I have a lot of work to do today, but dropped in here to check "progress" and post a query of my more knowledgeable brethern per the desired "SMART" temperatures of hard disks.
The XP-90 was a great cooler for a number of reasons. Per your Allendale, I see you have it overclocked a full 50%. I'm mostly guessing that the TDP spec for that processor at stock settings was only 65W, just as it is for my E6600. I'll have to check. Other Core-2-QUAD processors, like my Q6600, are rated at 105W -- a bit toastier.
However, since you OC'd it from 2 to 3Ghz, you must also have bumped up the VCORE. The question then is "how much?" And an increase of 50% in the overall speed would be another linear boost in TDP, while the voltage-delta would increase thermal power proportional to the voltage-delta-squared.
Before you go out and buy another cooler -- and I can recommend a couple -- see what you can do to improve case airflow and improve the efficiency of the XP-90. The idle temperatures still seem too high. You can use the idle temperature alone to gauge progress in improving airflow and case temperatures.
The XP-90 has a thermal resistance somewhere in the range of 0.13 to 0.14 C/W. You can do better than that. For instance, the now-dated XP-90C or Cu has a thermal resistance of only 0.11 C/W, which I believe marginally trumps the original Ultra-120 cooler. Since the Ultra-120-Extreme probably can show a thermal resistance of only 0.095 C/W with the proper airflow, and since the Ultima 90 is only slightly less effective than the U-120-Ex, I'd guesstimate that the Ultima 90 probably fits in the range of 0.10 to 0.11 C/W.
Also, I think you were referring to the Ultra-120-Extreme's IDLE temperatures, as opposed to LOAD values. With a properly ventilated case, you could expect a moderately over-clocked E6600 or even Q6600 to show idle temperatures between 22C and 30C depending on room ambient. The peak TCASE temperature in the Anandtech review for the ThermalRight coolers falls between around 47C (120-Extreme) and the mid-50's.
Those of us who're overclocking the Q6600 processor are showing TJunction (core) temperatures between the high 50's and the low 70's, depending on the overclock setting, the efficiency of airflow, the room ambient and other factors. I consider the low 70's to be unnecessarily warm. @ 78F room-ambient, my Q6600 2.4 @ 3.2 Ghz peaks at 65C for the hottest core. But since these processors (probably) all have similar thermal limits before they throttle, I'd say 70C -- while it may be safe -- is still too close for my comfort.
I'm making an educated guess (based on limited information) that if you're reporting core/TJunction temperatures, the XP-90 is adequate for stock and mildly over-clocked settings, but if you want to do better, there are more recent models of coolers by ThermalRight and others which can improve. And I don't think your overclock setting is "mild." Remember also -- if there is a matter of how your case is set up, or whether you want to use the same fan -- the XP-90C is the same shape, size and basic design -- with a much better thermal resistance, although heavier for it's being copper. And to achieve that thermal resistance, i think you'll need to run your 92mm fan up to around 4,000 rpms.
I can suggest some tedious mods that offer a payoff for any of these coolers, such that you would construct a duct for the cooler so that its exhaust is ported directly onto the motherboard components and then to the case exhaust fan, and you'd want to minimally purchase a $5 piece of foam-art-board, the right glue, an Xacto knife and other few incidentals.
But if you were willing to take the time and trouble to do that, I'd think you'd be inclined to also get a more effective cooler. Otherwise, I'd suggest backing off your OC setting so you can lower the VCORE enough to reduce the CPU load temperature.
Incidentally, these thermal resistance measures I've mentioned seem like small differences between small numbers, but consider:
delta-TR = delta-C / W
If W is your measured thermal wattage at load, it could be a number in the range from 120 to 150W given your OC setting -- another ballpark guess. Suppose it's 100W. Then the drop in temperatures you'd get from replacing the XP-90 with even the XP-90Cu would be:
(0.135 - 0.11)C/W = delta-C / 100W
and delta-C = .025 x 100 or nearly 3C degrees. If the thermal wattage is higher than 100, then delta-C would be proportionately greater. If you've increased the thermal wattage of your E4400 to 150W, then the temperature improvement for getting an XP-90C to replace the XP-90 is closer to 4C degrees. And if you can improve on the cooler's performance -- effectively reducing thermal-resistance again by improved airflow at a desirable noise-level, it would be even better. And it's possible that the 50% OC has actually boosted your thermal wattage higher than I'm showing here by way of example.
You can add another few degrees to the temperature improvement by ducting, and maybe 2C degrees to that by using a diamond-based thermal paste if you're not already doing so. In combination, a more recent and improved cooler-design, the choice of TIM and ducting for better airflow could gain you in the neighborhood of a 10C improvement.
Facebook
Twitter