My Bike Lights - Info, Teardown pics, Beamshots, Current Measurements, etc. . .

You’re setting yourself up for more grunt work :-).

Interested to hear your thoughts on the 3x XML2 light

Which one? You mean these? (Though they're usually XM-L.) Or do you mean the top secret one I have coming for review? (Shhh . . . don't tell anyone :) ).

-Garry

The one I posted about on MTBR with the purple trim ring. High>Low>Strobe, mine says 2400lm, yours is listed at 3600lm. We both know they will be about 1000lm tops.

You said you bought yours from Gearbest, I’d get mine from Amazon.

Ah, yes that one. It's on the slow boat right now (tracking isn't showing anything). I'm kicking myself now that I didn't ask to review the one Cat recommended trying in that thread. Actually, that light might not have been on Gearbest's site when I chose a light to review.

-Garry

UPDATE: New Flickr Album Link: https://www.flickr.com/photos/garrybunk/albums/72157662737583937

Just received my GearBest 3x XM-L2 bike light for review light yesterday. I hurried up and grabbed some beamshots last night while the weather is still decent and there's no snow on the ground. This thing is a seriously bright flooder! Check these out:

What really impresses me is that this light only pulls 0.92A from an 8.26v battery pack on high. It also runs on a 4.2v pack (same two cells in my 2P1S battery carrier) and pulls 2.02A from that. So that's roughly 0.67A per LED. That calculates out to roughly 5.5+ hours of runtime on high on an 8.4v 5,200mAh 2S2P pack (your average 2600mAh cells)! After reviewing in stock form I'd like to see if I can mod the driver for higher output. I haven't torn it apart to look at heatsinking or the driver yet. Stay tuned for a full review.

-Garry

Update on the KD Duo/Yinding Clone. Good news! KD responded to my inquiry about buying the driver used in sku#S023788 (the "10-step user programmable" one reviewed by ledoman). They will sell the driver for that light for $8. They also reminded me that for my 4.2v version light the LEDs would need rewired to series (I appreciated hearing this from them as it showed me they had some knowledge on what they were telling me about driver replacement). If interested, you would just have to contact them and say you wish to purchase the driver used in bike light sku#S023788.

-Garry

UPDATE: New Flickr Album Here: https://www.flickr.com/photos/garrybunk/albums/72157691589882004

So I received my NiteFighter BT40s yesterday (unexpected fast shipping) and I'm quite impressed. It's definitely a step up from the $15 (+ shipping) Ebay MJ-872 clone. Bought it on sale for $23.XX.

First off, the lights are identical in size; I thought the NiteFighter would be a tad longer, but it's not. They are pretty much identical in construction, except that the silicon button (one "button" works the up & down switches) is much better on the NiteFighter. Like Andy has mentioned previously, the nubs on the inside of the silicone button are in the opposite direction of the switches, but it works fine. The rear of the light unscrews and reveals the driver (which I have not pulled out all the way yet). There is a nice o-ring at these threads. The front bezel unscrews easily (tip - use snap ring pliers to remove these easily), however there is no o-ring here. The bezel does screw tightly up against the optic though which will keep it mostly water resistant.

The optic was a little tough to pry out, but after working around and around I finally got it to pop up. It's not secured by anything other than the pedestals resting in holes, the openings pressed over the LED domes, and the bezel screwed against it. The emitters are mounted on a single pcb which is mounted against a solid backing which is part of the light body!!! Woohoo! (Note - I checked the $15 clone and it is identical too.) There was even a good amount of thermal compound in place evenly spread out! (Note - the $15 clone had way too much of a very runny compound in place.)

My light tested at the following current draws on 2 fully charged unprotected cells in a Pannova case:
Low-0.43A (20%)
Med-0.95A (44%)
High-1.43A (67%)
Turbo-2.14A (100%)
EDIT - I believe the original "2.14A' measurement was a fluke (perhaps higher current draw due to lower input voltage?) Turbo should be 1.75A for 0.875A per emitter before losses.

These currents are higher than I tested on the $15 clone (which reached 1.48A on high. The 2.14A on high roughly translates to 1.07A per LED. An XP-G2 on aluminum star @ 1.0A puts outs roughly 350 lumens (not accounting for losses), so we're talking roughly 1,400 lumens. I do believe this light could be pumped up quite a bit more. (Note: I was experiencing quite a bit of voltage sag even when I tried 4 cells, and I wonder if that is due to the battery pack / lights cabling. It's possible that the light would pull more current with better cabling, high-drain cells, or from a power supply.)

There is driver hum in all modes except high. (The $15 clone also has hum but I could only hear it when held up to my ear.) I don't believe this hum would bother me in actual use. I'm not susceptible to PWM, so I can't comment on noticing any. There is a slight fast fade up and down from one mode to the next.

I compared the tint to a XM-L2 T6 3C and an XM-L T6 4C and I'd say the NiteFighter is between the two, but very close to the 4C, a nice creamy white, definitely not orangish.

My PhotoBucket gallery is here, but here are a few pics posted:

Driver (same Andy?):


Bezel removed revealing star mounted with thermal compound:


Solid backing! Direct thermal transfer to the body!


For comparison, my older $15 clone:


Two side by side (NiteFighter on the right):


Again, NiteFighter on the right;


NiteFighter on the LEFT this time:


Sorry, no beamshots yet. On my basement wall it defintely looks like a wide swath of light as expected (same as my $15 clone, but a bit brighter).

UPDATE 10/2015

new close-up driver pics:

other side:


-Garry

This has been bothering me for a long time.

I have trouble seeing Garry's photos.

Here is an example of what I usually see:


If I quote Garry and maybe refresh my web browser (Google Chrome), I can usually see the images.

I've noticed that deleting "~original" at the end of the "direct link" that Photobucket gives you might solve the problem for posts 13 and later in this thread.

Is anyone else having trouble seeing Garry's photos (in posts 13 and later)?

I used to use Opera (which runs on Chrome engine) and got so fed up with this very problem that I switched to Firefox. I just did a test with my old Opera installation and could see the same problem. It took a while to get all my bookmarks and add-ons running on Firefox, but today I’m a happier man. Good riddance piece of sh*t browser!

This is a great thread Garry. I have to say though that after using CW on bikes so the past 2-3 years and finally getting a NW headlamp, I simply cannot be sold on a CW bike light anymore :slight_smile: NW FTW!

Sorry, I was not aware of this issue. I use FireFox on a PC and Chrome on my Android phone (tried Firefox, but it was too buggy and frustrating).

-Garry

No problem.

I'm sure it's all Photobucket's doing. :)

I had issues with a few other sites. I couldn’t even see my own photos from TinyPic sometimes, very annoying.

Link to the thread with my Convoy M1's intended for helmet mounted use. I compared a build with an XP-G2 vs. a build with an XM-L2.

-Garry

Link to ImA4Wheelr's Custom Mod Thread for my HD-016. My beamshots start at post# 149.

-Garry

Beamshot Comparison between un-modded and modded HD-016. I was in a rush to beat a storm coming, and didn’t notice the first shot was way out of focus (I did almost get some shots with lightning strikes in the background!) Also, the original beamshots had a wrong white balance setting and appear more blue than they should (should appear a crisp cool white). My modded beamshots are with the KD driver programmed at levels 2,7, & 10 (I think low is “2”, but I never confirmed that after my accidental reprogramming). Keep in mind the stock light was only driven at +/- 1.4A per emitter and now it’s modded to 2.4A per emitter. Original beamshots are only with both LEDs on (no sense comparing single throw / single flood shots).

Modded Beamshots, distance to the two trees is approx. 185 feet (I aimed at the tree to the left). Approx. 275 feet to the house off to the left of the tree.:

Low:

Med:

High:

Wooded simulated bike trail

Low:

Med:

High:

Mouseover’s with original Un-modded Light (Mouse Out=Original, Mouse Over=Modded):

Yard:

Wooded:

-Garry

Here’s an interesting comparison, Nitefighter BT40s on Turbo vs. my modded HD-016 on high (note: photos on different dates):

Yard:

BT40s:


HD-016:



Wooded Setting:

BT40s:


HD-016

Mouseovers (BT40s = Mouse Out, HD-016 = Mouse Over):

Yard:

Woods:

Very similar performance, with the BT40’s having some more throw. (Remember the BT40s is 4xXP-G2’s at roughly 1.0A per emitter.)

-Garry

That BT40s sure washes an area with a nice even light. Seems like a nice candidate for the 90+CRI 219C's when they come out.

This is not criticizing as you take much better pictures than me. The perspective is slightly different on the yard shots. Your much more into the beam on the BT40s as the spill hits the bottom of the picture. The HD the bottom of the spill is about 20% up into the picture. That being said, it appears that the beam of the BT is wider. Sure is nice and even, that BT beam.

Yeah, they were taken on different dates so it was hard to match them up (right now I at least have a spot of "dead grass" where I know to position the "table" (not my photo) to do my shots). I'm surprised they matched up that well, especially the wooded shot! I would have taken shots of the BT40s at the same time, but I was in a rush last night. I almost need a "fake bike frontend" to mount the lights to to simulate in the beamshots (that's going a little far I think; my wife would really question my sanity if she saw that in our backyard!). Perhaps I can at least put something semi-permanent on the ground to mark where to place the light and where to set the camera tripod (and keep record of my tripod heights used to keep that consistent).

-Garry