My Yezl Y3 Has Arrived!

LED Flashlights – General Info 26650 Flashlights
78

Ok, here's some results ....

Tailcap amp measurements (using 14 gauge high qual custom leads):

TrustFire 26650: 2.45A

KK ICR 4200: 2.51A

SONY 50A 26650: 2.52A

Efest 25A 18650: 2.51A

Samsung 20R: 2.50A

Powerizer: 2.50A

Amp measurements are pretty tightly grouped, so suspect it's fairly well regulated to 2.45-2.50 amps. Most of these cells are pretty hot cells, so there is definite amp limits in this driver.

Using the KK ICR 4200 for output measurements, Yezl Y3, XM-L2 T6:

lumens: 901 @start, 874 @30 secs, throw (from 5m): 62.6 kcd

Reference stock HD2010 TangsFire w/TF 26650, XML T6, measured 3.6A tail:

lumens: 908 @start, 898 @30 secs, throw (from 5m): 57 kcd

Certainly the HD2010 at the higher amps performs pretty good, but the Yezel still beats it in throw. The throw to lumens ratio goes to the Yezl, but of course it has the XM-L2 vs. XML, so it's different, but I don't think it has an impact on throw. My for-the-moment conclusion is the Yezl is the better thrower, probably attributable to the reflector, but by a small margin.

You cannot overlook the results here... A totally stock light doing 62.5 kcd on one cell at regulated amps of 2.5 is doing pretty good.

Beam Patterns

No pics yet, but compared side by side to the HD 2010, patterns are very similar - bright intense spot, surrounding flood zone. The spill zone is slightly smaller on the Yezl, to be expected because it appears to have the deeper reflector. Also the hot spot is ever so slightly smaller than the HD2010's, indoors that is.

There's a ringiness to the Yezl's beam, not in the HD2010's, both around the hot spot and at the outer edge of the spill. The HD2010 has the flower shape halo, but transitions are smooth. On the Yezl, the flower shape is not as defined - it's more round, but the ringiness is there.

Now about beams... This is with these 2 units I have here. There could or may be variations piece to piece about how the bezel and o-ring fit/mounted, how the reflector sits on the LED, etc. Also the Yezl appears to have an alignment piece that could block output, while the HD2010 has a pretty thin edged, low profile alignment piece - I won't know for sure until I dis-assemble it. Also, I could have caused the ringiness when I loosened the bezel to fix the bumped out o-ring. I had to use a strap wrench to loosen the bezel, but then I hand tightened it afterwards. To me, this level of ringiness is not so bad, definitely acceptable. I've seen much worse.

mudgripz - re-tested lumens with a Panasonic PF 18650 and get 901-867 lumens -- nearly identical as would be expected for an amp regulated driver.

Lens - definitely glass, size: 56.8mm x 2.04 mm

HD 2010 lens as comparison, size: 58 mm x 2.04 mm

Reflector - aluminum, size: 57 mm OD, 54 mm ID, 41mm depth

HD 2010 reflector as comparison, size: 57.75 mm OD, 51.5 mm ID, 38 mm depth

Hope this answered some questions. I'd say Joe was pretty dead-on with this light. His results of course may be different in what he saw, depending on the cells and amps being delivered in both the Yezl and HD2010 he used as a reference, specially the unregulated HD2010 that varies wildly depending on the cell used.

79

Yikes, I think I killed Joe's thread! Maybe post #77 was little too much?

I tried a resistor mod similar to what DayLighter did - calculated it to 4.8A, but so no improvement with one cell. On two cells though, I got like 3.5A at the batteries (??), so I'm thinking the bank of resistors only kicks in when the driver is bucking down voltage? Weird... At this point, I'll abandon this idea and go with an e-switch version of a Nanjg, maybe single FET ala comfy.

80

LOL… 4.8A for single LED is not enough for you Tom? :smiley:

81

Sorry - didn't make it clear -- the resistor mod did not work at all for a one cell configuration - still got 2.4A or so. Tried couple of other resistor mods to other resistors (higher values found on the board) using 50% value resistors, still, no improvement. Ugh - this driver is beyond me... Better for me to go with a custom FET Nanjg driver because it should get over 6A. Thinking this host can handle the heat well. The pill top is pretty thick and where the pill threads in to the front bezel, there's pretty good thickness of metal to get to the cooling fins.

Only bad thing is the lens is an odd size but thinking I can sand down a UCL/p lens to make it fit -- never worked on a UCL/p, so have to find out if it's do-able.

82

I’m pretty excited for this light. I imagine I’ll be happy with a well regulated and bright stock light for about a week or so before I have to either make it a hot rod or break it completely.

Tom- please keep us updated with your progress with the FET based driver. I’ve never messed with an electronic switch before, so I’d love to see how the masters do it.

Edit- Can you use a FET driver with multiple cells, or will you have to run only one cell?

83

Ok, will do - I got the driver built, not tested yet, waiting on the JB Weld to dry that's used to secure the FET.

I programmed my custom e-switch driver on it already. Time to bench test...

- This will be limited to one cell unfortunately because it's a simple Nanjg/FET based driver. Could do a Zener mod though, run 2 cells, driving an MT-G2.

84

Since I’m planning on using multiple cells, I might try to use the resistor mod to bump up the current. You said that that resistor bank only seemed to be in use when there are multiple cells, correct? If so, perfect! What resistor did you use to get 3.5a or with multiple cells?

Also, how easy is it to get to the LED? Was the head glued? Thanks!

85

No glue on the head, but used one strap wrench -- maybe you could use a jar opener pad...

The 3.5A at the tail with 2 cells is 7A to the LED, so, the resistor I used was a bit overboard. Think I piggybacked an R100 on to the existing resistors -- pretty sure there were 2 resistors in parallel, both under 1 ohm, so easy to spot. Probably need to use an R150 or R200 -- something higher to make the bump lower.

86

there’s 2 resistors on the board… it’s R150 and R180
you can open up the bezel with rubberized glove, but to access the driver or to change the LED… you don’t need to open up the head… the light came in 4 sections
the head, the driver/LED pill, the battery tube and the tail

87

Finalyy got the upgrade finished for the most part. Piggybacked the FET based Nanjg to the stripped stock driver because of the mounted e-switch. Lumens: 1,717 @start, 1,656 OTF at 30 secs, 125 kcd throw on a Powerizer, 4.21v, 5.72A tail. Springs copper wired, using 22 gauge wires to the LED, 22 gauge for the + wire from stock to piggyback, 3 24 gauge wires connecting ground between the boards, XM-L2 U2 1A on Noctigon using AS5.

Wow! The PWM value of 1 is pretty incredible - I measure 0.6 lumens. Here's the break down:

PWM Values - Lumens

1 0.6

4 16

16 92

100 663

255 1,656

Issues:

  • Thinking the 5.7A is lower than expected - could be the piggybacking with extra resistance, or maybe how I'm wiring the FET, dunno
  • an AR lens should help big time
  • will re-test with a charged up KK ICR 4000 or 4200

The 5 mode switching is working amazingly:

  • no mode memory
  • quick click, turns it ON in moonlight (very low moonlight, lower than what you can get with 7135's)
  • long quick (hold) of 3/10 of a second turns it on in high
  • once ON, quick clicks go up, long clicks (holds) go down
  • if you pause after changing modes for more than 1.2 secs, a quick click then will turn the light OFF

I'm finding the .3 secs and 1.2 secs times are working out well for me in this light. Here's the summary:

  • From OFF: single click to lowest mode, long click to highest mode
  • navigating modes is very easy: click to go from low to higher modes, long click for high to lower modes
  • after a pause, single click is OFF
88

awesome… ! :beer: :beer:

89

The 125k sounds ok but not compared to say a modded 2X or the like. Is the project torch not yet dedomed? - that could jump it to 250-280k?

90

Right - still with dome, that's why the high lumens #'s, and expect to see that range of throw when de-domed. This light is more a HD2010 replacement/upgrade for the same cost, and for that purpose, it's just about perfect - better quality, e-switch, better throw (a little). I don't see it competing with MaxToch 2 cell lights.

91

But does the userfriendly mode switch still work?

92

What mode switch? You mean my customized version? Yes - of course. It's pretty awesome.

93

I meant the stock white button for changing modes. The big selling point of this light, for me, was the “no worries of it going into any other mode” feature. But no doubt, what you have is even more awesome (please do some beamshots!)

94

Sure - the whole point was to utilize that switch. The tail switch now is simply is a "lock-out". Everything can be done on the side switch, but nice to have total power cutoff with the tail switch.

95

Do you have any suggestions for a small, floody e-switch light which could use a similar firmware? (and any extra parts required to get it all working)

I’m hoping to make one which uses an e-switch to control PWM speed in very small increments, to make a variable-speed strobe light. I think a UF-T50 might be a good starting point, but I haven’t figured out where to go from there. (my background is in software, not hardware)

96

The Small Sun ZY-T29 or ZY-T11 clones. I bought this one: http://www.banggood.com/CREE-XM-L-T6-1800lm-5-Modes-Zoomable-LED-Flashlight-Black-p-80813.html, but received the UltraFire, but it's the same thing. Cheap - I modded it to 3.5A, XM-L2 U2 1A on a Noctigon, with the e-switch custom firmware. The XP-11 at WallBuys.com I've also mod'ed but it's a different setup internally. Check out the T29 and T11 threads for details. The Small Suns have plastic lens reflectors unfortunately, at least at the time, but the UltraFire and Aleto's don't.

It's certainly not a simple mod - ZY-T11/ZY-T29 require to piggyback a Nanjg onto the existing driver because of the switch mounted on the stock driver. For the XP-11, the stock driver can be tossed and replaced with a Nanjg, but wiring in the e-switch can be a challenge, unless you are ok with very fine soldering.

Check post #19 here: https://budgetlightforum.com/t/-/21292

97

Chris at flashlightlens.com-UCLp_AR just listed a new size I requested for a UCLp: 56.8mm x 2.25 mm, that I requested. He'll basically do any size upon request for the UCLp, not the glass UCL. The stock lens on the Yezl Y3 is 56.8 mm wide and 2.04 mm thick. I'm assuming the 0.21 mm extra width will be ok, specially when swapping out the stock LED for a XM-L2 on copper, because it's slightly thinner. I'll be ordering one to try it out.