Best Available WB

[BillA]

ah, positive first
please - the 'not's are endless

[nikhsub1]

Originally Posted by Cathar

I think item #1 would be:

1) Thermal probes placed at the edge of the CPU die, and especially at the edge of an IHS, do not measure anything useful.

Oh, it get's WORSE

[Cathar]

Originally Posted by nikhsub1

Oh, it get's WORSE

Well, while sticking a thermal probe where they did does have its uses, I believe the biggest mistake in that review is the statement:

"actual temperature", which get repeated again and again.

Perhaps if these guys understood that the TIM layer != 0 resistance, and that ~1mm of aluminium also != 0 resistance, then they might be on to something.

If anything, it more or less tells me that Abit is somewhat close to the mark (as I suspected), and the Asus boards are, as I also suspected, reading more like 20-25C below what would be expected.

[pHaestus]

thermal resistance of an unknown amount of "silicone glue" between probe and heatsink? I am guessing that one IS 0. It's just like homemade cookies have no label and thus no calories.

[Etacovda]

http://www.bleedinedge.com/reviews/a..._temps_04.html

The monitors on the left hand side; the numbers change by 8 degrees, they up the fsb to 250, and the vcore to 1.6, and theres no change? righhhtttt.....

[jaydee]

Ha.

#1) An explanation of WHY better testing equipment and methodology is needed.


First they need to understand why they need to do better testing before you try and get them to do it. I can remember at least 2 threads started on other forums dedicated to slagging my ass for hammering them here about their lack of testing knowledge, methodology and equipment.

They usually get heir fanboy readers all huffed up about how stupid I am to think more than a mobo's CPU sensor (uncalibrated of course) is adequate for the job..... These people first need to understand and accept better methods and equipment is needed. Then inform them how to use it to obtain decent results.

My opinion anyway...

[bobkoure]

Originally Posted by Cathar

I think item #1 would be:

1) Thermal probes placed at the edge of the CPU die, and especially at the edge of an IHS, do not measure anything useful.

I think I'd like to generalize that to
1) Just because something shows a reading in degrees doesn't mean it's actually measuring a temperature you care about.
1a) socket thermistor
1b) CPU diode
1c) improperly placed thermal probes
... 1x) or whatever

And try to keep the top level as joe-sixpack-approachable as possible - just my take on things, of course.

I'm also wondering about the best way to do online collaboration. Is there something built in to wiki? Should we just start another thread (or keep going in this one)? Might work if we give one person the role of secretary (I'm willing, but one of the mods here might want the additional control that role might give). We then all spout as many things as we can think of as belong in the doc (not the actual text, just outline points). Secretary tries to organize, collate, etc., into a full outline (or outlines, given the multi-layer approach I hope we'll take). We have a good deal of discussion as to what order things belong in, things forgotten, things inappropriate, at which point we have an outline (or outlines) we agree on. It's then pretty easy to say "I'm working on section 2" or even "12b" and not step on each other. Secretary gets to put text into the outline (or does it make more sense to have mini-docs linked to the outline?).
If I end up as secretary, I'll want to collaborate with one of the mods here to make sure we produce something they'll want to publish.

Or is all this functionality already in wiki? If so, producing this first doc might be a good learning experience, so we have some knowledge in the community about what works (and what doesn't, of course).

Ph - are you planning to actually use wiki - or just end up with something that looks like it? (wondering if I should spend time learning how to create/edit articles in the Wikipedia

[pHaestus]

was planning on getting Joe to hack something up onsite.

Originally Posted by bobkoure

I think I'd like to generalize that to
1) Just because something shows a reading in degrees doesn't mean it's actually measuring a temperature you care about.
1a) socket thermistor
1b) CPU diode
1c) improperly placed thermal probes
... 1x) or whatever

And try to keep the top level as joe-sixpack-approachable as possible - just my take on things, of course.

I'm also wondering about the best way to do online collaboration. Is there something built in to wiki? Should we just start another thread (or keep going in this one)? Might work if we give one person the role of secretary (I'm willing, but one of the mods here might want the additional control that role might give). We then all spout as many things as we can think of as belong in the doc (not the actual text, just outline points). Secretary tries to organize, collate, etc., into a full outline (or outlines, given the multi-layer approach I hope we'll take). We have a good deal of discussion as to what order things belong in, things forgotten, things inappropriate, at which point we have an outline (or outlines) we agree on. It's then pretty easy to say "I'm working on section 2" or even "12b" and not step on each other. Secretary gets to put text into the outline (or does it make more sense to have mini-docs linked to the outline?).
If I end up as secretary, I'll want to collaborate with one of the mods here to make sure we produce something they'll want to publish.

Or is all this functionality already in wiki? If so, producing this first doc might be a good learning experience, so we have some knowledge in the community about what works (and what doesn't, of course).

Ph - are you planning to actually use wiki - or just end up with something that looks like it? (wondering if I should spend time learning how to create/edit articles in the Wikipedia

[bobkoure]

Originally Posted by pHaestus

was planning on getting Joe to hack something up onsite.

So... just use a thread in this (or some other) forum to collaborate?

[BillA]

how does it go . . a little knowledge
what would his reviewer's toolkit need ?

a technical education for sure, 1st yr physics would help a lot

re: http://www.bleedinedge.com/reviews/a..._temps_04.html

[Kobuchi]

Originally Posted by pHaestus

was planning on getting Joe to hack something up onsite.

If this company did the links pages script, why not use this product? It dovetails with their forum software too. So Joe needn't reinvent the wheel.

[bobkoure]

Originally Posted by unregistered

http://www.bleedinedge.com/reviews/a..._temps_04.html

I would certainly agree with this reviewer that reported CPU temps are probably wrong. Beyond that - I dunno.
He could have dropped the entire heat sink with embedded probe into a beaker of water with a decent thermometer to see how close his reported probe temps were to something reliable - failing that, the use of the phrase "actual temp" makes me cringe. Calling it a "digital probe" makes me cringe as well - but not as much
On a related note, isn't silicone sealant an insulator? Maybe it just plugged the end of the hole. Maybe better to use AS epoxy? Maybe the probe was a press fit. Another "I dunno...".
I don't know what the C/W of AS5 plus some thickness of aluminum, plus a possible gap between alu and probe (which might have silicone seal in it) - but I know it's something - and I would expect the temp at the probe to be lower than at the CPU. Or is the C/W so low as to not matter?
Just wondering....

Bob

[Genocide]

even if he didnt fill the hole with sillicon, air isent much better and even a tube of as5 wouldnt do it justice

as i see it he basically made the situation worse