TaiSol's new CGK742092 heatsink/fan just arrived today. Previously, I was using the TaiSol 733092 with excellent results on a 1Ghz Thunderbird.
With the cheaper 733092 heatsink/fan my idle temps at an overclocked speed of 1.1Ghz were high 20's Celcius with peak load temps in the high 30's. Rarely did I record a temperature in the 40's. But I was curious about the new forged heatsink/fan so I ordered one to play with. (I'm not rich, just curious)
Upon inspection I immediately noticed the bottom surface of the 742092 heatsink was very rough compared to the 733092. A serious lapping job was in order, but first I wanted to see how it compared out of the box.
I removed the old TaiSol heatsink/fan, popped off the CPU for cleaning, installed the new 742092 complete with TIM and booted up the machine.
I performed the identical burn in procedure with the new cooler as I did with the previous one. Using Via's hardware monitor I loaded up a session of Prime95 and CPU Heat, unplugged all case fans including the CPU fan, and proceded to 'bake' the TIM pad on the heatsink for maximum phase change. Upon reaching 55C, all fans were quickly reconnected and the computer was shut off. Remember, this process previously produced excellent results with the older TaiSol cooler, as my idle and load temps were very low as a result.
Unfortunately, after booting with my new 742092 and checking Via for temps after the burn in, I was puzzled to see that temps were actually much higher than the cheaper TaiSol cooler. I was idling at 30C right off the bat. Doh! not good. I then loaded up a few programs, prime95, CPU heat, Seti, and 3DMark 2000 looping demo's. I waited a while and monitored the temps. I was in horror to see temps climb into the low 50's. Yikes! This has never happened before. Heck, previously temps in the 40's were rare - what gives?
Off came the heatsink/fan, both heatsink and CPU were cleaned of all traces of thermal pad, a new thermal pad was installed (glad i purchased extra) and the process repeated itself. The results of the second round were almost identical. Off came the 742092 and back went the trusty old $15 TaiSol 733092. Temps were now stable and things were running cool like they before. idle temps in the 20's, peak load temps in the high 30's.
Before anyone says "heh, ditch the thermal pad and grab some Artic Silver!" Remember, the thermal pad (TIM) was producing amazing results for me with the older TaiSol. The thing is, I was expecting the newer expensive TaiSol 742092 forged cooler to produce at least the same or preferrably better. I was disappointed to find out that results were much worse than the cheaper cooler. I'm almost convinced the problem lies in the uneven surface of the 742092. It need to be lapped and that's what I am going to do next.
I'll post my results after I've properly lapped the new cooler. But for now, I just want to inform people that this cooler may require some lapping out of the box for maximum effeciency.
I'll keep you posted.
With the cheaper 733092 heatsink/fan my idle temps at an overclocked speed of 1.1Ghz were high 20's Celcius with peak load temps in the high 30's. Rarely did I record a temperature in the 40's. But I was curious about the new forged heatsink/fan so I ordered one to play with. (I'm not rich, just curious)
Upon inspection I immediately noticed the bottom surface of the 742092 heatsink was very rough compared to the 733092. A serious lapping job was in order, but first I wanted to see how it compared out of the box.
I removed the old TaiSol heatsink/fan, popped off the CPU for cleaning, installed the new 742092 complete with TIM and booted up the machine.
I performed the identical burn in procedure with the new cooler as I did with the previous one. Using Via's hardware monitor I loaded up a session of Prime95 and CPU Heat, unplugged all case fans including the CPU fan, and proceded to 'bake' the TIM pad on the heatsink for maximum phase change. Upon reaching 55C, all fans were quickly reconnected and the computer was shut off. Remember, this process previously produced excellent results with the older TaiSol cooler, as my idle and load temps were very low as a result.
Unfortunately, after booting with my new 742092 and checking Via for temps after the burn in, I was puzzled to see that temps were actually much higher than the cheaper TaiSol cooler. I was idling at 30C right off the bat. Doh! not good. I then loaded up a few programs, prime95, CPU heat, Seti, and 3DMark 2000 looping demo's. I waited a while and monitored the temps. I was in horror to see temps climb into the low 50's. Yikes! This has never happened before. Heck, previously temps in the 40's were rare - what gives?
Off came the heatsink/fan, both heatsink and CPU were cleaned of all traces of thermal pad, a new thermal pad was installed (glad i purchased extra) and the process repeated itself. The results of the second round were almost identical.
Off came the 742092 and back went the trusty old $15 TaiSol 733092. Temps were now stable and things were running cool like they before. idle temps in the 20's, peak load temps in the high 30's. Before anyone says "heh, ditch the thermal pad and grab some Artic Silver!" Remember, the thermal pad (TIM) was producing amazing results for me with the older TaiSol. The thing is, I was expecting the newer expensive TaiSol 742092 forged cooler to produce at least the same or preferrably better. I was disappointed to find out that results were much worse than the cheaper cooler. I'm almost convinced the problem lies in the uneven surface of the 742092. It need to be lapped and that's what I am going to do next.
I'll post my results after I've properly lapped the new cooler. But for now, I just want to inform people that this cooler may require some lapping out of the box for maximum effeciency.
I'll keep you posted.
Facebook
Twitter