Actual Lumen Readings in 10.5in Sphere.

[FroggyTaco]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

I think there are metallic particles in AS5, that might scratch the UCL coating of you try and wipe it off your lens. I am not sure if its good for the threads either.
:green:

The "particles" are silver hence the name & excellent thermal transfer properties.

 

[MrGman]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

Is some of the "liquid" part of that going to outgas over time and fog over the lens? That is a heck of a lot.

 

[joshconsulting]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

AS5 is slightly capacitive so it could be allowing more current draw as well the improved heat sinking.

Josh..you have to remember the clamping pressure associated with CPU to HSF/WB mounts. These drop-in are not being pressed together any where near what a CPU is.

IIRC, one of the beast thermal greases for large "voids" is one of the ShinEtsu variants.

p.s. Josh isn't it Arctic Cooling MX-3 instead of Arctic Silver? Even since IC7 Diamond has been released I haven't looked back on my PC builds so I haven't watched any recent developments.

Travis

I don't think the clamping pressure is a big issue. Like BigC said, the dropins are being jammed in quite hard. A few months or so ago I did some tests with a Heatkiller v3 waterblock and i7 CPU with a screw-in backplate to see how temperature varied, and it really didn't make much of a difference all the way up to where the cooler wasn't being held down much more then the weight of the block. You are correct, it's Arctic Cooling instead of Arctic Silver (same brand name, different product line), I just got used to the convenience of writing it as AS. I haven't tried the ShinEtsu series as it tends to be more expensive without benefits for CPUs, but if it can bridge larger gaps it would indeed do better. I thought about IC7 Diamond for my main build, but initial reviews didn't really show it beating MX-3.

One way to test if electrical conductivity is a factor would be to try AS Ceramic, it's designed to be non-conductive so you would have to remember to leave a space bare for electrical contact. Also, online you can get 12 or 20 gram tubes for ~$10-12, a lot cheaper then the 3.5g.

 

[bigchelis]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

I just purchased another tube of AS5. I figured the hight current P60 pills benefit great, but what about low current builds.

I have a DeCree XP-G by Electrolumens. The lens is not UCL AR/coated:sick2:, but using the same topped off AW 18650 2600mAh cell I re-tested it with AS5 around the pill section.

The improvements were not as super as the high powered MC-E or SST-50 pills, but it helped.:thumbsup:

Electrolumens DeCree XP-G________XP-G with 5 degree Optic________ 1 AW 2600mAh cell w/ 1.2A____________ 297.7_________________1 sec
Electrolumens DeCree XP-G__________XP-G with 5 degree Optic__________ 1 AW 2600mAh cell w/ 1.2A_____________ 275.4_________________30 sec
Electrolumens DeCree XP-G__________XP-G with 5 degree Optic__________ 1 AW 2600mAh cell w/ 1.2A_____________ 275.4_________________1 min
Electrolumens DeCree XP-G__________XP-G with 5 degree Optic__________ 1 AW 2600mAh cell w/ 1.2A_____________ 271.5_________________2 min
Electrolumens DeCree XP-G__________XP-G with 5 degree Optic__________ 1 AW 2600mAh cell w/ 1.2A_____________ 269.2_________________3 min
Electrolumens DeCree XP-G__________XP-G with 5 degree Optic__________ 1 AW 2600mAh cell w/ 1.2A_____________ 269.2_________________4 min
Electrolumens DeCree XP-G________XP-G with 5 degree Optic____1 AW 2600mAh cell w/ 1.2A___w/ArcticSilver5__302.3________________1 sec
Electrolumens DeCree XP-G__________XP-G with 5 degree Optic_______1 AW 2600mAh cell w/ 1.2A___w/ArcticSilver5__ 284.6_________________30 sec
Electrolumens DeCree XP-G__________XP-G with 5 degree Optic_______1 AW 2600mAh cell w/ 1.2A___w/ArcticSilver5__ 282.3________________ 1 min
Electrolumens DeCree XP-G__________XP-G with 5 degree Optic_______1 AW 2600mAh cell w/ 1.2A___w/ArcticSilver5__ 279.2_________________2 min
Electrolumens DeCree XP-G__________XP-G with 5 degree Optic_______1 AW 2600mAh cell w/ 1.2A___w/ArcticSilver5__ 277.7_________________3 min
Electrolumens DeCree XP-G__________XP-G with 5 degree Optic_______1 AW 2600mAh cell w/ 1.2A___w/ArcticSilver5__ 276.2________________ 4 min

 

[SCEMan]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

I wonder how the CL1Hv4 & MC-E would improve from AS5 treatment since it seems to have better heatsinking than a 6P?

 

[bigchelis]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

I wonder how the CL1Hv4 & MC-E would improve from AS5 treatment since it seems to have better heatsinking than a 6P?

Maybe Craig will let me tamper with his Dereelight hosts for testing.:naughty:

 

[ma_sha1]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

Big C.,

I just made a Triple XPG R5 with LXP optics, direct drive on Mag 1C using DX C-lion. It's crazy bright & I think it might beat out the SSR-50 5A mag. The host is bored to accept 26mm IMR 25500.

I am interested in sending to you for a test. Would you compare Dx C-lion & IMR 26500 the same way you did with the SSR-50 & post the numbers next to the SSR-50 light. I'd like to have two > 1000 lumen OTF lights side by side on your chart 😀

let me know

 

[bigchelis]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

Big C.,

I just made a Triple XPG R5 with LXP optics, direct drive on Mag 1C using DX C-lion. It's crazy bright & I think it might beat out the SSR-50 5A mag. The host is bored to accept 26mm IMR 25500.

I am interested in sending to you for a test. Would you compare Dx C-lion & IMR 26500 the same way you did with the SSR-50 & post the numbers next to the SSR-50 light. I'd like to have two > 1000 lumen OTF lights side by side on your chart 😀

let me know

Send it on over. PM SENT.

The 3 XP-G R5 is suppose to be more efficient than the SST-50 at 5A now we can conclude without a doubt if it is brighter too.

bigC

 

[saabluster]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

The host is bored to accept 26mm IMR 25500.

Man if an IMR 25500 won't get your blood pumping you've got problems.

 

[joshconsulting]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

I'd like to see a 5x or 6x XP-G setup in a 26650 host. The efficiency of the XP-Gs at <1A draw is amazing. More efficiency means more light and less heat. It would cost quite a bit, though. I'm halfway tempted to buy some parts and start trying it myself, but I've gone down that slippery slope before with various other hobbies, and it's a place I don't want to be. I always end up spending three times as much as I should and losing all my free evenings.

 

[Locoboy5150]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

Where did those so-called 630 OTF lumens claimed by the Fenix go?

That 630 lumen rating is a factory rating from Fenix and, as with most factory light output ratings, is generally marketing over-hype. Thus, that's why I basically ignore all factory provided ratings. That goes for my Fenix TK40 too, which was the one that was tested for this thread last week. When I was shopping for it I pretty much ignored the 630 lumen spec.

That brings up the point of whether it *really* matters or not. I've been using my TK40 for about five months now and not once have I ever thought, "gee I wish that this was brighter" during use. So, whether it actually puts out 630 lumens as stated boldly by Fenix or 550 lumens as measured by Bigchelis last week, does it really matter in real world use? To me, not really because the light works very well for my personal use. In an ideal world, the flashlight manufacturers would have an independent certified lab do all their light output measurements. That seems to be an impossibility though.

Thank you Bigchelis for being a warm host last week. Your tests are a *HUGE* eye opener for me in terms of how much these manufacturers over-hype their light specs. They really should get their act in gear.

Also, on the subject of Arctic Silver 5, I bought a tube of it for my 4D cell Maglite with the Terralux TLE-300M-EX that you tested. Without the thermal compound, the body of the light did not get warm at all when using it on high mode. With it, the body does get quite warm after a few minutes of use, so it does work and work well at dissipating heat. The tests that you did last week were with the Arctic Silver 5.

I bought the same size tube that you did but I got mine at Radio Shack. It also was $9.50:

http://www.radioshack.com/product/index.jsp?productId=2216879

 

[jirik_cz]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

Locoboy: just wondering how old is your TK40? Is it from the early batches with K-bin or newer with M-bin?

 

[Justin Case]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

I doubt that Arctic Silver 5 is really necessary. You can save your money and use Arctic Alumina compound or Ceramique.

I tabulated thermal conductivities of the various Arctic XX products here.

Arctic Silver 5, 8.89 W/m-K
Arctic Alumina compound, >4W/m-K
Ceramique, >5.08 W/m-K

Thermal resistance is given by bond line thickness/(thermal conductivity * effective contact area).

I did a back of the envelope calculation for an LED turbo tower in a KT1. The contact area is an annulus (a ring) with OD ~0.8" and ID ~0.6". Thus, the contact area is pi*(0.8^2 - 0.6^2) = 0.88 sq in = 0.00057m^2. If we use Arctic Silver 5 compound, we have a thermal conductivity of 8.89 W/m-K. If we assume a bond line of 5 mil (0.000127 m), we get a thermal resistance of 0.025 C/W. So it looks like the thermal resistance is essentially nil, meaning you should get good heat flow from the tower to the TH and thus good heat sinking by the TH.

The 5 mil bond line assumed above is probably larger than actual since the LED tower is pressed tightly against the under surface of the TH. However, you could have 5 mil slop in the threads of your pill and in the fit between the TH and the flashlight neck. Regardless, the thermal resistance isn't overly sensitive to the bond line thickness (linear dependence). We could double the bond line (and 10 mils is a big gap) and the thermal resistance will become 0.05 C/W for the above turbo tower calculation. Still negligible.

The thermal resistances for the other compounds scales with their thermal conductivity. Thus, for AA compound, the thermal resistance is <0.056 C/W. For Ceramique, it is <0.044 C/W.

BigC, your setups most definitely will have far more contact area than what I've used here for an LED turbo tower in a KT1. Probably orders of magnitude more. Any minor differences in thermal conductivity for Arctic Silver 5 vs some other magic goop is completely swamped by the huge contact area.

Keep the bond line thin and the contact area large and you probably can use toothpaste.

 

[MrGman]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

Is the Alumina and Ceramique thermal compounds electrically conductive? Aren't these both insulators? Putting them on the threads of the pill might make for a very resistive path. In other places it may not matter but the way big C slathers it on he may find he has an open or highly resistive circuit and a very dim light using these, yes?

 

[Justin Case]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

The pill's electrical contact to the TH reflector is irrelevant. That's not part of the flashlight circuit path. The ground path is pill shell to flashlight tube neck. You are right regarding avoidance of using electrically insulating goops elsewhere, such as on the outside of the TH and pill where they may contact the flashlight neck.

Still, IMO it is overkill and unnecessary to slather the head all over with thermal compound. Even for what I would guess to be a large bond line, you need only a square inch or so of contact area to reduce thermal resistance to a negligible level. Thus, I'd put thermal compound only on the TH threads that connect to the rest of the flashlight. I'd leave the remainder of the TH clean. All of that excessive compound is just a mess that is going to cause other problems.

It would be interesting to test for any lumens output difference when using Arctic Silver 5 vs AA compound.

 

[recDNA]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

So the Mini on High with an IMR 16340 would run about 15 min total? (I know not to run it for 15 min straight. I mean the total of let's say 5 three minute runs)?

Is there any clue when the IMR is getting too low? I noticed mine is at 3.7 v yet high is still maximum brightness.

I assume on a conventional RCR123 it wouldn't be discernably brighter than with a primary?

 

[psychbeat]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

no its waaay brighter on high with a regular rcr than with a primary.
15min or so sounds about right runtime wise. Ive had mine on for about that
long using on my bike but was mooooving and I sprayed water on it 😉

my NB sst50 is showing up today I think so Im gonna stop by rat shack
n get sum AS5. Ill try not to slather it toooooo much but I know it will
be tempting!!

 

[ky70]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

Am I reading this right? On a cr123, the Romisen RC-N3 pushes out more lume than a qmini 123 (150 v. 142)? I wonder what the Romisen would do on a rcr123?

 

[psychbeat]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

silly ? but do u guys think AS 5 would do anything for a qmini123?
I use mine with 16340 pretty much exclusively:devil:
a little on the threads might help transfer heat to the body?
😱 or maybe just make a mess....

 

[Locoboy5150]

Re: Actual Lumen Readings by bigchelis in 10.5in Sphere.

Locoboy: just wondering how old is your TK40? Is it from the early batches with K-bin or newer with M-bin?

Gee, I don't know. I bought it in October 2009. It is serial number 1N7J5A00407 if that helps.