Thermal Test: Fujik vs RTV Silicone vs Insulcast

Reading my posts, you may think I only do thermal tests…today I’ve built 3 potted drop ins built out a Convoy M1…and yes, did another thermal test.

I’ve seen comments made often about Fujik Heatsink Compound and it’s effectiveness compared to RTV Silicone. I only use Fujik as a thermal thread lock, but saw an opportunity to set up a new test.

The contenders:
Fujik Heatsink Compound (50 ml tube - Silicone Filuid* - greaselike - white……….*yes, they misspelled it on the tube)
Permatex Black RTV Silicone
Insulcast RTVS Potting Compound –2pt

Most of you probably have never heard of the Insulcast, but I’m including it because its what I use to attaching my mcpcb and it’s a good soft set potting compound.

The test setup:

I used a piece of 1”x1”x1/8” aluminum angle 2’ long. I wrapped the angle with monofilament .5mm diameter. The monofilament was wrapped in a spiral fashion around the aluminum angle and the wraps were spaced 1/4” apart. (more on this later)

I cut 3 pieces of aluminum flat stock 1”x1.5” -sanded one side flat to 600 grit. I then spread each of the materials to be tested on a 1.5” section of the monofilament wrapped aluminum angle.
I pressed a piece of the 1”x1.5” flat stock into the test material -I weighted each test piece with an 8oz weight, placed it in a curing oven at 65c for 24 hours.

The monofilament created a .5mm gap between the aluminum angle and the 1x1.5 test blocks…this gap was filled with only the test material (and the monofilament wraps)…any heat passing from the aluminum angle to the test blocks had to pass through .5mm of the test material…consistent gap -no metal to metal contact.

A fly in the ointment: I like the Insulcast because it is thin. When I mount a mcpcb with it, I add about 3 or 4 drops into the pill -yes drops -not globs or smears. Fujik is a little thicker than toothpaste -Insulcast is about as thick as syrup (Aunt Jemima not Karo…there is a difference). When I press the mcpcb into Insulcast, any excess is easily squeezed out from under the mcpcb and fills the voids around the perimeter. Also it sticks, but not as tenaciously as Fujik or RTV silicone making removing the star easier. The Fly: when I set up the test block with the Insulcast, a good portion of it dripped out of the gap (remember, it’s thin) -I would estimate I still had only about 75% coverage -since I’m probably the only using Insulcast, I decided to proceed with the test.

Enough about Insulcast -this is really a test between Fujik and RTV silicone.

I mounted the angle onto two steel spring clamps. I attached a piece of blue masking tape to the top surface of the aluminum test pieces (I used an infrared thermometer to take the temperature readings, the masking tape corrects the emissivity of the aluminum surface). I cut off the exposed monofilament strings leaving only the portion embedded in the test material. I heated the underside of the angle with a propane torch, concentrating on the leg that did not have the test pieces on the opposite side. Each torch pass lasted about 1/2 seconds.

I immediately read the temperature of each test sample and had someone record it. I continued reading right to left -left to right -right to left -ect. I chose the set with the highest temperature difference.

I allowed the assembly to cool then repeated the test a total of 6 recorded times -each run had a different number of torch passes:

Higher temperature means better thermal conductivity.

R1 R2 R3 R4 R5 R6 avg
Fujik 97 105 118 122 136 152 121.7
RTV Silicone 96.5 105 118.5 122 135 152 121.5
Insulcast 103.5 112 126.5 130 142.5 163.5 129.6

all temperatures are F

Well I’m surprised, I always though Fujik had some kind of thermal filler…but it looks like Fujik is no better (or worse) than RTV Silicone…actually, it is just RTV Silicone.

Yup, white tub caulk RTV silicon

Care to test Ultrablack or Ultragrey sensor safe silicon?

Most excellent though…shows that the heat will pass thru it

Nice test and report. Thank you Pflexpro and welcome to the forum.

Do you have any suggestions as to the best place to purchase Insulcast?

Thank you for the test(s). Those are useful tests you do. Keep it up. I sure am reading them :-)

Thank for the test pflexpro, I appreciate the detail test about the setup and testing. Fujik is overrated/very popular and because of that better alternatives are not coming out to the surface, so to speak.

Thanks for the data. Very informative.

I wonder if Stars-922 is just Fujik under a different name, and if so, how many other “Fujik” clones are out there?

The commonly product refereed to as Fujik is BAB 900 (Thermal conductivity 0.88W/m.k)

The ST922 is rated 0.671W/m.k

I do not think they are the same, usually rebrands have the same specs. Who has them both can tell you more about physical appearance.

Am I the only one that uses Arctic Alumina Thermal Paste or Arctic Alumina Thermal Adhesive?

When using the adhesive, I usually do a pre-assemble to have the adhesive set up in actual conditions, then dis-assemble and connect the driver to the emitter. This allows me to get everything centered and exactly in position in relation to how it will fit together in use before the Arctic Alumina Thermal Adhesive sets up, and it sets up quite quickly! (3-5 minutes it’s firm, pretty well set in 12, fully cured in 8-12 hours if I’m not mistaken.)

Thanks for the tests and measurements, knew I didn’t care much for Fujik for a reason…

To me, silicone is an insulator…we use it around here for potholders and oven mitts to protect things from heat! Hmmm, not exactly what the situation calls for in order to conduct heat away from the emitter. I like using copper pills and copper stars and re-flowing the two to eliminate this issue.

I'd like to see Permatex Ultra Grey and JB Weld tested on the same rig (and also a control, with NO thermal compound, will help put the whole thing in perspective).

Might also be useful to eliminate the .5mm spacer and see how they perform when assembled like they would be in an actual application.

All this is making me wonder if it might not be prudent to start screwing down the star to the pill. A solid tight connection is always a good thing, right? Probably not the easiest thing, with possible centering issues in some lights, but still…

It depends... two tiny little screws way out at the edges stand a good chance of either not enough clamping force for the paste/compound to work properly, or, just as bad, if they are tight enough the MCPCB can warp and bow up in the middle, exactly where you don't want a gap.

I suppose the 'right' way to do it would be to use screws no smaller than M3, and very slightly bend the MCPCB so the bottom is convex. The clamping pressure from the screws will pull the MCPCB down flat with maximum pressure in the center right under the LED.

The only problem with AA and AS thermal adhesives are that they are a permanent epoxy. I’d love to use them, but if I have to replace an emitter (especially in a multi emitter light), it’s not coming off without a fight, damage to the pill or a nearby emitter in the process.

I buy Insulcast from a local distributor -the drawback -it’s a 2 part silicone-to purchase, you buy 1 gallon of each at about $125 per gallon -probably not for the budget flashlight builder. I build enough to ‘almost’ make it worth while.

To be a good thermal conductor -it seems silicone needs to be filled with a particle, like aluminum oxide. Since I use both (filled and non-filled), I can somewhat tell if its filled with a thermal particle…maybe you can too. You can squirt out a bead a few inches long on a piece of wax paper, let it dry, then stretch it like a rubber band. If its just silicone, it will stretch about 2x in length. If its filled, it will break after stretching about 10%.

Wow. I tend to over purchase, but that's even too much for me. Thank you for the follow up info.

That’s what I like about Insulcast -the bond is strong enough without making the star too hard to remove.
If I create a star removal scale, it would look something like this:

10 (hardest to remove -don’t bother trying) epoxy

8 (removable but not fun -heat helps) Fujik and RTV Silicone

3 (requires a little prying and twisting) filled silicone, Insulcast

1 (easy to remove with just a twist) thermal grease

The most important is not the type of compound, but the amount of metal-metal contact, in other words: the layer must be very thin, just filling the space. You know why chinese manufacturers get away with hollow pills with no thermal paste?, that is because the metal-metal contact between board and rim is actually very good in this set-up ( fully circular and near the heat'exit'). When I place my led-boards on the aluminium reflow block to reflow the led, I can see the solder-flux start evaporating almost instantly, no thermal paste there, nothing screwed tight, just two good flat metal surfaces touching.

It has a one year shelf life -I wish it had a 20 year shelf life.
I very generous when I mix it…if I doing a single noctigon, I need 2 drops of finished material -I will dispense about 5 drops of each part to make mixing easier. I’m mixing 10 drops to use 2 drops. At this wasteful rate, I will only be able to mount about 18,000 16mm noctigons…but I need to do them all in a year.

lol. You better get started then.