Arctic silver vs Arctic alumina? Cast your vote.

Arctic Siver vs Arctic Alumina?


  • Total voters
    16

Lumencraft (Matt)

Flashlight Enthusiast
Vendor
Joined
Jun 19, 2008
Messages
3,223
Location
Missouri
I have been a cpf member for long enough now that I have built, rebuild and modified a reasonable number of lights. I have used both AS an AA to create a thermal path between leds and heat sinks. What I like about the AS better is if needed the led can be removed. Not so easy with AA, in fact I have destroyed emitters trying.

So whats your opinion and why?
Has anyone had an emitter on a light come loose and get damaged as a result of using AS rather than an epoxy?

A. I prefer Arctic Silver
B. I use Arctic Alumina (epoxy)
C. I use another product
 
Last edited:
The Arctic Silver Epoxy is electrically conductive while the Arctic Alumina Epoxy does not conduct electricity (some but not well), So it can be used for electrical Isolation too.
 
Both are electrically non-conductive. Arctic silver has a small amount of capacitance, so is not recommended for driver boards. AA has no capacitance but has slightly lower thermal conductivity.
 
I often use both. I use AS for optimum thermal path but also a little of AA Adhesive (on the side of the mpcb) to fix the LED's position. AS stuff is a PITA to clean though :popcorn:
 
Can't vote, I use both.

If I need some adhesive and thermal transfer it's AA and if I need just thermal transfer I use AS 5.
 
I mostly have used the AS in "pill" type lights because in my case it meant that the reflector would aid in holding the led to the heat sink, and in my xp. it does create a better thermal path. With the 2 part epoxy it seems there is far less room for error. Dont get me wrong it does work but getting a very thin very even coat seems a bit harder to do. I will say that the P7 light of mine that has probly seen the most real world use is glued with AA. It has been driven fairly hard for extended periods of use (in a cave) and still working fine, in fact as I recall this was my first P7 build.
 
The Arctic Silver Epoxy is electrically conductive while the Arctic Alumina Epoxy does not conduct electricity (some but not well), So it can be used for electrical Isolation too.
Arctic Silver is not electrically conductive, it's electrically capacitative. Because the individual silver grains are made of metal, an electrical charge in close proximity can induce a sympathetic opposite charge in the grains -- but that will happen with an anodized aluminum heatsink too, because the anodizing is more than thin enough for slight charges in close proximity to cause induction in the aluminum.

Now, if you try to put a 1000V charge on the back of the LED, then yes, there will be some leakage, but if you expect any heatsink paste (or anodization layer, for that matter) to perfectly insulate the heatsink from that charge, you're out of your mind. But for low voltage applications it's a complete non-issue.

Think about it; if the electrical capacitance of Arctic Silver were really a problem, they wouldn't use it on bare silicon CPU dies, which are MUCH more sensitive to current leakage than LEDs are. But it's not a problem, so it's fine.

Use Arctic Silver. Metal conducts heat better than ceramic does.
 
They make Arctic Silver epoxy, you know. :thumbsup:

I prefer AA epoxy as it is an insulator, no capacitative characteristic :)
Often I put a couple of blobs of AA Adhesive around the LED (or aluminum base of the LED) just to make sure the led isn't moved around freely, but still easily removable with a bit of force when desired.
 
Last edited:
Think about it; if the electrical capacitance of Arctic Silver were really a problem, they wouldn't use it on bare silicon CPU dies, which are MUCH more sensitive to current leakage than LEDs are. But it's not a problem, so it's fine.

Use Arctic Silver. Metal conducts heat better than ceramic does.

Problem is : I have ruined CPU's with Arctic Silver.:mecry: I have also unlocked or changed CPU multipliers with a line from a pencil.
The CPU dies are not bare. AMD have some chips that do have bare spots on them and Artic Silver will destroy them if you are not careful.
There are people who have fixed electrical traces on circuit boards by making their own Arctic Silver/EPoxy combinations.
 
Last edited:
The AMD Athlon XP in my home computer has a chip in one corner, and if you look really really closely you can see a perfect 90-degree angle inside the chipped area, perhaps the edge of the actual circuitry inside the die. I have used AS5 on it and other AMDs with similar damage for years and they all work perfectly.

Most AMD Athlon CPUs were actually destroyed by torque from the heavy heatsink bending the weak die carrier and breaking wires inside it, because they stopped using ceramic carriers and switched to something resembling regular circuit boards instead. These could not stand up to the stress of having a heatsink clamped onto them for long periods of time, and nowadays the only ones that still function are the ones that were braced with thin metal shims that redistributed pressure around the outer edges of the carrier, directly above where the pins lock into the CPU socket. Newer CPUs deal with this problem by using aluminum covers installed at the factory, which serve several purposes: conducting heat from the die to the heatsink, redistributing the heatsink clamping pressure to the edges of the die carrier, and preventing the die from being chipped. In no case is the heatsink thermal compound to blame for the failures of AMD CPUs -- assuming the user used it properly, of course. There's no accounting for user error.

I have used AS5 and AS epoxy on dozens of CPUs and motherboard chipsets, as well as modding a couple of flashlights now. AS5's electrical capacitance is not, and never will be, a real issue, and Arctic Silver would've done better to have never mentioned it as a possibility in the first place. They ruined the market share of what should be the most popular product of its kind on the market, by giving people a reason to blame it for their own mistakes.

EDIT: Without seeing the circuit diagrams of the circuits "repaired" using Arctic Silver epoxy, I can't say why they worked, but my guess is the traces in question carried signals that changed voltages at extremely high frequencies, in which case the capacitance of the Arctic Silver would be sufficient to transmit the signal down the wire. It would not work in a situation with a constant voltage being applied across the damaged area.
 
Last edited:
Top