Copper Wrap Variations and Tests

On a previous post, I tested the effectiveness of various thermal wraps. In a nutshell, Copper Foil was the best while Copper Foil Tape was surprisingly bad and aluminum foil was somewhere in the middle. After seeing how well Copper Foil performed, I decided to test some variations. One heated question (pun intended) is ‘what effect do the multiple wraps of copper foil have on the thermal transfer rate’.

My personal goal is to build drop ins with the lowest possible Junction Temperature. I don’t want to be concerned with how long I can run my light on high or how much lumen drop I will have with extended run times. Driven at 3 Amps, my drop in has a stable maximum Tj of 87°C at 28 minutes while being hand held. Standing on it’s tail cap, my Tj is 112°C at 28 minutes. At these temperatures, I can run the flashlight with maximum output as long as the battery will let me, but I still looking for ways to save a few more degrees.

I have tested three copper wrap materials:
4 mil copper (fewer wraps, fewer wrap layers)
2 mil copper (more wraps, more wrap layers)
Copper corrugated (fewest wrap layers, but smaller thermal pathway)

Wrapping methods:
4 mil Copper: 4 mil copper wrapped around the reflector shaft until slight insertion resistance was achieved, then it was finished with a 2 mil strip for a tight fit.

2 mil Copper: 2 mil copper wrapped around the reflector shaft from start to finish.

Copper corrugated: 2 mil copper was wrapped around the reflector shaft 1½ times, the corrugated was slipped on and another wrap was made with 2 mil copper. The 2 mil copper is a continuous strip about 6” long.


Which was easiest?

  1. The Copper corrugated was the easiest because the number of wraps was not critical. If too many wraps of 2 mil were added, the corrugated will compress. The corrugated is reusable a few times, but a longer strip of 2 mil may be necessary for a good fit. The copper corrugated should not be used without the 2 mil strip on top and bottom because it doesn’t stay tight – I’m not sure why, but it becomes loose.
  2. The 4 mil copper was easier than the 2 mil just because there were fewer wraps to deal with.
  3. The 2 mil copper took a lot of wraps and depending on your host, it could be an insane number of wraps.

.
.
.
Preparing the Reflector:


This is important! The reflector shaft is the area where the wrap picks up the heat from the drop in. The contact area of the reflector shaft is about 1.4 square inches…or is it? When measuring, I subtracted the area of the groove in the middle of the shaft (there may be 1, 2 or 3 of these grooves). The large groove is not the only groove that should be subtracted from the total area to yield the usable contact area…you probably have about 70 more micro-grooves to subtract…if you don’t have these micro-grooves, your lucky. These micro-grooves reduce the usable contact area from 1.4 square inches to about .7 square inches. Removing the groves will double your wrap contact area and give a very big thermal improvement. I use 600 grit sandpaper, followed by 2000 grit sandpaper. For the final finishing, I use a thermal grease as a polishing compound. Most thermal compounds have a fine abrasive grit which will give a excellent finish. When sanding, aluminum oxide will form on the surface before you can remove the sandpaper and look at your work. Using a thermal grease as a final polish will remove the aluminum oxide and block oxygen from the surface so aluminum oxide doesn’t re-form. You can wipe off virtually all of the thermal compound and still have a grease layer thick enough to inhibit the formation of aluminum oxide.


.
.

Preparing the Reflector for sanding and polishing:

It’s very easy to damage the reflective surface of your reflector…it’s so delicate, it’s almost non-cleanable.
Before starting, I seal the reflector as follows:

  1. Tear off a piece of masking tape (I use 3M Blue Painter’s Tape).
  2. Place the tape sticky side up on a cutting board and press the open end of the reflector onto the tape.
  3. Trim the excess tape with a sharp knife.
  4. Put a small piece of tape over the small emitter opening.

Sanding the reflector: I use 600 grit, 2000 grit then polish with a thermal compound. You can use only 2000 grit sandpaper, but it takes longer. (sanding and polishing time –10 to 15 minutes.
.
.
.
.
.
Now your ready to wrap

I use 5/8” wide copper –it doesn’t tend to slide as much when I insert it in the tube, but there is one thing to consider. If your host uses the large spring, be sure the diameter of the spring is small than the reflector shaft –It’s usually not –especially after sand a few thousandths off the diameter of your reflector shaft. If you wrap over the spring, and the diameter is larger than the reflector shaft, when you put it in your host, it will be like trying to put a cork into a wine bottle…upside down.
I trim the first 3 inches of my wrap down to 1/2 inches wide. Using the 5/8” wrap has a few other advantages: the insertion end can squeeze smaller and grip the spring or around the end of the reflector–this stops the wrap from sliding up the reflector when you insert it. It also gives a little more wrap contact area on the inner walls of the flashlight tube.

I always wrap the copper strip counter clockwise facing the open end of the reflector. When I insert the drop in, I twist clockwise. If you wrap clockwise and twist the drop in into the host counter clockwise, the pill may unscrew from the reflector.

Wrapping with thin copper (2mil):
This was the hardest of the three tested…mainly because it took a lot of wraps. In the host I used for testing it took about 12 wraps.

Wrapping with thick copper (4mil):
This was the best performing wrap. This method took 5.5 wraps of 4 mil and 1 wrap of 2 mil. This is actually a dual gauge wrap –the 4 mil was used for the majority of the wrapping but, for the final fitting, I used a few inches of 2mil. I wrapped the drop in with 4mil copper until it had slight resistance going into the host –then I slipped a short strip of 2mil copper under the last layer of 4 mil until I had a tight fit. Besides having the best performance, there is another advantage –the thicker copper strip is more durable and can be reused many times –the short strip of 2 mil can be replaced to easily adjust the fit if you need to take the drop in out or put it in another host.

Wrapping with Copper corrugated:
I’m still working with this idea. The copper corrugated has the disadvantage of large air gaps surrounding the corrugation, but makes for it by providing a near seamless path from the drop in to the host. The corrugated still needs a 2 mil wrap underneath and on top. Without these two layers, the copper corrugated becomes loose. When wrapping with the copper corrugated, I start with a 2 mil wrap, 1.5 times around the drop in, slip the corrugated on the drop in (the 2 mil slips through the slit in the corrugated, then I continue wrapping until I have the fit I’m looking for. When inserting the corrugated wrap, the corrugations compress for a good fit –even if the wrap was made too thick. Reusing the copper corrugated usually means replacing the 2 mil wrap with one that is a little longer.
.
.
.
Now The Results:


.
.

Conclusion:
As expected, the interface layers do reduce thermal transfer.
In order of performance:
4mil with 2mil finishing strip (8 interface layers)
Corrugated (5 interface layers but much less surface contact area)
2 mil (13 interface layers)

At 5 minutes, best to worst had a spread of 4.9° C solder point temperature.

Even though the 2 mil copper wrap finished last in this test, it still outperforms aluminum foil wrap and everything outperforms copper foil tape.

The 4 mil copper wrap with a 2 mil copper finishing strip seems to be the best thermal wrap…for hosts with large openings, a 5 or 6 mil copper wrap with a 2mil finishing strip would probably work best, but I wouldn’t suggest anything thicker than 6 mil.

Cost of copper: (based on a Solarforce L2 series host)
1 –12.5” strip of 4 mil would be the base wrap and is reusable many times.
1 –12.5” strip of 2 mil would allow several wraps / rewraps (3 or 4)

4 mil 6”x100” (enough for 72 average flashlights) $36.40
2 mil 6”x100” (enough for 288 wraps / rewraps) $29.43
Cost per wrap: $.92

(I’m not suggesting someone has 72 P60s to wrap –but someone could buy the material and cut each in 6” x 12.5” strips and sell the packs to other members –each pack would be enough for 9 wraps)

Again neat testing and information!
This helps me a lot.

What does 2mil mean and how does it translate to metric numbers?

Edit
I’m in for one pack of strips, who’ll organise the groupbuy? :stuck_out_tongue:

2 mil = 0.0508mm

-> Convert mils to mm - Conversion of Measurement Units

1 inch = 1000 mil

2 mil is .002” or .05 mm
4 mil is .004” or .1 mm
I use 2 mil because I have 2 mil
It’s not critical for it to be 2 mil, but thin enough to make fine adjustments

the little things add up, don’t they?

I never would have considered polishing the reflector or that the machining grooves were such a negative in heat transfer.

Thanks for the information.

Who has the copper wrap material ?

Thank you all for details regarding mil -> mm.

Awesome to see the data. I’ve also wondered whether the microtexture on the reflector would inhibit proper heat transfer. Looks like it’s time for me to get out the sand paper and polishing compound.

I think the inside of solarforce hosts also have this weird texture from machining.

EDIT: I don’t know if you saw my post in your previous thread but lilkevin has a mod to move heat from the star directly to the reflector which should greatly help heat transfer.

Thread mentioned is located here :wink:

Nice test results. Are the pill/reflector threads bare or do you use any sort of thermal compund?

Please post the link to keep all useful info together.

@pflexpro

Thanks a lot for the testing and your time put into it!

It's really appreciated.

Could you by any chance also do a direct wrapping without the polishing and sanding, just for the reference!?

I wrap my threads in .7mil copper foil with a small amount of heatsink silicone on both pill and reflector threads. I like to lock the threads because I also install a thermal pad on the top of the MCPCB. This heat path is bypassing the pill and going directly to the reflector…this lowers the Tj and directs some emitter heat away from the driver.

I actually have test data on all the thermal modifications I’ve made…I will have to look it up -I think, at the time I ran that test, I was using a straight 2mil copper wrap.

Well hell, if that's all that matters then skip the wrapping and just remove the battery.

Great info .

Thanks for sharing .

table(table#posts).

Test 4
Noctigon MCPCB
XM-L2 U2 1C
Isolation pad: plastic
Thread wrap: yes cu
Thread silicone: no
Shaft polished: no
Wrap: 2mil cu
Power: 3.00A (power supply)
Ambient: 24.9 C
Host: Solarforce L2N

Test 5
Noctigon MCPCB
XM-L2 U2 1C
Isolation pad: plastic
Thread wrap: yes cu
Thread silicone: no
Shaft polished: yes -AS
Wrap: 2mil cu
Power: 3.00A (power supply)
Ambient: 25 C
Host: Solarforce L2N

Test 4 Test 5
min Tsp Tsp
5 75.5 69.7
10 88.8 82.3
15 95.8 89.0

Differences between test 4 and 5: Test 4 -stock shaft Test 5 -sanded and polished shaft
Test 4 - 24.9 C ambient Test 5 - 25.0 C ambient (no data adjustment was made)

Please note: If you add up all my thermal improvements, I should be running 25 degrees better than baseline…when all improvements are tested together -it only lowers the Tj about 15 degrees C (they’re not additive)

…haven’t thought of that, but I am making some headway with liquid nitrogen…

I prefer the term “thou” to “mil” because mil sounds like a metric unit but is an inch unit.

Where do you get copper foil?

What did you use as a thermal pad to get heat directly from the board to the reflector, any pictures?

Try doing the test with thermal compound on the copper strip… without it you have air gaps that will behave a lot like your micro-grooves.