Cooler reviews - Recommendations :)

[tkpenalty]

So yeah I've been noticing a lot of the cooler reviews are probably compared with older reviews from different days. This results sometimes in unfair results that come from the review. I'm not saying our reviews are bad, but we could improve.

Taking an idea from frosytech, why not have reviews have a rise over ambient temperature, instead of just the temperature itself.

This leads to my second point, are our reviewers using the onboard temp monitors or a temp probe? Its a good addition to the system, and certainly a great thing to use in reviews as they are inexpensive and accurate.

Thanks to the TPU staff for continually providing already very comprehensive reviews! With this the reviews will be even more accurate, and comparable on one single chart.

EDIT: Point 2, fixing it up a bit, if we are using the software to monitor temperatures in reviews, it isnt as accurate in contrast.

[tkpenalty]

*bump*

[beyond_amusia]

Point 2... I assume they just use an app like speed Fan to see the temps.

[Kursah]

It would be a cool idea TK, I'd like to see something like that implemented for Cooler reviews and maybe an overall chart for all reviewed coolers too. Good stuff man!

:toast:

[tkpenalty]

Point 2... I assume they just use an app like speed Fan to see the temps.

Inaccurate. As I stated, onboard sensors are very inaccurate due to manufacturing quality and the way that the motherboard manufacturer uses them.

Heres a fact:

My E6300 runs at 34*C with the ITE temperatures on my main PC's P31 DS3L. Now, throw it over to my second PC's P31 DS3L and the temps drop 10*C.

Core temp also has the same problem, except now its on the CPUs. The core temperatures are by no means accurate, they MAY be but not anymore thanks to manufacturing quality differences and the fact that intel's using different parts in different batches. My friend's E6320 runs at 33/32 for core temp, while my E6300 runs hotter-and he was on stock cooling!

See what I mean? Using the temperature probes is FAR more accurate, on the LGA775 CPUs, you generally sandwich the probe onto the black part, where the IHS drops down a bit. On AM2, you usually would sandwich it on the side of the IHS as well.

I'll probably try to get a temperature probe soon, might want to review something :)

[SK-1]

I'll probably try to get a temperature probe soon, might want to review something :)


D:toast:o it.

[DanTheBanjoman]

Inaccurate. As I stated, onboard sensors are very inaccurate due to manufacturing quality and the way that the motherboard manufacturer uses them.

Heres a fact:

My E6300 runs at 34*C with the ITE temperatures on my main PC's P31 DS3L. Now, throw it over to my second PC's P31 DS3L and the temps drop 10*C.

Core temp also has the same problem, except now its on the CPUs. The core temperatures are by no means accurate, they MAY be but not anymore thanks to manufacturing quality differences and the fact that intel's using different parts in different batches. My friend's E6320 runs at 33/32 for core temp, while my E6300 runs hotter-and he was on stock cooling!

See what I mean? Using the temperature probes is FAR more accurate, on the LGA775 CPUs, you generally sandwich the probe onto the black part, where the IHS drops down a bit. On AM2, you usually would sandwich it on the side of the IHS as well.

I'll probably try to get a temperature probe soon, might want to review something :)

Not really relevant when you're testing with the same sensor. Even if it is 5% off you still have the same difference between coolers.

[tkpenalty]

Not really relevant when you're testing with the same sensor. Even if it is 5% off you still have the same difference between coolers.

Explain to me why i get different temps on the E6300 on different boards? Remember that all our reviewers dont have the same hardware.

I meant, all the review results put into one for every review for a very comprehensive comparison-certainly something that will make us a very reliable and reknowned review site :)

[Jelle Mees]

#1 source for best cooling:
http://www.frostytech.com/top5heatsinks.cfm

[twicksisted]

Explain to me why i get different temps on the E6300 on different boards? Remember that all our reviewers dont have the same hardware.

I meant, all the review results put into one for every review for a very comprehensive comparison-certainly something that will make us a very reliable and reknowned review site :)

This may not be relevant in your case as to why you get different readings with the same chip, but I have found that each time I re-install my cooler & thermal paste I will get different temps.

For instance I made 2 rigs with the same asrock boards in... both with E6600 chips and retail cooling... about a 10c difference as I applied too much thermal paste on the one... after taking and swapping chips around and re-applying paste it was within 2c difference.

Mostly nowdays its pretty close as I have found the best way of applying thermal paste that works for me with the lowest temps.... still, different boards, the 775 dodgy clipping mechanism, thermal paste & the fact that no 2 cpu's are the same (concave convex IHS etc..)... all these factors play into why the temps will bbe different

[tkpenalty]

This may not be relevant in your case as to why you get different readings with the same chip, but I have found that each time I re-install my cooler & thermal paste I will get different temps.

For instance I made 2 rigs with the same asrock boards in... both with E6600 chips and retail cooling... about a 10c difference as I applied too much thermal paste on the one... after taking and swapping chips around and re-applying paste it was within 2c difference.

Mostly nowdays its pretty close as I have found the best way of applying thermal paste that works for me with the lowest temps.... still, different boards, the 775 dodgy clipping mechanism, thermal paste & the fact that no 2 cpu's are the same (concave convex IHS etc..)... all these factors play into why the temps will bbe different

Yes I did take that to consideration, but however a 10*C difference in temperatures? I mounted the coolers several time to see, but the same thing happened.

[marsey99]

have you got your multi meter out to see if they are both delivering the same volts to the cpu? its not only intel/amd who can let qa slip.

i do agree with you that a delta from ambeint should be included too tho.

[W1zzard]

would be nice if we could find someone to do more cpu cooler reviews.

requirements:
- tech skills & knowledge
- testing platform s775 + am2 (main pc as testing rig = bad idea)
- ability to take good photos
- english skills
- enough time to review several coolers a month

[WarEagleAU]

Ooer, I may have to volunteer for that then :D

[Pinchy]

would be nice if we could find someone to do more cpu cooler reviews.

requirements:
- tech skills & knowledge
- testing platform s775 + am2 (main pc as testing rig = bad idea)
- ability to take good photos
- english skills
- enough time to review several coolers a month

I wouldnt mind doing reviews and such in my spare time, but I would like to know one thing (which is what has held me back from doing any reviews).

(This is based on the cooler reviews but can be applied to any reviews. eg, CPUs/GPUs). How do we compare different heatsinks? i.e, do we need to start off with (say) ~10 heatsinks/coolers to compare or is there a different method?

[WarEagleAU]

I was thinking the same thing pinchy. I got two spare PCs here (my old opty) and an AMD X2 3800. I got an awesome camera. So Id love to help out if needed. Let us know Mr W1zzard!!! :D :D :D

[Nitro-Max]

Seems to me that its the heatsinks with the biggest surface areas that seem to get the best results like the thermaltake ultra 120 its huge man!.

[Pinchy]

I was thinking the same thing pinchy. I got two spare PCs here (my old opty) and an AMD X2 3800. I got an awesome camera. So Id love to help out if needed. Let us know Mr W1zzard!!! :D :D :D

Yeah I wouldnt mind. I got my secondary core 2 duo rig to test anything on.

Another problem I may face is that im in Australia. Might cause problems sending the cooler to me and me sending it back.

[W1zzard]

coolers usually dont go back

yes you have to build your own set of comparison numbers over time. i doubt we can get several reviewers together that would get exactly the same results for a pair of heatsinks on different rigs

[Pinchy]

I already have three coolers if that counts :D (LOL including the intel stock cooler :p)