Hi Tom, Have you cleaned up the heatsink plate or have they used Kapton tape 'Instead of' thermal compound ?
Edit - oh, realized I didn't answer your question directly - yes, the pic is after cleaning up the thermal grease. They used white colored grease, not gray, so it couldn't be AS5, could be Fujik, or could be another Arctice grease though, couldn't tell.
I sanded the heat sink (alum pill top) and star at 8 levels, up to 2500 grit, that's all, then I used AS5.
Oh! The gas de-doming only took like 2 hours, finished and tested. No battery carrier mods at all (yet), but using Pana PD's (not even fresh):
lumens: 1,332 at start, 1,275 at 30 secs
throw at best, measured at 4.31 meters: 347 kcd, which is 1,178 meters at 1/4 lux (moonlight)
So, got the results I was hoping for. Actually, the throw should be higher, and I'm thinking will be if measured at a longer distance like 15 meters, but not sure. Wow - it's really an intense beam.
Thanks very much indeed Tom, and congratulations on the fantastic results. I'm looking forward to your beamshots if you have any :)
Wow, that is some thermal insulation.
I have this terrible feeling that most of us are going to find this tape on the back our copper MCPCB's
Thrunite owner is aware but his response wasn't all too comforting. The language barrier is difficult and I'm not entirely sure he knew exactly what I was talking about first time I mentioned it to him.
After showing him this thread and the actual picture he replied with "I see the tape it is for preventing the anodized on the copper"
It should have been removed prior to assembly and I've highlighted the heat transfer issue but I'm yet to receive a further response.
MCPCBs are not being anodized...
Reading between the lines he likely meant to say Oxidized instead of Anodized. But nevetheless, it should have been removed prior to assembly because it's a serious QC issue
So he actually tries to explain why there is a high temperature resistance tape there on the PCB, even so that should not be there in the flashlight, also as I see there was no thermal grease anyway.
Ooops, if you are referring to my photos, there was thermal grease there, but I cleaned it all up before taking the picture.
Yes.
Ok, so you cleaned the head base and also the MCPCB yellow tape.
Yep - I cleaned up both sides for the pics. I mentioned in an earlier post they had a white colored grease there, could have been Fujik or a better Arctic grease, but certainly wasn't AS5. I love this design though - big, thick MCPCB screwed down with a direct thermal path in copper - better than the 20 mm SinkPAD's because it's bigger and thicker.
Sorry about that.
I do not think anyone uses Fujik out of the higher-end companies. As for AS5, I do not think anyone uses it, there are better alternatives from Shin-Etsu which is an industrial manufacturer (also OEM for others), you buy with kilograms and I presume that sort of grease can be use much better in a industrial manufacturing plant than a product that caters to retailing.
How do you know it's a direct thermal path copper MCPCB?
I mentioned in an earlier post, actually without certainty, I think it is because I get electrical continuity between the pad and the underside of the PCMCB with a DMM, so I'm assuming so -- pretty sure on PCMCB's without a direct thermal path you can't get continuity. Actually I was hoping AlexGT or anyone else knowledgable with the TN31 star would respond to confirm or not confirm. I was wondering maybe this was changed by ThruNite, not sure, because AlexGT definitely thought it was not a direct thermal path based on what he did. From his OP:
8 Replaced the stock XM-L U2 with a XM-L2 U2 1C emitter from illumination supply, I direct bonded it to the copper of the stock PCB by filing off the central pad’s dielectric layer with a dremel, do not put too much presure and over do it, stop once you see the copper.
In the photo I posted of the PCMCB, you can clearly see the raised areas of the positive and negative lan's, but where the thermal pad is, there's nothing.
I'm doing a TN31 XM-L2 model now, resistor mod and de-doming. Strap wrenches got her open, and driver is identical to the TN31 XML U2 version, even the same label/date on it of "TN30 Rev:B 2012-3-12". Good thing so the same resistor mod will apply. Two more good things: I'll keep the stock XM-L2 U2 and simply de-dome it, and, there's no kapton tape on the back of this copper star this time. Same copper star - still looks to be a direct thermal path so I'll re-use it as is.
They are still using crappy 26 gauge wire - I don't understand it, all that quality and money - why use thin wires? Maybe they are afraid of too much power to the LED?
Just went through the whole thread. This sounds exciting.
A better than SinkPAD stock copper star, ~1300 lumens, and 347kcd sounds pretty beastly. Quite nice indeed!
Just a few questions, have you measured the current on the highest setting? And does the resistor mod only affect the highest level, or all levels meaning there’s no more moonlight low.
Here's the measurements I made along the way for the XML based TN31:
Stock TN31 XML version:
Lumens: 1169 – 1125 (1st # is at start, 2nd is at 30 secs), throw: 128 kcd
After resistor mod:
Lumens: 1411 – 1329, throw: 145 kcd
After upgrading to XM-L2 U2 1A, keeping the stock PCMCB, AS5, removed kapton tape, and de-domed the emitter:
Lumens: 1332 – 1275, throw: 347 kcd (1,178 meters)
Stock TN31 XM-L2 numbers, on fresh KeepPower 3400’s:
Lumens 1,326 – 1,281, throw: 156 kcd
So you can see the results of the resistor mod alone what the impact is, more lumens than a XM-L2 version, but less throw. Unfortunately I didn't do amp measurements - having trouble doing that. AlexGT mentioned in the OP of getting 4.6-5 amps, and stock it does 3.7-4.0 amps, and that seems consistent with the output measurements I've gotten.
Excellent, thanks for the reply.
Though longevity and reliability are key for me, the stock mcpcb appears to do a nice job at pulling away heat (along with as5).
This is a nice result 
Oh - also, I don't see any effect on the moonlight mode from the resistor mod - it's really low even on the mod'ed light. Think someone else mentioned that too.
Dedomed my TN31 some days ago and today I did the resistor-mod.
Didn’t noticed any difference…only the light gets hot faster near the selectorring.
Did you recognize any difference by bare eye? Does your TN31 gets warm in the same area?
Toppel
Well, depends where/how you are checking by eye. Sometime indoors, I'm thinking a 1,400 lumens HD2010 is nothing great, but outdoors in a range of 50-100 meters, wow, incredible amt of light and throw. So I believe to really see the result of the resistor mod on an already de-domed LED, you have to see it stretch it's legs in a real distance shot. The meters don't fool you - every reading i've gotten is confirmed outdoors at distance - kind of all makes sense - I can see a 125 kcd light having the edge over a 110 kcd light, etc. Only thing that really shows up outdoors is the pencil beam vs. shear power throw - Small Sun T10-S vs a BTU Shocker for example.
Also, did you have the kapton tape and remove it? Did you re-use the star and reset it in AS5 or equivalent, or use a SinkPAD?