http://systemcooling.com/thermalright_xp-90c-01.html
Admin Edit for hotlinked image.
XP90-C review
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
-
frankgehry
- Posts: 1424
- Joined: Sun Dec 12, 2004 5:00 am
- Location: New York, NY
That's a pretty good review, but with a couple of nagging flaws that leave an unsatisfying aftertaste:
1. Why not test them all with the same fan, or at least adjust the voltage on the 120mm to get the same CFM. Wouldn't affect the results, but it'd be interesting to know by how much the 90C beats the 120 with the same airflow. Apples-to-apples, you know.
2. No attempts at testing them with anything besides the mobo-controlled fan speed. That introduces a mess of variables. Since the load temps are different, the fan voltages may be different, so essentially you're testing the thermal-control system's ability to adjust the fan speed to hold the temp under a given set point (could be 50°, judging by the results) rather than purely testing the heatsink itself. The XP-90 could be right at some critical set-point in the fan speed ramp-up, and be running at 10v, while the 90C is just below the ramp-up point, running at 7v.
3. Why use the dubiously accurate Enermax diode for temps? The CPU itself has a perfectly fine diode right in it, and after calibration it will produce comparable, even if unaccurate, results. Each time you remount the heatsink, the chances of you sliding that thermal probe back in exactly the same way, at the same contact point and pressure with the IHS is nill. You're better off with results that are consistently wrong than results that are randomly wrong.
It may sound like nitpicking, but when you're talking about a <1° spread between first and last, the little things can make a big difference.
The results themselves aren't surprising though...knowing what we know about the XP120 and the XP90.
1. Why not test them all with the same fan, or at least adjust the voltage on the 120mm to get the same CFM. Wouldn't affect the results, but it'd be interesting to know by how much the 90C beats the 120 with the same airflow. Apples-to-apples, you know.
2. No attempts at testing them with anything besides the mobo-controlled fan speed. That introduces a mess of variables. Since the load temps are different, the fan voltages may be different, so essentially you're testing the thermal-control system's ability to adjust the fan speed to hold the temp under a given set point (could be 50°, judging by the results) rather than purely testing the heatsink itself. The XP-90 could be right at some critical set-point in the fan speed ramp-up, and be running at 10v, while the 90C is just below the ramp-up point, running at 7v.
3. Why use the dubiously accurate Enermax diode for temps? The CPU itself has a perfectly fine diode right in it, and after calibration it will produce comparable, even if unaccurate, results. Each time you remount the heatsink, the chances of you sliding that thermal probe back in exactly the same way, at the same contact point and pressure with the IHS is nill. You're better off with results that are consistently wrong than results that are randomly wrong.
It may sound like nitpicking, but when you're talking about a <1° spread between first and last, the little things can make a big difference.
The results themselves aren't surprising though...knowing what we know about the XP120 and the XP90.
-
Dragon Puppy
- Posts: 13
- Joined: Sat Apr 24, 2004 6:34 am
another review
[DeLurk]
I don't know if its been posted , but overclockers have an review of the XP-90C here http://www.overclockers.com/articles1211/ .
[/DeLurk]
I don't know if its been posted , but overclockers have an review of the XP-90C here http://www.overclockers.com/articles1211/ .
[/DeLurk]
