that will be better than the stocks (convoy m1 and BLF A6 Fet + 1 driver) ?
Thanks in advance.
Alberto.
Djozz did some interesting work re. springs here:
I use the Fasttech ‘carobronze’ springs to avoid the need for a wire bypass.
Not the best, but much better than steel springs and does not heat up at moderate-high currents.
https://www.fasttech.com/product/1347100-batterydriver-contact-support-spring-for-flashligh
Big issue with the A6 drivers is probably the off-time cap and related firmware. The short/medium/long press duration all gets stretched as driver board heats up. Build quality may also not be consistent. I tried a few and they are OK, but this is with my own firmware and not using the OTC. One pleasant surprise was that the stock wires are silicone-insulated and the spring is non-magnetic (so not steel).
Anyone knows what metal is used for the springs on these drivers?
I believe what Mitko is talking about is high resistance and low amps, compared to other DD FET drivers.
Not only that TOm, i got 3 of those drivers, they lose modes on a random base, they have a moon mode issues that i mentioned early in the BLF A6 topic yet since i do respect TK efforts and the others that worked hard about A6 project i only mentioned it, and yes, they have kinda high resistance too
WHy the hell they changed the design i dont know…whatever
Any reliable source of U4 1A? I have never tested that emitter though
I got U4 1A's from Cutter - blew out the first one in the BOSS1 (believe the bind wires failed) when running off of 4 EFEST 4200's, it survived on one EFEST 4200. I got two left of the U4 1A's - not many, haven't tested.
For the BLF A6 drivers, I'd think the FTE is suspect, and definitely the caps must be high quality and think they skimped on costs with them.
Interesting Mitko, TomE.
Probably I got lucky, 5+ A on an XP-L HI V3 without much effort. The MCU seems stable as well (not using the stock firmware).
But I suspect the assembly quality of the A6 drivers vary wildly.
As much as I hate to skew off topic I will have to chime in on the BG FET drivers as well.
I really thought this was going to be what we had all been waiting for. A ready made mosfet driver. Not so much. Certainly better than anything that was available before, but I too noticed the higher resistance vs ones made using OSH park boards and BLF suggested components. In a setup that would typically draw around 3.2A I got only 2.6 ish IIRC.
I have not even tried them with the stock firmware (not why I bought them). The only UI issue I have had that I know of after flashing with a version of star firmware was a very long amount of time it takes for the memory function to take hold. Like 10-15 seconds.
Using said firmware the moonlight and other low modes run through the 7135 were great. All in all still a good driver, just not nearly the ready made unicorn we have all been dreaming of.
OSH is now offering 2oz traces and I will have those in hand in a few days, if all goes well asflashlights will be selling mosfet drivers by next week :)
My a6 FET driver is working fine. All the funtions working well and no issues so far. Maybe is the lastest bach? cause i buy a month and a half ago.
BTW mythe driver that i received have great silicone wires and the negative contact ( for driver ring) its bigger that the bangood photos.
Thanks to all.
Alberto.
I wish I had heard about these A6 driver issues beforehand as I have 15 of them in the mail.