Review: Kaidomain KD C12 XPL HI

bibihang wrote:

What? Sorry I’m really confused…

EDIT: I also believe that the copper base is not anything like DTP, so the heat dissipation could be still quite poor regardless if the substrate is made of copper or aluminium.

Give the following djozz thread a read (link in picture). The copper base with dialetic layer does pretty darn good in that testing.

With that being said, I could be completely wrong about it being a possibility.

Both torches had a 3A driver althouh not the same one. I did test out different reflectors to assess the impact it had on the LUX readings. As expected a SMO reflector leads to a higher LUX figure. The SMO cleans up the beam which removes rings in the C8. I wasn’t able to work out why the overall brightness was lower on the C12. I tried multiple batteries too. The sample C12 model I had doesn’t produce the brightness levels I’d expect of XPL-HI. I also found in my results that the C12 light output over time wasn’t quite as I expected. I saw results where as time passed the LUX readings increased. Of the two drivers I found that the KD V2 was the better driver.

Rolz,

First off, Great review!

I have one of the older C12’s with an XPL from Aliexpress when they first came out and it is a Smooth reflector, so when this XPL-HI version came out, I jumped on it and figured why not get the OP reflector and then could swap them back and forth. Annnnd, NO, will not work properly, as they are just that little bit different, same with the pill threading.

Anyway, the pill threading should be the same on your 2 lights from KD. So could you swap the pills on the C8 and C12 and see what results you get then?
Not full blown graphs, just Lux would be interesting.
I am confused how the output can go down and then come back up on the C12, never saw that happen before but there is a first time for everything.
Maybe the thermal paste is changing properties based on heat?
My original C12 actually will put out 3.58 amps on an LG HE2 so it is brighter for sure on white walls then the new XPL HI and the throw is right there with it but with more spill and bigger hotspot (as expected).
Tint is better as well on orig XPL.

Overall the XPL Hi is a disappointment to me. Was expecting way better throw. YMMV.
It may just need a re designed reflector to show off it’s stuff.
Thanks!
Keith

Keith,

Thanks for your comments.

1. could you swap the pills on the C8 and C12 and see what results you get then? -> Will aim to do so tonight and post my thoughts/observations
2. I am confused how the output can go down and then come back up on the C12 -> I was surprised by this result. I ran multiple tests using different batteries and found similar results. My immediate thought is that it’s driver related. The C8 and C12 have the same LED, assume similar wire size to LED, and assume similar switch resistance. I expected that the output initially would be similar and that the better cooling efficiency of the C12 would come into play in a 10 minute test. The C8 was hot after 10 minutes, whereas the C12 wasn’t too hot to hold, it was very warm.
3. My original C12 actually will put out 3.58 amps on an LG HE2 so it is brighter for sure on white walls then the new XPL HI and the throw is right there with it but with more spill and bigger hotspot (as expected). I have a C12 XML L2 which performs well. I’ve not checked the tailcap current, but may do so tonight and include some numbers as a comparison.
4. Overall the XPL Hi is a disappointment to me. -> The C8 with a XPL - HI performed very well. The KD V2 driver provided very good results. I’ve a C8 (XML U2) and a C12 (C12 XML L2) which I can compare against. Again I’ll do some more general tests as a comparison. The hot spot on the XPL-HI is clearly smaller in size.

I believe a XPL-HI in a C12 host with a more effective driver would be an excellent throwing torch for it’s size. If I had to re-design the KD C12 purely based on my findings I would swap the driver. The inability to take non protected cells without a spacer is bit annoying. I may find that if I can manage to swap over the pills as per the dot point 1 above I’ll have an excellent torch.

All thanks for the followup discussion. The response to the review has been very pleasing for me. I’m new to producing and publishing reviews on the web and have found it to be an interesting experience. It does take much time, more than I expected. I try and improve the content and presentation each time to make it more interesting to all. I’m thinking of ideas on ways of assessing the total light output. I don’t have a sphere but I’ll work something out.

I will produce more in future, with hopefully new models. If you know of areas that I should improve please PM me, I always appreciate all feedback, good or bad. Thanks Again :slight_smile:

“I’m thinking of ideas on ways of assessing the total light output. ”

.

A ceiling bounce and a light meter will do very well here.

Thanks,
-Chuck

Chuck thanks it helps me confirm my thoughts. I have a room where I can do so and take a baseline for a known torch. I was considering stating “light units” as the measure for comparison.

Hi Rolz,

You just need to be sure that everything is in the same place each time, and then it doesn’t really matter what you call it, because it’s all relative.

Just go down to the middle of the page here and see how Selfbuilt does his:

http://www.candlepowerforums.com/vb/showthread.php?381868 “Ceiling Bounce Max Output” or whatever…

Hey! - I really am enjoying your reviews! - Keep up the good work!

Thanks!
-Chuck

Thanks again for your help on how best to proceed and your feedback :slight_smile:

your second review?Superb job !I am impressed with the entire revirw,format side by side comparisons ,the graphs and I am suprise by the out put and throw I mean same reflectors driver driven the se except one bieng a Nanjing 105 the other a KD V2 ,I tough it was a great side by side !

Thanks, I was considering a mouse over for the outdoor shots but I didn’t want to over complicate the task being only 2nd time around. I wasn’t 100% happy with the final layout of photos, but next time I’ll fine tune it a little to improve the presentation. I’m still developing the testing/measurement methods. All aimed at making for a better review and hopefully simplifying the task.

Followup to the questions and suggestions by Muto

1. could you swap the pills on the C8 and C12 and see what results you get then?

I was able to swap the pills around. The C8 pill gitted perfectly into the C12 host. No issues
The LUX readings over a 10 minute period were similar to those from the C8.
C12 SMO reflector with LG HE4
starts off at around 55k LUX , peaks at 56k then over 10 minutes gradually drops down to 51k.
The temperature readings were better than the C8 as the peak temp after 10 minutes was only 37C

2. I am confused how the output can go down and then come back up on the C12 ->

3. My original C12 actually will put out 3.58 amps on an LG HE2 so it is brighter for sure on white walls then the new XPL HI and the throw is right there with it but with more spill and bigger hotspot (as expected). I have a C12 XML L2 which performs well. I’ve not checked the tailcap current, but may do so tonight and include some numbers as a comparison.

I tried my c12 XML2 SMO @ 3.8A tailcap using a Samsung 30Q starts of at around 36k LUX. Hot spot is larger.

4. Overall the XPL Hi is a disappointment to me. -> The C8 with a XPL - HI performed very well. The KD V2 driver provided very good results. I’ve a C8 (XML U2) and a C12 (C12 XML L2) which I can compare against. Again I’ll do some more general tests as a comparison. The hot spot on the XPL-HI is clearly smaller in size.

I ran LUX test on the C8 fitted with the NANJG 105C driver and a LG HE4 battery. The results were

start 44k Lux @ 22.2C
end 39k Lux @ 44.4C
The drop off was linear

Overall assessment here is that the NANJG 105C dropin isn’t performing as well as the KD V2 driver dropin. The results on the NANJG were still good, but were a fair margin off the KD V2 driver. The C12 host clearly dissipates heat more effectively. The best pairing is the C12 with a KD V2 XPL Hi dropin. It also overcam the issue with the shorter batteries needing a spacer. The KD V2 pill is slightly deeper, enough to overcome the issue.

Hi Rolz, perhaps useful for future reviews, for a build instruction of a simple integrating sphere that will be more accurate than the ceiling bounce method, click on 'integrating sphere 4' in my sigline.

Djozz, thanks for the tip. I’ll have a look around at craft shops for a sphere, if I can find one I’ll have a go at making one up. I’ll also experiment with the ceiling bounce as another method of measurement. I’ll try various options and determine what’s best.

The KD site has now discounted the C12 torches to $19.49 see first post above Review: Kaidomain KD C12 XPL HI - 18650 Flashlights - BudgetLightForum.com for link details to individual models, otherwise visit

http://www.kaidomain.com/Search/SearchResult.Cree%20XP-L%20HI/-1 to see all discounted XP-L Hi products.
.

Yes I have read djozz test multiple times, and thanks for that. Maybe copper base with dielectric layer is okay for running at 3.5A, and not so okay for anything above 4.5A, not sure, no data to back this up yet so it is just my feeling.

Anyway I still don’t agree with the part that LED need to reach a certain degree of temperature to produce maximum brightness… at least from what I understand LEDs love to run cool, and HID however loves to run hot.

^ Just threw it out as a possibility because heat sinking is one of the differences in the lights being compared.. I don't know how the XP-L Hi behaves at different temps. I agree that it probably is not the cause of the lower output, but I don't think we know enough yet to rule it out as a possibility.

The results I present where the maximum brightness occurs after the temp increases only occurred with the NANJG 105C driver. The results obtained for the KD V2 3.04A driver were consistently, start high and then a gradual decline as the volts drop and the temp increases. The 105C driver didn’t work as well, resulting in slightly reduced brightness and the dip and then rise in brightness. Of the two drivers the KD V2 wins out. The LEDs were the same in both dropins.

Nice review Rolz! Liked the charts and graphs.

I've seen these variations in output before, with the ups/dns, mostly in stock budget lights. The output flakiness always clears up after fully modding the light. On the C12, the light is generally under-performing across the board, so, there is something wrong. The drivers should basically be the same output at 3A, but the output is significantly different. To me it seems like there is some intermittent or poor connection in the electrical path, probably ground. Could be the use of the magnets with the under-sized springs, not sure. Springs could be heating up, effecting the output (I think conducts better w/heat), etc. Or there's a ground problem with the driver mounted in the pill, but it's most likely something in the ground path. Generally a poor connection will start out low, then increase over time - this indicates from heat, conduction is improving. Could be a poor spring, or poor spring soldering job, poor driver ground, etc.

A tailcap measurement typically doesn't reveal this problem - only when the light is assembled and output is measured. With a tailcap measurement, I would typically press down the cell pretty good, so it's different when fully closed up.

I'm thinking I might have fixed issues like this just by doing the spring bypass's, not sure. Many times I'll replace suspect springs, and if I don't like the solder job of the springs, beef that up too while I'm doing the bypass's anyway. The bypass's solve a lot of problems, or potential problems: stiffens up the springs, better conduction for loose fitting batteries, eliminates poor conduction of some budget springs, improves on the soldering of the springs to the boards, etc.

Tom thanks for passing on those ideas. Soon I’ll have tidied up my workbench area in my garage which will enable me to get on with some practice with Mods. I can alsoremove the driver and look for any issue and experiment with the torch to see what may be the cause of the inconsistent results. Clearly something isn’t quite right as I would have expected better performance. Thanks again