XC-997B review

I needed to order some other stuff so I thought I'd waste a lobster on one of these just to see what it was like. I've got an interest in finding a good light for commuting, mostly on road, some on an unlit bike path, that is good enough to be well seen and bright enough to light my way.

Source: http://www.dealextreme.com/p/xc-997b-cree-3w-200lm-3-mode-white-led-bicycle-bike-light-4-x-aa-108640
Cost: US$19.95
Delivery (to Australia): 2-3 weeks


Various teardown images: (equivalent xc-997a, same I believe except for the batteries/charger) http://www.dealextreme.com/feedbacks/browseCustomerPhotos.dx/sku.82138




Rated output:
3 Watt LED, but power draw is apparently measured at 1W
20 lux (according to instruction sheet)


Videos:
Construction: http://www.youtube.com/watch?feature=player_embedded&v=gaPMxyoktiw#!
Street: http://www.youtube.com/watch?v=rmAL784uxVI
http://www.youtube.com/watch?v=VTaWAgXomZg


Random discussions:
http://www.candlepowerforums.com/vb/showthread.php?324828-Build-thread-double-XM-L-road-lights
http://www.candlepowerforums.com/vb/showthread.php?323176-Fork-mounted-lights
http://forums.mtbr.com/lights-night-riding/commuter-lights-throw-vs-area-784993.html




My reference points:
Previous: Niteflux Commuter4 (4W, symmetrical beam, terrible artefacts in the beam shape)
Current: Magicshine MJ-808 (10W SSC-P7) - I run two of them. Because I can.
Also a BMW E46 330i with HID projectors - my benchmark


Out of the box:


Poor instructions (model variant and mounting leaflet only)
Batteries leaking (went straight in the bin)

Construction appears OK but nothing special.
Plug is the same size as Magicshine, but reversed orientation.
Have tried a Magicshine extension cable but oddly it doesn't work, though it plugs in.
waterproofing looks adequate, not yet tested. The plug grommet pulled out with the plug cap at first - my mistake, I couldn't figure which part to pull out.
Battery pack is a 4S design, sealing looks marginal (gasket doesn't compress much).
Internal boss/screw inside battery pack lid broke off, but doesn't seem to affect operation.
Image before it broke:



Beam:
Pros:
Good sharp cutoff at the top.
Brighter at the top than the bottom
Wider at the bottom than the top projected on a wall
This translates into a fairly evenly illuminated rectangle on the ground. Enough to light up an area ahead of you on a bike track.


Cons:
Very narrow (maybe 2m wide at 6m distance) - not good for seeing anything except what is right in front of you.
Very very shallow - not much light spilled close in front of you, and above the cutoff you can't see anything.
Almost total lack of spill. OK for a bike path but you will not see any overhanging branches
Weak - not up to the requirements of a wet road.
Artefacts: In beam, a bit patchy left to right
Out of beam: some random artefacts but quite minimal. Probably useful for side/angle visibility


Brightness:
at centre, comparable to the 4W Niteflux. But the hot spot is smaller.
Visible alongside a Magicshine, but weaker in the centre and drowned by the much broader beam.




Oncoming view:
From inside the beam, just as dazzling as a Magicshine. Luckily this will be aimed lower than most oncoming observers.
From outside the beam, it is much weaker, much less dazzling. This would be ideal on bike paths and in urban traffic. However it is possibly to the point of not enough spill to warm oncoming traffic/people pulling out from side streets. In Australia, drivers aren't expecting to see bikes; you pretty much have to scare them into submission.


Colour:
Much bluer than either Magicshine or Niteflux. Poor.


Conclusion:
What do you expect for under $20? In that criteria, it's probably OK, though so far I have no reliability info to report. For seeing things on unlit paths, it needs improvement. Maybe mounting 4 of them would do the trick, but by that point you might as well buy something decent.




Possible modification:
Make the beam brighter.
LED upgrade to XP-G (possible drop in) or larger?
Replace driver to suit.
XM-L? Might spread the beam pattern, this may be no bad thing except for the upper cutoff.


Make the beam wider and deeper.
Moving LED rearward should do this, or the use of a larger-surface emitter (XM-L anyone?).
Could two LEDs be fitted side by side in the housing? Probably cooling issues...
Use two lamps side by side?


New housing would be required to provide adequate cooling. Billet machined? $$$. The shape of the reflector (more specifically the position of the emitter) doesn't easily lend itself to using the reflector dropped into a different housing easily.

Thanks Matt, I have been interested in how this light performs. Pity that it doesn't live up to its promise

Cheers

Bruce

find_bruce said:

Thanks Matt, I have been interested in how this light performs. Pity that it doesn't live up to its promise

Cheers

Bruce

Click to expand...

Oh I think it promised very little so no loss there. But as a basis for modding I think there is some potential...

I've never used a shaped-beam light before, I suspect many of them (having seen beam shots on line) have the same issue of a narrow and shallow beam, to make the most of limited power input (there are dynamo versions of this same light)

Some beam shots in the back yard. ISO 100, 6 sec @ f/4.0
(Forgot to get head on shots...)
Cold night, there is a fair bit of dew on the ground.

Reference (no lights)


XC-997B - on high


Niteflux Commuter 4 (rated at 300 lumen by manufacturer)


Magicshine MJ-808 on high (typically described as ~550 lumen, but it poops on the one above from a great height)


Two Magicshines


So the basic conclusion is that it needs a lot more width and more foreground illumination.

I got the 997B when it was on sale for about USD18. I rode it maybe two nights, and my conclusion is that there's more artifacts than there is beam. The center beam is O.K. at best, but the overall output looks to be only in the double-digits lumen-wise. Its about an order of magnitude too small where directing the beam with an asymmetric reflector would add any major benefit. Reflector losses are the more dominant factor here. I think MK's photos make it look brighter than it is.

I don't think the reflector losses are that bad but you'd need some teardown and an integrating sphere to know. Some thought has gone into the reflector, it wouldn't surprise me if it's a common part off something more expensive.

Dropping in a much more powerful emitter would be really interesting, though only useful if the side and downward spill can be improved. Though as I said earlier, moving the emitter backwards should accomplish both of these

If I get a chance I'll get some artefact shots. The dark patch in the beam shot above is due to the different surface - less dew under the lemon tree (the darker patch is visible in all the images).

Did some teardown last night. The reflector (as seen in one of the links above) has some posts on top, and where the LED PCB sits is slightly recessed.
The hole into the reflector is 5mm diameter. It is about 3mm from the rear of the reflector. So some cutting would be required to fit a larger LED in. eg the P7 from my Magicshine that I also pulled down is 9mm diameter - this means it really couldn't be moved any further rear, however the surface of the emitter would partially be rearwards.

I tried holding the P7 up to the hole but it was impossible to get close enough to properly see what's going on. This is due to the posts on the reflector, and the body the P7 was sitting in.

Enlarging the hole would be tricky without damaging the internal surface of the reflector immediately below it. Might be possible with lots of care, but cutting plastics is rarely a simple task.

The light has been sitting in the cupboard for a while. I decided to take it out last night to play again.

This immediately resulted:




The battery case and lid parted ways.

You can see here that the retainer clip for the battery terminals that broke off earlier has caused the terminal to flop around. I think this was the underlying issue - I didn't notice that this part had got in the way of the lid closing, and this consequently over-stressed the hinge.

So, if you were think about one of these, get the 997A version with the recharger so you won't have to open the lid.

Reviewing the beam and reflector more, there are 6 "segments" to the reflector - one square immediately under the LED, and 5 narrow rectangles beside and below this. Partially blocking the beam indicates that:
a) the square under the LED fills the whole of the beam shape, including foreground.
b) the other 5 segments do the "hotspot" at the top of the beam

The reflector segments appear to be essentially toroidal - ie, different radius of curvature in the vertical and horizontal planes. Though possibly more like parabolic in the vertical axis...

Gee Matt, a cheap chinese light that is not entirely robust. Colour me surprised

I never expected it to be high quality. I'm more interested in what can be done with this unit as a donor of its shaped-beam reflector. The results will be interesting even if not effective - might end up being a bit narrow. And if the beam is OK, then the next issue is figuring out how to mount it into a well-cooled housing.

But the beam with about 10W should have quite a few lux...

I have posted it already elsewhere, but for the reference I think it would be useful also here (mods: feel free to remove if you consider it redundant)

Here are beamshots from XC-997 (bottom) and from XC-997 with the emitter replaced with an XM-L (top):


And 2 real life shots: (exposure ISO800 0.75s f2.8 which is roughly equivalent to ISO100 6s f2.8)
The original:



and with XM-L emitter: