pH doing waterblock/HSF testing again? - Poll

[Joe]

Since pH has pretty much lost any love with doing testing, and figures no matter what he does with it, he will be torn apart. Because of the recent "flip flop" if I can use that term... of testing methodology thats been thrown around in the last couple months... I think its fair to say this whole area is in flux.

So I figure I would put up a poll to see what people think. Should pH start testing blocks again? get his rig back up and running? And if so what kind of testing would you be satisfied with?

[pHaestus]

This poll is flawed

If everyone votes for testing with IHS-clad AMD64 it doesn't affect me. I won't do it cause I will quickly destroy the CPU

And there's no poll for "pH should sit on his ass and drink beer"

[Joe]

I should have noted that the IHS Clad option was there just for info, because I know you wouldnt do it. If people pick that, you can just tie it to "sit on your ass and drink beer" I guess

[pHaestus]

OK I vote that one then

[flatline]

voted socket A as thats what i use

[bobo5195]

picked bare core, but would be tempted am equally tempted by beer to give the poor guy a break, since knowing us we will only argue anyway.

What about if he cooled beer?

[j813]

Test =
"Big Player
Making Big Money"

[jaydee]

Went with the first one. How is the dP thing going?

[BillA]

a sad day when I an unable to answer such a question
my views have created such confusion and dismay that I will pass on a vote/position

fix the dP setup, test in any manner
the results will still have to be correlated to something (no temp is "real")
change source, start anew

[bigben2k]

No clear answer here either.

I support the use of a processor, just because of the (painfully arranged) diode reader (and how the data can be used by other testers).

Not entirely sure about the consistency of it all though; I'd like to see one single block (calibration block?) tested on a regular basis. I'm not invalidating the procedure, I'm just suggesting that since this is (relatively) cutting edge (i.e. unique), some kind of check might be in order.

[snowwie]

my confusion is the diff between measuring tempuratures from the cpu or from a TC. I always thought that the accuracy of the on-die diodes was unknown or bad, and a TC or RTD or whatever should be used to make the measurements and thus the test "scientific".

but then again everyone loved PH's testing when he measured from the athlon xp's on-die diode. of course as billa said:

the results will still have to be correlated to something

I suppose THAT was the factor that has set PH's testing apart from the other review sites and even joe nowadays.

[BillA]

pH did independently cal his diode
the interesting question is what that temp is wrt the allegorical 'die temp'
yes it is a die temp, is it THE die temp ? (all are NOT the same)

[stev]

Originally Posted by BillA

pH did independently cal his diode the interesting question IS what that temp is wrt the allegorical 'die temp' yes it IS a die temp, IS it THE die temp ? (all are NOT the same)

Define the word "IS" --- Bill Clinton.

(I just couldn't pass up the pun in that one.)

[BillA]

let me get to my desk

ahh, being is

[pHaestus]

Um Ben

In the SocketA days I always had 2 "std" waterblocks with which I tested. Any time I was forced to move things around I would (a) recal the diode reader/CPU setup (in case some slight change in the solder joints) and (b) run 2 separate wbs through a dT vs. Q testing (5 day process). THEN I could feel confident that I was getting numbers that were always comparable to previous work and so the interactive graphing thing was as free of error as was humanly possible.

[bigben2k]

Oh good! Thanks for sharing that; it's hard to consolidate everything that you've posted already.

[Joe]

I would note that the overwhelming majority either want pH to just test, or test with an IHSless CPU.

Lets get cracking!