Can someone with some real knowlege answer this question???

[BillA]

the 2 preceding posts have links to the answers to all those questions

[MadDogMe]

All ten of them or one in particular? ...

Anyone else have a short answer? ...

[gone_fishin]

Originally posted by MadDogMe
BillA!, if AS is such a big variable for testing , why use it?. The block and CPU would get along fine without it would'nt they?(albeit with a DegC or three rise ), but a Cu to silicon(sp?) join would'nt vary at all would it?, apart from the pressure...

Has anyone tested just the block on CPU with constant temp water before?(no rad), I'd be interested to know how far above ambient it holds the CPU(on one of the 'great5' blocks ), it'd give an idea of the % of the frailtys of the rest of the system would'nt it?...

What kinda temps do people get with H2o nowadays?, above their ambient I mean?, I average 10DegC with a poorly tuned setup. What is a good No?, 7c?, 5c?... does'nt leave a lot left in the block when you take the rad's inefficiency into count...

When you ask "how far above ambient" what do you mean? At what flow rate and heat output? If you mean a die temp of 10C above room temp with an average overclocked AMD processor(100watts guestimate), even at higher than average flow rates, then I think something is wrong with your temp measuring procedure. A review of posted testers temps would reveal a higher delta than that. A lot of links were provided here already to sift through

Here's some recordings by my primitive testing of a modified Spiral with a true 1/2" ID inlet/outlet.

[MadDogMe]

Has anyone tested just the block on CPU with constant temp water before?(no rad), I'd be interested to know how far above ambient it holds the CPU(on one of the 'great5' blocks ), it'd give an idea of the % of the frailtys of the rest of the system would'nt it?...

The temp of the water(constant held temperature)...

What kinda temps do people get with H2o nowadays?, above their ambient I mean?, I average 10DegC with a poorly tuned setup. What is a good No?, 7c?, 5c?... does'nt leave a lot left in the block when you take the rad's inefficiency into count...

I meant their room temp for this one, my room temp is 20DegC and I idle/browse @ 30DegC(34~5 load) at the moment, these are just diode temps, but mine are'nt what I'm interested in at the mo', it's what a good temp is, I've heard people claim 7degC above such and such(20~22:shrug: ), I know it's not accurate but it's a ballpark figure I'm intersted in...

PS.My mistake!, I should'nt have said 10DegC was my average temp, I should have said average idle ...

PPS. I'll have to sift through the links when I have more time, I was hoping there'd be an average 'best' for an average (100watts is right) temp. basicly where the wall is with the big boys in a room temp that not at extremes(not below 20DegC?). I hav'nt learnt to equate Delta yet, how to make it into a 'real world' figure in my mind. so I rely on CPU temp above X ambient . I'll get there though, when I have time! ...

[BillA]

g_f
always good to see more people testing
any chance of a more legible image ?
(I am not afraid of a link)

[gone_fishin]

Originally posted by unregistered
g_f
always good to see more people testing
any chance of a more legible image ?
(I am not afraid of a link)

Host troubles at the moment. Here's a temporary link, don't expect it to last long though.
h**p://www.theforumisdown.com/uploadfiles/0802/Modified_spiral1.jpg

[jaydee]

Originally posted by gone_fishin
Host troubles at the moment. Here's a temporary link, don't expect it to last long though.
h**p://www.theforumisdown.com/uploadfiles/0802/Modified_spiral1.jpg

My host problems seem to be over for now so here is the pic for people that are afraid olf links.

[BillA]

define series 1, 2, & 3; is that the vcore ? (C, D, & E)
so the load temps dropped at 3/E ???
look at the air/water differential - ?? (should be increasing as the load increases)
2.2, 4.2, 3.0 ????

at the same flow rate, with the same die size, the C/W will be the same at all power levels
(hundreds of tests this way; if your's are not, there is an error)


what is "Ambient load Temp" ? = air temp under load ?
Water the same ?

you have too many uncontrolled variables
-> unless you REALLY like multivariant analysis (which I do NOT)
and you'll need a LOT of data points

[Les]

Originally posted by unregistered

at the same flow rate, with the same die size, the C/W will be the same at all power levels
(hundreds of tests this way; if your's are not, there is an error)


-> unless you REALLY like multivariant analysis (which I do NOT)
and you'll need a LOT of data points

Testing using a CPU die may always be a multi-variant situation. The enviroment(e.g motherboard) temp increasing with v-core and/or MHz..
Possibly illustrated by :


g-f. Ta for the impingement cooling link in the other thread. Some useful data.

[gone_fishin]

Originally posted by unregistered
define series 1, 2, & 3; is that the vcore ? (C, D, & E)
so the load temps dropped at 3/E ???
look at the air/water differential - ?? (should be increasing as the load increases)
2.2, 4.2, 3.0 ????

at the same flow rate, with the same die size, the C/W will be the same at all power levels
(hundreds of tests this way; if your's are not, there is an error)


what is "Ambient load Temp" ? = air temp under load ?
Water the same ?

you have too many uncontrolled variables
-> unless you REALLY like multivariant analysis (which I do NOT)
and you'll need a LOT of data points

A quick glance up at the numbers represented by each line in the chart will show you.
But here goes

There were three tests taken in all, variables being room temps noted at each test and each test at a different heat output. The three resulting C/W's are added together and averaged for the resultant (C) for water and (C) for air, hence the two C/W's.

Series 1 is the numbers taken from the chart for "ambient temp" which is the air temp measured at the rad(the at the rad part I failed to make a note of in the chart)
Series 2 is is the water temp at load
Series 3 is the P4 cpu die temp reading

There are 3 numbers for each series, one taken at a different vcore/mhz as noted in the chart.

.168, .174, .224

I got these readings (computernerd watts)as you can see,
@ stock = .224C/W
@ 81w = .174
@ 101w = .168

They are then averaged to .189C/W

I am well aware that heatload does not change C/W but these are the repeatable readings that I get, a tendency towards higher C/W at lower wattage. Perhaps the wattage used in the equation is buggered? But you can see from my chart the conclusions people will come to by doing their own testing. My water temp sensor has a resolution of .1C and as you know the P4 die reading is a .5 resolution.

I included the three most commonly quoted wattage figures that people have seen in the bogus reviews across the web. I am not claiming that what my chart says is the real C/W of the block, just what I am able to measure on my system with the thermometer that I have.
The room temps were different when each test was taken, as noted. The middle load air to water differential is greater (and room temp was also greatest) and it does bother me, but that is what I recorded. This is only one seating and the purpose of my posting it was to the question at hand, does the average user see a cpu temp of 10C above room temp when watercooling? I propose not, even though my test is not valid when compared to yours BillA I do not believe I am off by 10C

[gone_fishin]

Les, your welcome. The thread didn't go very far though, wonder why I even hesitated to post? But I'm glad it was useful to someone

[BillA]

"does the average user see a cpu temp of 10C above room temp when watercooling?"

from that sentence I will select one word . . . see
and we both understand that what they will SEE has no necessary relation to what actually IS
-> with the "pHaestus Thermister Calibration Mod" in fact they can SEE whatever they wish

the nerd term is GIGO, no ?

my error, I was presuming to talk of real numbers - sorry 'bout that

in answer to the above question you said "I propose not"
my response:
- in terms of what they 'see', who knows ? (your post is an example, but disproves your own answer ??)
- factually, you are wrong (at any 'typical' power and flow rate)
period (there is a chart previously alluded to if you want facts)
for the link challenged:

at 40W and above there will certainally be a 10°C differential

Les
this is the whole problem with software/CPU "Watts"
some guy with a pencil stated what he 'thought' (NOT data from Intel or AMD)
- there is not a single iota of factual test data to substantiate ANY of those 'values' (sic)
- indeed secondary losses are not a constant related only to the applied power, but are also affected by the various temp differentials

[gone_fishin]

Originally posted by unregistered
[B in answer to the above question you said "I propose not"
my response:
- in terms of what they 'see', who knows ? (your post is an example, but disproves your own answer ??)
- factually, you are wrong (at any 'typical' power and flow rate)
period [/b]

I was obviously implying that it would be more than 10C when I said "I propose not" and a reasonably minded viewer should get that idea even though I did not expound on said proposal, because the chart I submitted shows more than 10C. Was my point really that unclear?
I was originally responding to this question,

Originally posted by MadDogMe

What kinda temps do people get with H2o nowadays?, above their ambient I mean?, I average 10DegC with a poorly tuned setup. What is a good No?, 7c?, 5c?... does'nt leave a lot left in the block when you take the rad's inefficiency into count...

"Seeing" is definately done with blinders at times, whether on purpose or through ignorance and your point is well taken.

[BillA]

well, I would say the 'seeing' problem is mine

add 3 or 4 for the coolant/air differential and the CPU would have to be below 20W to even be close to a real 10°C differential

(and yes, I did not understand your 'intent')

[MadDogMe]

I did'nt realise how much the temps increased in ratio with ambient increase!, I've only got my rig to go by, and a poorly setup creature it is too. I see 10DegC above(20DegC room) when browsing and 34.5 when load*, so that's what I went by, I forget the 28C room temps some SoCal'ers have (lucky sods!) I could live with 45+ temps if a year~round tan went with it! . I'm gonna have to learn to read Delta~C/W ...

*what's the best proggie to stress with?, I have BurninTest that someone gave me, I usualy run that with Sandra and HW monitor for temps.

BillA, what would you recomend for the average Joe Deer to read their temps by?, drilling a diode into the block BP above/near the core?, or touching core edge between block and socket?. then what about readers?, do DD5's or compunurse type readers have a resolution higher than 0.5?. are they accurate enough(not for testing of course). what's the best affordable setup?(<£100). PS. you sig should be changed to 'AirBourne~BillA' on account of going in feet~first ...