Need tips for HSF testing for reviews

[sieb]

I currently have one of the new Swiftech x370s on its way, along with a Millennium Glaciator. Im hoping to do reviews on them considering Im probably one of the few having fairly annoying cooling problems. I still cant run my full speed of 1.4Ghz due to heat. But I wanted some tips (without getting too technical) on the best places to record temps.

I currently have a new Antec SX1030, 400W PSU, (5)80mm fans, and my digi doc5 monitoring 3 thermisistors, One touching the die (tab trimmed off), one wedged in between the HS fins, and another floating about monitoring in-case temps, (as well as the mobo thermisistor behind the proc). I was going to use this same setup to test my new HSs when i get them in along with ASII.

I wanted to know what some other good places or ways to monitor temps? Then I was going to post the results here in a review for everybody. What ya all think?

 

[klosters64a]

I suggest looking up past HSF reviews on OCer's.com's front page. Reading a few of them should reveal the right way to measure CPU temp.

 

[William]

reading our reviews is probably a good idea, ataching an external probe to the side of your core is probably the best way to get an accurate reading of your temps. I recomend using the AS Epoxy to attach it!

 

[Kingslayer]

There is one thing that I will recommend if you are going to test heatsinks. DONT OVERCLOCK. If I find ANY review in which a piece of hardware or software was tested on an overclocked machine I will instantly skip that review.

Overclocking can and will produce inconsistant numbers. Especially in AMD chips. AMD's can hit the same speed with different multipliers and FSB speeds and cause different temperatures at the same speed.

Use a baseline HSF (stock HSF is usually good) to show your review. This will give people an idea on how to judge a HSF against a stock one.

Ambient temperature. Sometimes it can be a pain to control, but keeping the ambient temperature the same for all HSF reviews is a must. If I see a test with wavering ambient tests, I will skip it also.

Case configuration. Build a test case. And NEVER change it. Don't change fans (unless they burn out) don't change cards, dont change wiring, don't change anything that will alter the flow of air in the case.

HSF's. Most come with TIM's. What you do with it is up to you. But, if you are going to test it with the TIM, then RadioShack thermal paste, then AS2, make sure you thoroughly clean the base of the HSF. Also remove any dust that might have accumulated on the HS or the fan.

I worked for a testing facility for the Marine Corps my last 3 years. I know how to test things. And the more that you can control the test environment the more reliable your test will be. And the more reliable your data will be. And that's basically what you are doing in a review.

Also, if you are doing this for a hardware site, invest in a good multimeter that will read temperatures as a comparison. And seeing the Asus boards are known for inconsistant temp readings I would rule it out. If you monitor your temps with software, use the same version every time.

All this will read to good consistant data, and a review that's worth looking at.

 

[Hoot]

If you do anything human in your review, I threaten to skip it. Then you'll be sorry.

Hoot