skyray king "pill" mods

so i popped the top off my king and was looking at the disc that the LED’s mount to and it looks a little thin for having 3 LED’s pumping heat into it. I borrowed this image to show what i was talking about. that aluminum disc that the LED’s are mounted to looks a little thin and is only contacting the body of the light via the threads and a little “shelf” that it sits on.

has anyone thought of recreating this in copper or brass and making it thicker?

Here is a MS Paint drawing of what I am thinking of.

looks like there is a little room to make it go deeper into the driver cavity but that would require a step just below the threaded area to clear the light. if someone were to machine it super precisely, the stepped thicker flat section could make contact with the wall of the light (perhaps with the aid of some fujik) and transfer even more heat to the body, making the thermal sag less.


I think that making it thicker will not help at all, you will have more thermal inertia (I don’t know if it’s the correct term), I mean that with thick copper you are going to need more time to heat up it, but once it’s hot it’s the same. What you need to take apart the heat is a lot of surface, like fins in a heatsink.
A mod that could help, is putting XM-L with copper base, in this case it do help because the led generates a lot of heat in a little surface and with copper you do a better job.
I would like to install XM-L U3 with copper base in my KING… but not sure how much could it improve visually, if it worth the swap or not….

I think it would help. To optimize the use of the copper, you want to maximize the surface area with which it can contact the body of the flashlight. I have tweaked your picture to illustrate what I am thinking (and did with my SST-90 mag). Bore the body out through to the driver area and reshape the heatsink so it will replace the ridge that was milled out. This will create a ‘H’ section with the legs of the H in contact with the body. Copper can conduct heat faster than aluminum so if you give it a large path to the aluminum body the thin plate of the heat sink where the stars mount won’t be as large an issue.

I accidentally hit the wrong button?

-The thicker copper will ( as mentioned ) take longer to heat up, however, it will take much longer to cool down.

—-> A larger path to outside the system will draw heat out better not worse

-The amount of heat being transferred to the body may not increase because of the larger contact area,

—-> Increasing the contact area between the heatsink and the body will definitely increase the amount of heat transfer to the body. Ever played with copper foil around a P60 drop in?

there is a chance the heat may stay inside the body - the amount of air within the body creates a thermal buffer, helping to dissapate the heat before it heats through to your hand. Decreasing this pocket of air decreases the amount of room the same amount of heat has to move around in.

—->Air is an insulator and replacing it with a highly thermally conductive material (Copper) that is directly in contact with heatsinking (the body of the light and on to your hand) will definitely get the heat out faster than pumping the heat into a void

I don’t think this mod will make noticeable improvement.

thare are large thermal sag between LED’s thermal pad and pill
used in metalcore PCBs have poor thermal conductivity on it dielectric
and emitter board is glued to pill. this also increase thermal sag.

A simple mod is to add mass , so why not just glue a disc underneath , more mass should be better [ should ] ..

I have done this before [ add metal ] to a pill thats to lightly built , just make sure you improve the thermal path as well [ side contact ]

What you want !

Is for the pill to transfer heat , before becoming too hot itself .. Now not enough metal [ light pill ] will mean the pill heats up fast , faster than the amount of mass can transfer heat away from the source .. Can you have too much mass ? wow . err possibly , but in a flashlight ? well one thing is sure , it cant be worse ..

Yes more mass will hold heat longer , but it will also transfer it faster ..

Actually think of it this way .. A 6 lane highway will allow twice as many cars to travel at the same speed as a 3 lane highway ...

i think ill just put my light back together and use it then lol.


I think you may be may representing the problem. There is already something between the torch and the steel, an inadequate heat sink. What has been suggested is a more massive heat sink that also has more surface contact with the housing. The existing pill is a bottleneck in the path to ambient. It heats up too quickly and cannot act as the buffer it should. If the heat sink has enough mass and the surface contact area is great enough, then the heat generated relative to the heat dissipated will stabilize.

In the case of a P60 with XM-L and high current (ie: 2.8A or 3A) it produces more heat than it can dissipate. in my experience, what happens is that the LED starts somking and the body doesn’t get hot, but once you do the aluminium paper trick, the body gets hot and smoke from the LED shouldn’t happen again.
In the case of the KING, I think that the pill is good enough to repel heat, a friend accidentally turned on my KING and left it in maximum until battery protection cut the current, and all that time that the flashlight was in high the LEDs never produce smoke but the flashlight’s body got really really hot. For me, that represents a good pill with no bottleneck effect.

Or it may be that your friends king is suffering from corrosion resistance and isn’t sending max current to the LEDs. Also, some people mod the driver to increase amps increasing the need for improved heat transfer.

No, mine is clean and works awesome, it puts around 2200Lumens OTF.
I’m going to do the driver mod to get more current as soon as I get SMD resistors, and then I’ll see if heatsink is good enough. If 3A is too much maybe I should mod it to something like 2.5A for example…
But in stock current condition, I think it doesn’t need pill modding.

Sorry, your light. That’s bad form on my part. To me, the point of the thread was whether or not altering the heat sink could improve the transfer of heat from the emitter junctions to the host and I believe this is so. Would better/copper pcb’s help as well? Likely, if they are well lapped and screwed down with a thin layer of compound. If stuck in with gobs of Fujik, maybe not. Increasing the contact area between the sink and the host will improve things also(#2, nice picture Ford Perfect). Best would have to be a Match-made or ChicagoX-made sink with the emitters direct soldered to it and the sink milled for a precise fit. The driver is most susceptible to heat when it is stuck to the same pill as the emitter(s) as in p60 drop ins. Air being an insulator does a fine job with only minimum of space(1-2mm). Which mod provides the most benefit depends on where the bottleneck(s) are and these will vary depending on assembly methods. Good luck with yours.

Well , if your theory on heat transfer works , fire up your computers cpu without the cooler on it ! [ The CPU will die ]

We are not talking about a gas torch here , we are talking about thermal paths , An EMITTER is much like a CPU , both have a function to perform , and the bye product of that function is heat , and this heat needs to be removed - transferred - conducted away from the source .

Now , the type of material and how much of it there is , is directly related to removing heat from a source ...

Now if you like blow torches , here is a correct analogy :

Take a coin , a small one , and put the hottest part of the flame to the coin , QUESTION ! how quickly does the coin heat up ? how hot does it get ?

Now lets try that again with a one ton block of steel , put the hottest part of the flame to the block , how quickly does the block heat up ? How hot does the block get ?

More mass = better thermal absorption , better transfer of heat [ larger path ] and better dissipation ..

This was an extreme example .

That has sense, but in my experience, that didn’t work.
Look at this custom bike light: Yet Another Custom Bike Light
I put it on the freezer, and after 20min it was TOO hot to touch, but when I insert the KING on the freezer for more than 20 minutes, it is barely warm (I can say it is at ambient temperature).
When I designed it, I tough that MORE material was BETTER, but I realized that it wasn’t that way.
The first time I saw the KING, I remembered that custom light and tough: “If this custom light gets HOT even in the freezer, the KING is a crap”, but no, the KING is always cooler…

If correctly utilized , bigger is better ! Just look at CPU coolers .. They get bigger for a reason ....

Now obviously there are limitations - size of flashlight , amount of space available , etc ...

Torches use aircooling because its too expensive to incorporate water cooling , as for engines , even motor bike engines these days use water cooling , why ? because water is more dense than air and can transfer more heat in less time ..

As for how much material goes into an engine - Wow - You have it so wrong ...

Motors are the way they are because of weight , and they are water cooled , again to keep weight down . If air cooling was anywhere as effective , there would be more aircooled cars around .. How many aircooled cars does Ford make ? GM - Toyota - Hyundai - ?

Aircooling is only superior to no cooling ..

And as for enclosed or not , no difference , the laws of physics still apply ..

It comes down to how something was done - if something is done poorly or badly or in such a way as to be detrimental - its not the fault of the laws of physics , one must look else where .

Say you have a penny that has been heated with a torch and you let it hang from a piece of small thread. Then you take another penny that has been heated to the same temp and placed in between two large pieces of copper. The penny sandwiched by the copper will cool down faster because the copper can pull the heat away better than air can.

Making a larger pill that has more contact with the body will allow the heat to travel through more surface contact. The only place for heat to get out is through the body so the faster you can get it to the body the better.

Keeping a flashlight cool is basically a two-step-process: First, get the heat away from the emitter and into the body with a good thermal path. Then you have to transfer the body heat to the surrounding air (increase surfeace area - cooling fins!)

The bike light mentioned above, looks like it has a very good thermal path, but it has a small surface and almost no cooling fins. So the heat just accumulates in the body, it gets hotter and hotter and has no ways of getting rid of the heat fast enough.

Sky Ray King has a bigger surface and therefore stays cooler.

Exactly, in case of PCs heatsinks, both CPU and GPU, when MORE material it is better, yes, but the important part is to have MORE SURFACE. Look for example a CoolerMaster V8, it weights near a kilogram, but aluminium is really thin and has A LOT of surface in contact with air:
Now imagine the same heatsink, but SOLID. It may take more time to heat up, but also it will take much more time to get cooler. In my custom flashlight, I didn’t have that in consideration, I though that the only important thing was to use a lot of material, but the important thing is to use a lot of surface.

In the case of the KING, it transfers correctly heat from the pill to the body, so I don’t really thing it will get any better with a bigger pill (at least with stock current).

Exactly. But the OP didn't only want to add mass, he was also looking to improve the thermal path. The analogy to the cpu cooler would be to add more heatpipes frome the copper core to the cooling fins.

Now the question is if the King's thermal path is already very good or is it just good enough with room for improvement...