Here is an upgrade I did on my nearly new AKOray K-106 3-mode programmable light. I liked pretty much everything about the light except that it wasn't as bright as it could be. So I bought a Cree XP-G R5 LED from KaiDomain and put that in there. The upgrade itself was pretty easy because the original LED and board were not even affixed to the pill. So I just had to unsolder the two wires to take out the old one. Once I had that out, I found a decent size hole, maybe 10mm across underneath. That seems kind of silly: here's a board that gets pretty hot and they put air underneath it? It would be better to fill it in, I guess, but I just wanted results. I put some STARS thermal compound that I got from DealExtreme and soldered the new LED in place. I didn't want to use an epoxy in case I didn't like the results and, I don't have the FUJIK thermal adhesive in yet from another order form DX. It was a little tricky getting the LED centered and I still don't have it quite right (I'll add some pictures tomorrow).
Here is the LED soldered into place (the little drips were there when I got it, the fried looking black lead is me getting sloppy with the soldering iron):
Here is a look in the reflector. The hole at the bottom of the reflector is made for a larger XR-E so there is space around it and the LED probably doesn't sit up quite as high in the reflector as it should. Notice it is also a little off-center. When I took it apart to take these pictures I was able to center it better. I also knocked down the overly large blob of solder on the red lead that you can see above (it was flattened by bearing against the bottom of the reflector).
Results? The reflector is made for a Cree XR-E LED so the hole is bigger than it needs to be. There is a dark blotch on one side of the hotspot, maybe because the LED is still a little off-center (doesn't really show up in these pictures, but it is easy to see when the beam is moving). The increase in output doesn't seem that noticeable. Here are some side-by-side shots vs. a Fenix L1D (with a Cree XR-E Q5) before the upgrade and after:
On NiMH batteries. The first picture is the upgrade, the second is the original. At 1/25th second:
Same thing at 1/100th second:
Same thing at 1/1600th second. The hotspot is bigger with the upgrade, but loses some intensity. I think before the L1D was brighter and now they are closer to being equal. So a little bit of improvement, but it's not dramatic.
Now with a fresh lithium-ion battery in each at 1/25th second:
Li-ions at 1/100th second:
Li-ions at 1/1600th second. Same sort of results as before. Whereas before the K-106 lagged a little, now it is closer to an even match, with 106 having a bigger hotspot. In these pictures, the beam actually looks like it improved, losing the donut of darkness that is typical of the XR-E. I could try putting some shiny foil underneath the reflector to see if that helps the new beam.
Now for some test outdoors. These are 4-second exposures. The stick in the ground is 25 feet away and the beam is focused on the fence posts at the center of the picture which are 120 feet away.
Here is the L1D on a NiMH battery:
Here is the AKOray. It's a little blurry, but you can tell there is less throw because the fence posts are almost invisible. The L1D has a smooth reflector which really helps throw:
Now here is the AKOray with the original LED, taken a couple of months ago:
Here is a zoomed in detail of the fence posts for the L1D, AKOray R5, and AKOray Q5 original:
Now the L1D on a li-ion battery:
And here is the AKOray on a li-ion:
And the AKOray with the original LED from a couple of months ago (kind of hard to compare since there were no leaves then):
Now here is the detail of the fence posts for all 3: L1D, AKOray R5, and AKOray Q5 original:
Battery Life
The battery life should be the same, I would think. The driver should be regulating how much voltage or current is going to the LED. But in my test it seems like the XP-G draws a higher current, but that doesn't shorten the battery life too much (I paused the test at 2 hours, 26 minutes to check the battery and it was below 1V, so I stopped testing and couldn't get a reading on the current). This may have more to do with the state of the battery, though both times I was using a 2000 mAh Eneloop.
Time (hr:min) | Volts (XRE) |
Current (mA) | Volts (XPG) | Current(mA) |
---|---|---|---|---|
0:00 | 1.45 | 1190 | 1.345 | 1220 |
0:15 | 1.31 | 950 | 1.281 | 1140 |
0:30 | 1.28 | 880 | 1.276 | 1130 |
0:45 | 1.26 | 850 | 1.270 | 1110 |
1:00 | 1.26 | 840 | 1.264 | 1080 |
1:15 | 1.25 | 710 | 1.255 | 1070 |
1:30 | 1.24 | 930 | 1.248 | 1050 |
1:45 | 1.23 | 780 | 1.229 | 990 |
2:00 | 1.21 | 750 | 1.200 | 930 |
2:15 | 1.18 | 720 | 1.147 | 770 |
2:30 | 1.11 | 560 | 0.970 (2:26) | n/a |
2:45 | 0.92 | n/a |
On a lithium-ion 14500, it looks similar. The current draw is higher and the battery life is shortened a little (maybe 5 minutes if I had stopped these at 3.60V):
Time | Volts (XRE) |
Current (mA) | Volts (XPG) | Curent (mA) |
---|---|---|---|---|
0:00 | 4.21 | 610 | 4.19 | 670 |
0:15 | 3.91 | 660 | 3.88 | 750 |
0:30 | 3.69 | 720 |
3.65 | 820 |
0:45 | 3.55 | 770 |
3.52 | 860 |
Summary
This was a decent upgrade, but spending $10 to upgrade a $20 light and improving the output only a little bit probably doesn't make any sense. At the same current, the R5 is only 22% brighter than a Q5 (and I have not way of telling if this LED is really an R5 since R3 and R4 LED's don't look any different). It would be good if AKOray would upgrade the K-106 to a XP-G, drive it a little harder, and add a smooth reflector designed for this LED. Then they would really have something.
I have a couple of XP-G R4 neutral tint LED's on the way. If the tint turns out really good, I might put one of those in the K-106.