Sold/Expired Custom built triple CREE XP-E <> P60 drop-ins - SALES CLOSED FOR NOW

[TechnoBill]

Conductive is just faster/easier.

To quote Dogbert, "Simple!" I just have my weak hand replaced with a RoboCop or Borg hand, run nitrogen through the capillaries, and I'm all set :laughing:.

 

[Sgt. LED]

Rip the left sleeves off of your shirts and stick it in your armpit. :twothumbs

Oh wait, nah maybe not.

 

[manoloco]

Moddoo,

At the moment i have never spent so much on a drop in or part, heck only twice have i spent a similar amount on a whole flashlight, but i have already started to save for one if you decide to make them again, i cant believe i missed this.

My plan is to put it on a 6P head and an E2C adapter so i can use it 3p style on an Aleph 1x123 or 6p style on an Aleph 2x123, on the tail end: Aleph trim tailcap with a McE2S and 22ohm resistor or maybe 60, still on the fence on the value, planning to run it mostly on rechargables (imr16430 or maybe will buy a host for more power in the future so i will probably have to save more for the 3v-14v version, ouch). All black host

In other words, i really hope you find time to make these again!

 

[Moddoo]

I might do a few of these again sometime...
:thumbsup:

Moddoo,

At the moment i have never spent so much on a drop in or part, heck only twice have i spent a similar amount on a whole flashlight, but i have already started to save for one if you decide to make them again, i cant believe i missed this.

My plan is to put it on a 6P head and an E2C adapter so i can use it 3p style on an Aleph 1x123 or 6p style on an Aleph 2x123, on the tail end: Aleph trim tailcap with a McE2S and 22ohm resistor or maybe 60, still on the fence on the value, planning to run it mostly on rechargables (imr16430 or maybe will buy a host for more power in the future so i will probably have to save more for the 3v-14v version, ouch). All black host

In other words, i really hope you find time to make these again!

 

[Justin Case]

So the finned head isn't quite as heavy as the M2 head... I wonder if the possible poorer thermal conductivity(?) resulting from the shock isolation in the M2 head will outweigh the added mass in the end...

Cheers,

Chris

Unfortunately, I have only a current version Z44 bezel and an original version Z32 bezel. I weighed each on a digital postal scale. The Z44 weighed in at 0.8 oz (22.7g) and the Z32 weighed 0.9 oz (25.5g). The M2 bezel is reported to be about 39g, or about 1/2 oz. heavier than the Z32.

I installed a Triple XP-E R2 WG drop-in in an FM 1xIMR26500 host. The drop-in uses Moddoo's heat sink and I obtained the triple XP-E star and optic from the same source. However, I built a single-mode, 2-board sandwich consisting of 6xAMC7135, vs the single-mode 8xAMC7135 driver used by Moddoo. My driver measured 2140mA drive current (~700mA per XP-E).

I removed the tailcap and the IMR26500 from the FM host to gain access to the drop-in. I threaded the power leads from a bench power supply to the drop-in. I also inserted a thermocouple down the tube and pinned the thermocouple in-between the bottom face of the drop-in and the inside of the FM host's "neck". I probably would have preferred measuring temp at the aluminum drop-in's heat sink itself and as close as possible to the XP-E star. But there is no access to do that if I want to install the drop-in with a bezel and a host body.

Basically, the test concept is similar to what I did here.

I ran a temp vs time test, first using the Z44 bezel. Then I let the works cool down and re-ran the test with the Z32. For each run, the light was bezel down on the bench top. Perhaps in theory, this might help heat escape out of the flashlight tube like a chimney. But maybe this also compensates for the loss in mass (and heat sinking contribution) of the IMR and the tailcap. Then again, with the flashlight bezel down, some of the heat from the LED is trapped since the bezel window is blocked (or mostly blocked -- the bezel was partially cleared so that I could visually verify that light was still coming out of the drop-in).

In any case, after about 40 min, each run seemed to settle in to a final equilibrium temp. For the Z44, the final temp was 154F (67.8C). For the Z32, the final temp was 168F (75.6C).

If we assume that these temps represent the solder point temp (and I actually suspect that they may be lower than the solder point temp), then the junction temp is estimated at

Tj = Tsp + P*(9C/W)

where Tsp is either 67.8C or 75.6C and P=3.2V*2140mA ~6.85W. Thus, Tj(Z44)~68C + 62C = 130C and Tj(Z32)~76C + 62C = 138C. The max junction temp spec is 150C. For my triple XP-E drop-in, the data suggests that I should run the light for a max of about 3-4 min if I want to keep the junction temp below 100C (an arbitrary limit, but easy to remember). Of course, there is some uncertainty in that estimate since this test isn't an exact mimic of real-world flashlight usage. The light was stationary, the tailcap and battery were removed, and there was no hand holding on the light. For the Cree XR-E, the 50,000 hr life was based on 70% lumens maintenance with Tj <= 80C. I haven't seen any comparable data for the XP-E, but maybe we can assume the same. I've seen a graph for the Seoul P4 driven at 700mA that gives Tj~77C for 50,000hr/70% lumens maintenance. 25,000hr occurs at Tj~95C. 12,500hr looks to occur at about 115C. Unfortunately, I haven't seen a similar graph for the XR-E or XP-E. Just the one 70%/80C/50000hr figure. But if the XR-E and XP-E behave like the Seoul P4, then it looks like running the triple XP-E at 100C junction temp max should still give thousands of hours of 70% lumens maintenance life. I also note that the AMC7135 chip has a thermal protection circuit that cuts off the LDO regulator when Tj reaches 150C. My triple drop-in never cut out, so I assume I never approached 150C for the driver.

I would assume that the Moddoo drop-ins with the higher-powered drivers will run even hotter than my drop-in.

I tried to do a third run with no bezel at all, but just couldn't get all of the jumble of wires to sit still. I assume that without the pressure of a bezel to pin the thermocouple in place, I was getting anomalous readings and so I gave up for now.

I'll post a graph of the temp vs time results later.

 

[ElectronGuru]

Unfortunately, I have only a current version Z44 bezel and an original version Z32 bezel. I weighed each on a digital postal scale. The Z44 weighed in at 0.8 oz (22.7g) and the Z32 weighed 0.9 oz (25.5g). The M2 bezel is reported to be about 39g, or about 1/2 oz. heavier than the Z32.

A single graph showing both lines would be outstanding.

Let me know if you want to try out my M2 bezel. It measures in at a full 39g.

 

[TechnoBill]

If we assume that these temps represent the solder point temp (and I actually suspect that they may be lower than the solder point temp), then the junction temp is estimated at

Tj = Tsp + P*(9C/W)

where Tsp is either 67.8C or 75.6C and P=3.2V*2140mA ~6.85W. Thus, Tj(Z44)~68C + 62C = 130C and Tj(Z32)~76C + 62C = 138C. The max junction temp spec is 150C.

Sir Case,
Is it possible to define to what "junction temp" refers in layman's terms, or would the simplification enervate the material of the definition? My goal is to delineate between whether it is a physical place or a functional construct of electrical engineering.

Thanks,
TechnoBill The {in this case exceedingly} Obtuse

 

[Justin Case]

Here are the T vs t curves for the Z44 and Z32 on a single set of axes.

 

[Justin Case]

Sir Case,
Is it possible to define to what "junction temp" refers in layman's terms, or would the simplification enervate the material of the definition? My goal is to delineate between whether it is a physical place or a functional construct of electrical engineering.

Thanks,
TechnoBill The {in this case exceedingly} Obtuse

This pair of graphics might help:

Basically, Tj is the internal temperature of the LED die (specifically right at the so-called "pn junction").

Ts is the slug temp. Tb is the metal PCB temp. Ta is the ambient temp. In-between, you have Rjs, Rsb, Rba, and overall Rja. The assumption is series thermal resistance, just like series electrical resistance, so that Rja = Rjs + Rsb + Rba.

Also Tj = P * Rjs + Ts (here, the slug temp Ts is the same as the solder point temp, or sometimes called the case temp; P is the power dissipated by the LED). That's how you can estimate Tj, even though you can't get direct probes inside of the LED itself.

From junction to ambient, the temp steadily decreases because of the thermal resistances involved. That's why external temp could be misleading wrt Tj. Also, you can see my thermocouple measurement was not truly the solder point temp. It was closer to Tb. So that's why I stated that potentially Tj is higher than calculated. But then there are the compensating factors arising from real-world usage -- moving air, more mass from batteries and tailcap, hand contact that improves (decreases) thermal resistance to the ambient. But since I don't know how those real world factors affect the data I measured, I went with the data I had, not the data that I wish I had.

 

[Moddoo]

nice to see some speculation, measurement, experimentation, teaching, learning, etc. going on here.

Please note, these are not table lamps.

As I stated earlier, walking with this light in your hand will result in a hot, but not uncomfortable temperature.

 

[Neo Mara]

Sounds like these lights excell with short bright bursts which is perfect for a weapon light 😀

Someone mentioned using a thermal compound earlier in the thread to help heat dissipation. I wasnt going to bother but after seeing how much these guys heat up I think I might go for it. I have some Artic Silver 5 around here somewhere from CPU installs so I think I might as well give it a try. Can't wait to get mine!

 

[TechnoBill]

This pair of graphics might help:

So the junction is A)The ambient atmosphere above the die B)The top surface of the die C)The "die-attach" also described as the 'interface' between the die and the heat sink-slug D)Cosmic debris ~ quarkoid dust bunnies proximal to the die but in a parallel Universe?

 

[manoloco]

Theres some stickyness to the AS5 that makes it hard to clean (that stickyness is probably the same reason it works so well on processors)

but for something you have to remove once in a while to make maintenance i would choose AS ceramique, imo.

theres is also the aluminum foil wrap which has been documented on another thread.

 

[supasizefries]

I just finished another hand held (ie. highly subjective:naughty🙂 full power test, except this time with the M2 bezel I got from E Guru. Remember, this is based on feel so I would call these observations and not "results". 😉 The M2 bezel seemed to take about 4-5 mins longer to reach the same level of heat as did the finned head. But here's the kicker, the M2 bezel seemed to keep the rest of the light cooler than the finned head did. I stopped my first test because the body was approaching the heat levels of the finned head, whereas this time around, the body stayed cooler. I think I'm gonna stick with the M2 bezel. I actually like the way the light looks better than with the finned bezel.

 

[Justin Case]

I just finished another hand held (ie. highly subjective:naughty🙂 full power test, except this time with the M2 bezel I got from E Guru. Remember, this is based on feel so I would call these observations and not "results". 😉 The M2 bezel seemed to take about 4-5 mins longer to reach the same level of heat as did the finned head. But here's the kicker, the M2 bezel seemed to keep the rest of the light cooler than the finned head did. I stopped my first test because the body was approaching the heat levels of the finned head, whereas this time around, the body stayed cooler. I think I'm gonna stick with the M2 bezel. I actually like the way the light looks better than with the finned bezel.

Cooler is bad. The M2 bezel is acting as a better insulator than the finned bezel.

 

[russthetoolman]

The answer provided by the multiple choice would be "C". The backside of the LED.
Russ

So the junction is A)The ambient atmosphere above the die B)The top surface of the die C)The "die-attach" also described as the 'interface' between the die and the heat sink-slug D)Cosmic debris ~ quarkoid dust bunnies proximal to the die but in a parallel Universe?

 

[Linger]

Now introducing:
:candle:The Official priority list:candle:

Thank-you for your participation, and you're pacience. I'm please to announce that our arrangement has been legitimated, the list is approved.

A second run is coming.

...parts are on order....I have made several improvements in design and manufacturability.
parts on order for run two, including some upgrades
:thumbsup:

ElectronGuru's guidelines for the Official priority list

There will be many fewer configuration options this time around

There may not be enough units to satisfy all reservations

Available units will be provided based on slot priority

One slot per person

One drop in per slot

More details soon!

Thanks,


(At this time the number of your slot is all that matters. I've continued to post your requests beside each name. Just remember these are only your requests and do not indicate what options will available to your position)

1. russthetoolman
2. Paa100
3. ¿😵?
4. Kid9P
5. manoloco
6. Linger (1 + heatsink)
7. bshanahan14rulz (drop-in or kit if available)
8. JustinCase (one or two more heatsinks)
9. AW (warm tint)
10. Stillphoto (single mode or low-med-high. For running on a single cell)
11. smopoim86 (warm tint)
12. Xfinity
13. mfrey
14. QtrHorse (triple 3 mode drop-in)
15. Jimed315
16. jefft
17. christian gpr
18. Litefan
19. hornady708
20. thegeek (heatsink for single emiter)
21. DimeRazorback (warm-tint narrow)
22. RocketTomato
23. nfetterly
24. openbolt1
25. EvilPaul2112 (narrow, cool tint. single mode)
26. LightJunk (Cool,narrow. single 100% mode)
27. Chodes
28. supasizefries (triple XP-G, narrow optic, unfrosted)
29. thegeek (hole off to the side)
30. River Runner (XP-G, narrow optic, unfrosted)
31. mcmc

 

[Justin Case]

This is where you'd like to measure the solder point temp for the Cree XP-E:

Hard to do this with the optic installed and the drop-in inserted into a bezel.

 

[bshanahan14rulz]

Unofficial waitlist:
russthetoolman
Paa100
¿😵?
Kid9P
manoloco
Linger

and BShan (dropin, or kit if offered)

 

[Justin Case]

I would be interested in one or two more of the heat sinks.