Trainwrecker's Heatsink Tests

[Mikewarrior2]

Hey Everybody,

Not exactly real world testing, but some nice steps in the right direction and a nice "ranking" of heatsinks.

Trainwrecker's COP tests


Mike

P.S. The Acclaimed G-Orb was equal to the INtel Retail Sink.

 

[bcterps]

Thanks Mike! The more info the better. Its hard to find reliable info on heatsinks, I'm sure I'm not the only one that its pretty confused, even after looking at your pages hehe. I appreciate all the hard work though!

--Ben

 

[GiZzO]

Very interesting how the hedgehog handled a 52 peltier which is simulating a Thunderbird/Duron did pretty poorly, with with the intel sim (32w pelt) scored the best.
Pure bulk of the PEP66, kept the pelt hot side about the same on 52w and 32w, tho i think hes got his numbers backwards...hes got the 27 cfm fan doing better then 38.

 

[Mikewarrior2]

Gizzo,

Higher numbers=Better...(Tr-Tc). He's measuring the Air temp- Cold Side Peltier temp...

So The intel and G-orb tie for worst(the peltier + either these guys = very few C difference), whereas the hedgehog has some nice cold-side temps... Not very realistic, but can be an effective "heatsink-ranking" system.


Mike

 

[GiZzO]

I was comparing the hot side of the pelt

i guess most important would be the cold side, that would show how effective its pulling heat away, and i guess would result in a hotter hot side, doh um still learning how to read the peltier temps.

 

[Mikewarrior2]

why would you want the hotside hotter? The more effective the CPU is at cooling the peltier, the cooler both hot and cold side is(ie pulling the 58W of heat away from the peltier).

The hotside, with an inefficient heatsink, will "hold" the heat longer, allowing the hotside to heat up because the heatsink is incapable of disappating the 58W of power.


Mike

 

[ALBPM]

Sorry about the confusion GiZzO.

I wasn't real clear on how to evaluate the results. Look at the cold side temp and then the Higher the T(t)-T(c) value the better the performance.

The COP value or Efficiency score shows the design efficiency and how air flow or CFM can can affect this.

So the Hedgehog with a 27 CFM fan will out perform all the other Heat Sinks tested so far.

The one thing I wanted to show was how each of the Heat Sinks would perform with 23, 27 and 38 CFM fans. This was so the end users could decide if the performance gain is worth the extra noise levels, etc..

I will be adding Taisol and Swiftech Heat Sinks to the database in the near future.

Paul