Lapping the Thermalright Ultra 120 Extreme Black
My previous build utilized a
Tuniq Tower 120, and was my first experience with lapping a CPU heat sink. My black theme for this case narrowed down my choices of coolers - namely to the TRUE Black and the Xigmatek S1283 Dark Knight. I liked both, but people implored to me that most of the benchmarks don't show the processor overclocked very high and so the Dark Knight is able to keep up with the TRUE. I intend to have a 3.8GHz overclock 24/7 minimum so I want the best and the brightest, and this landed me with the TRUE Black despite the increased cost.
The TRUE is nickel plated, so I knew going into this that I'd have to sand through a layer of nickel; I wasn't sure what to expect from the black coating, and frankly underestimated the amount of nickel I'd have to sand off the base. I did anticipate needing more sandpaper than last time, though, so I ordered a couple of glassless kits from
EasyPCKits, both of which you can see next to the typical Thermalright discrete packaging:
You can see that I purchased a 1366 bolt-through kit for the TRUE as well. Thermalright does offer a Socket 1366-compatible TRUE, but it is not black. Yay for spending more money!
Thermalright packed the TRUE quite well.
The white box contains the included bolt-through kits and back plates in individually sealed packaging.
Taking the TRUE out of the box, one can see that the black finish is evenly coated over the entire cooler, and that the base is relatively smooth. In fact, it was flatter than I had expected based upon people's comments.
You can see the reflection of the plastic packaging off the base - but don't be fooled as this was not a mirror finish (I wasn't expecting one, mind you). Pictures of lapping results are often taken at angles because this allows a reflection to occur even if the base is not entirely smooth or a mirror finish. This is why I take pictures head on, as I did with my Tuniq Tower 120, to show the true quality of the finish.
I had glass left over from previous lapping kits (2 pieces to be exact), and I've learned that it is easier to put the sandpaper on sideways and tape it to the back.
Duct tape is the handyman's secret weapon!
This creates a flat surface of sandpaper upon which the heat sink base is moved.
I have a larger piece of glass from an old frame that I use as a solid base underneath the sandpaper setup. The duct tape creates enough friction with the large piece of glass that the sand paper setup remains stationary when moving the heat sink over the sandpaper side.
What you can see in any of the pictures is the bucket of water I use to lubricate the sandpaper so that metal buildup does not occur (I wash off the piece of sandpaper every couple of minutes in the sink as well), or the can of compressed air that I use to remove the dirty water from between the heat sink base and the heat pipe cover plate.
An easy way of making sure an even lapping is being provided to the heat sink base is to draw on the base with an indelible marker.
I started at 400 grit sandpaper, thinking I wouldn't have too much to remove. I am glad I had extra sandpaper from the Tuniq Tower 120 lapping!
A couple of swipes later it is apparent that the middle of the base is higher, indicated by the lack of the black finish and indelible markings on the center of the base.
After several more sets of swiping and "x" markings there is still some black finish left on the base.
You can also see water on the sandpaper in this image; a few sprinklings of water was sufficient in preventing metal buildup between washings.
Several swipes later and it is obvious that the base is becoming more level, but still no copper...
This was an hour or so into it and at this point I was about ready to settle for getting the mirror finish on the nickel and calling it good, so I moved to 600 grit.
The base started to look much better, but of course the copper started to appear (see the corners of the base), requiring me to continue until all of the nickel was removed...
I moved back to 400 grit sandpaper, and gradually more and more copper started to appear.
The majority of the copper first appeared near the middle of the base, indicating that it was still not yet flat. I didn't need to use any markings on the base during this transition because the spread of the copper performed the same purpose.
At this point I was becoming inpatient so I gradually decreased the grit number, from 400 to 320, 260, 220, and 180 grit before I started to notice the nickel being removed in significant amounts. It's the wrong order, but I wanted to remove as little material as possible so I didn't want to overestimate the necessary grit size.
At this point, 2 separate areas began to appear where the nickel was apparently much deeper than on the rest of the base.
However, I was getting close to removing all of the nickel so I went back to 400 grit.
I then moved to 600 grit, and ended up having to use 2 pieces of it before the nickel area remaining was small enough where I was comfortable moving to a higher grit number.
After a while longer, the nickel areas remaining eventually were much smaller, but their presence was taunting me so I decided to apply rougher sandpaper directly to those areas to remove the remaining nickel entirely.
Probably not the smartest thing, but I figured I'd get rid of the roughness in later stages.
Indeed, the rough copper patches seen above gradually disappeared as I went to finer grit sandpaper.
I did apply another "x" marking to see how level the base was, and after only a few swipes nearly the entire base marking was gone, meaning that the base was near level. My not-so-straight straight-edge confirmed this. With a flat base, the goal was no longer to remove copper but to get a better finish.
To that end, I did the usual steps of gradually increasing the grit number to get the desired base finish. From the 600 grit used above, I went to 800, 1000, 1200, 1600, and 2000 grit before stopping.
This would be that typical misleading picture angle I mentioned before.
And here is the view head on:
You can see that this finish is not as good as what I achieved with my Tuniq Tower 120. I still have 2500 grit and a few micron-sized grits remaining, but I am saving these to remove any oxidation that develops on the copper between now and when I mount the heat sink (not entirely sure when that will be). Creating a finer finish would not achieve significant gain for me for the effort it would entail, which is why I stopped here.
To help prevent oxidation prior to use, I wrapped the base in Saran wrap and placed it back in its original packaging.
Mmmmmm... Diet Dr. Pepper...
And there you have it!