Quote:
Originally Posted by
charlie310 
So I spent a few hours reseating my Noctua and testing the temps. I used the X method, and noticed no difference in temps from the "spread" method. So far, I'm disappointed with the Noctua. It provided, at best, a 1C degree improvement over my Dark Knight. Maybe a lapped Dark Knight with 1x120mm Yate Loon Fan is comparable to a stock Noctua dh14. Perhaps the case airflow has something to do with it as well.
One pro I see is that the Noctua runs noticably quieter than my DK.
In any case, I'm going to lap the Noctua whenever I have some more time. But, as far as I can tell, it is not worth the money. I bought the DK for $20, the Yate Loon fan for $9, and sandpaper for $5. That totals $34 vs. the Noctua I got for $75. Paying more than double for lower dB is not worth it IMO.
So, I spent a few hours lapping my Noctua. I used 1 sandpaper each of 200, 400, 800, 1000 grit. The contact plate is actually micro-grooved, which obviously makes is far from flat. The nickel plating is actually really thin, and I was able to sand it off within seconds of lapping. I didn't lap for a mirror shine because I've read that lapping with higher grit than 800 doesn't help with performance and is just for aesthetics. When all was said and done, I used a rice grain of the Noctua TIM that came w/heatsink. I fired up my PC and tested temps w/Prime95.
I got a 2 degree improvement, which is pretty good considering the stock Noctua was only a 1 degree improvement over my Dark Knight. You can check my rig for OC settings. I tested it at the same room temperature as my previous test of my stock Noctua. Keep in mind that the TIM still hasn't "settled", so I expect temps to go down by 1 degree on top of the 2 degrees. I really don't know why more people don't lap their heatsinks. It really does make a big difference, especially for budget heatsinks like the Dark Knight and Hyper 212 (since the contact surface on those are unbelievably uneven). You only have to spend $6 on sandpaper (buy an assorted grit 4-pack at Autozone). It only takes a couple hours, and the results last a lifetime.
It actually took me 3x as long to lap my Dark Knight because of how much nickel plating I was sanding off (harder to sand than copper), and because of how uneven it was. I'm surprised that a well-lapped Dark Knight with a 2200RPM Yate Loon Fan has almost the same cooling performance as a stock Noctua D14.