OCUK Official IC Diamond/ Perihelion Test Results

WingZero30 said:
Posted:)

Let me know again when you have received it



After some thought I decided to conduct the lapped testing in June which is meant to be one of the hottest month in UK. It will be ideal time.

Lapped testing methodology will be similar to unlapped testing :cool:
Click to expand...

received it today perhaps results by the end of the week
 
i7-920 21×185 Megahalems with GT 1150×2
My results:
MX-3
Ambient 22°C
idle 42, 38, 43, 41°C
load 81, 78, 78, 78 °C

IC Diamond
Ambient 22
idle 41, 39, 43, 41°C
load 81, 79, 77, 78 °C
 
Tested at ambient 21°C

Radeon 6950
Stock Paste 42°C Idle 88°C Load
IC Diamond 40°C Idle 82°C Load

AMD Athlon II 620 @ 3.6Ghz 1.475v Scythe Mugen 2
Akasa AK 750 35°C Idle 58°C Load
IC Diamond 32°C Idle 52°C Load
 
Posting the Lapped Contact and Pressure test results here:)



Raw Image

rawimage.png




Pseudo Colour Representation

pseudocolourrepresentat.png




Pressure Statistics

pressurestatistics.png




Pressure Histogram

pressurehistogram.png




Line Scan Profile

linescanprofile.png




3D Image

3dimage.png




Grid Raw Data 10 x 10

gridrawdata10x10.png



I am posting the email you sent me so further discussion can take place

Adeel,

Results were a surprise to me, here is my take on it. The mountain was made into mole hills. One of the properties on the paper is if there is no contact or no impression that area could be .0005 mil or a mile a kind of go - no go situation.

you are more likely from the thermal result to be on the "close to contact" margins. The other alternative was a miss-fire on the mount or test although I think as you have done a couple already it is not likely. Your contact area was down to .38 inches and your pressure dropped to 53.58 psi.

"Removing the mountain" may have been enough material to lighten your mount pressure (53.58 psi is a good + range for any sink mount and ICD) but as your contact was broader and less concentrated contact in areas that were not adding any benefit the contact improvement was where it counts was the dominant factor.

A tightening of the sink would most likely add to the contact profile and add a couple C to your overall performance.

If you have a test left an increased pressure test on the same sink would be interesting.

All The Best

Andrew (IC Diamond)
Click to expand...


I am surprised by lapped results also.

I think I know what you mean by mole hills. 'Mole hills' are those tiny dots which are spread across the centre of contact. Making the contact area broader has definitely resulted in lower pressure as I mentioned in post 390 of this thread.

I think I may know the reason why I get temps drops after lapping even though overall total contact area and pressure have decreased (thanks to 10+ hours of lapping :rolleyes:) and correlates with what you said: "the contact improvement was where it counts was the dominant factor"

contactcomparison.png


The four cores as drawn and as shown in the pic with rough superimposition are positioned in the upper half of Q6600 and that is where the contact area and pressure are greatest.

Below those cores are cache units and that is where we get the least pressure and contact area.:)

Since only cores are the dominant factor in producing heat especially under load, the cache doesn't make difference.

However if the situation was opposite with cores residing at the bottom and cache at the top, then there would have been major problem with temps.

The sink was fully tightened so there was no further room for tightening the screws.

As for increased pressure test, how would it be done?

With the contact surface as it is (which is still not perfect) between IFX-14 and Q6600, increasing pressure would only show itself in those areas where the greatest contact is already i.e top and bottom. There still won't be any difference in the lighter areas or areas where there is no contact i.e centre right

The only option is to produce near perfect lapping, so more protruding metal is shaved away, but I think this is difficult with hand lapping.

Here are my better lapped pictures and include the heatsink aswell

imag0148a.jpg


imag0151gi.jpg




imag0152fy.jpg


imag0154ln.jpg




I am in the process of Lapped testing:)

WingZero30:cool:
 
IC Diamond

I forgot to mention earlier that with regards to 'close to contact' margins and looking at the general contact profile, I also felt that the gap between IFX-14 and Q6600 surfaces is very very small indeed. Even the area where there is hardly much contact, we can still see tiny spots or 'mole hills'. So there is contact there but it is just a 'touch'.

1 mil is 1/1000 of an inch or 0.0254mm.

So 0.0005mil = 0.0005 x 0.0254 = 0.0000127mm

It is astonishing that a very very tiny fraction of a mm can make all the difference:eek:.

Could it be that while lapping, some of the grit particles managed to get stuck into the Cpu IHs or heatsink base, thus preventing any further contact?

What would you suggest?

In my lapping experience the only sandpaper I have found most effective in shaving the metal is 120 grit.
 
Last edited:
WingZero30 said:
IC Diamond

I forgot to mention earlier that with regards to 'close to contact' margins and looking at the general contact profile, I also felt that the gap between IFX-14 and Q6600 surfaces is very very small indeed. Even the area where there is hardly much contact, we can still see tiny spots or 'mole hills'. So there is contact there but it is just a 'touch'.

1 mil is 1/1000 of an inch or 0.0254mm.

So 0.0005mil = 0.0005 x 0.0254 = 0.0000127mm

It is astonishing that a very very tiny fraction of a mm can make all the difference:eek:.

Could it be that while lapping, some of the grit particles managed to get stuck into the Cpu IHs or heatsink base, thus preventing any further contact?

What would you suggest?

In my lapping experience the only sandpaper I have found most effective in shaving the metal is 120 grit.
Click to expand...

I would not worry about the grit.

Key thing to remember - increased contact does not increase pressure, but generally increased pressure will increase contact. Pressure is the more dominant of the two.

Let's flip the problem, In engineering a common device used to limit heat flow to sensitive components in an assembly is a “Heat Dam” shown here in figure #3 Where the heat flow is “thermally choked” by reducing the contact area to material limitations. Restricted heat flows with misaligned contact, too little contact, too little pressure, too little compound anything that can restrict flow creates a heat dam.

Heat flows on a differential from hot to cold but if you are thermally choked at a point in the thermal cascade downstream improvements in thermal conductivity offers little difference in performance and shows up as reduced deltas of only a degree or so between compounds or what I sometimes refer to as homogenized test results.

Now back to your situation, your analysis of having contact in the right area is correct rather than forcing the heat flow across the IHS to the sink which is a thermally choked condition pushing to the material limit of the copper. I believe the molehill change from mountain allows a thermal compound to work as designed now by being able to bridge the gap from IHS to sink as in Figure 1. These are the thresholds or tipping points for you guys and all your ducks need to lined up in detail to squeeze the most out of your systems.

For further improvement possibly a degree or two from improved contact and another 1-2 from increased pressure although I lean towards more pressure at this point.

If you could figure a shim to tighten down the sink some more...


heat_dam.png
 
Last edited:
IC Diamond said:
I would not worry about the grit.

Key thing to remember - increased contact does not increase pressure, but generally increased pressure will increase contact. Pressure is the more dominant of the two.

Let's flip the problem, In engineering a common device used to limit heat flow to sensitive components in an assembly is a “Heat Dam” shown here in figure #3 Where the heat flow is “thermally choked” by reducing the contact area to material limitations. Restricted heat flows with misaligned contact, too little contact, too little pressure, too little compound anything that can restrict flow creates a heat dam.

Heat flows on a differential from hot to cold but if you are thermally choked at a point in the thermal cascade downstream improvements in thermal conductivity offers little difference in performance and shows up as reduced deltas of only a degree or so between compounds or what I sometimes refer to as homogenized test results.

Now back to your situation, your analysis of having contact in the right area is correct rather than forcing the heat flow across the IHS to the sink which is a thermally choked condition pushing to the material limit of the copper. I believe the molehill change from mountain allows a thermal compound to work as designed now by being able to bridge the gap from IHS to sink as in Figure 1. These are the thresholds or tipping points for you guys and all your ducks need to lined up in detail to squeeze the most out of your systems.

For further improvement possibly a degree or two from improved contact and another 1-2 from increased pressure although I lean towards more pressure at this point.

If you could figure a shim to tighten down the sink some more...
Click to expand...


Nice insight :)

I agree and I did state in an earlier post that with unlapped setup, heat accumulated faster at the Cpu IHS/ IFX-14 base than it could be trasnferred via the IFX-14 heatpipes resulting in higher coretemps overall for all the thermal compounds I tested.
Now I know that the technical word is heat dam which was greater in unlapped setup.

However with lapping, as the contact area has become broader and especially the more pressure and contact area is located in the right place where cores reside, not only heat transfers faster via heatpipes, but being so 'close to contact' margins, the thermal compounds are now carrying out their main function more efficiently i.e. filling in tiny gaps between the mating surfaces.

However with regards to shim or should it be called CPU Shim, would it really force the mating surfaces to come even closer?

Edit: Would washers do the job aswell if I placed one or two between the mounting plate and the top surface of heatsink base as showed in post 291:

applicationofifx14mount.png

Application of Mounting System
 
Last edited:
WingZero30 said:
Nice insight :)



Edit: Would washers do the job aswell if I placed one or two between the mounting plate and the top surface of heatsink base as showed in post 291:

applicationofifx14mount.png

Application of Mounting System
Click to expand...

I believe you are onto it. Just a matter of taking up the slack realized from the base material loss and a shim under the mounting plate center would do it, perhaps a coin or a few layers of duct tape
 
OK so as things stand at the moment:):

I am currently in lapped testing process with the current lapped set up and will repeat the same length time-wise procedure which I carried out for unlapping as this is my official pure raw lapped testing:p.

However with the final washer modded lapped testing, I won't spend that long. I will apply each thermal compound for about 2 days, heat it up with prime for about 2 hours and then letting it settle for the remaining time and carrying out testing. This will be my final testing :(:cool:.


IC Diamond will you be able to send me 3-4 contact papers for washer modded lapped testing?

Now all I need is few coins or may be washers.:)
 
I would like the info so I will send you some more. Your anomaly of light pressure, light contact, improved thermal adds to our library of results while it's always about contact and pressure there are gradiations in between that are always not easily explainable so the addition will add depth to our troubleshooting efforts.

Nice work, good analysis btw - have to summarize all the results makes a great article
 
IC, quick question, just got a tube of IC Diamond and IC Perihelion today so gonna start testing tonight with a few different oc's-probably stock 2.66ghz, 3.0 3.2 and 3.4 and if time allows 3.6&3.8

but what kind of times are folk generally letting their prime95 run for on this test, is there an average length of time or a fixed minimum we must use?

too lazy to read the entire thread to find out if my question has already been answered, sorry :(

cheers for the TIM btw
 
The table on the first page should have an extra column for better representation of the cooling difference. As well as temp difference there should be temp difference minus ambient difference. EG the third result in the table has a load difference of -10C but the before ambient was 31C and the after 21C, giving an ambient difference of -10. In this case the temp difference minus ambient difference is -10--10=0C which accounts for the change in ambient.
 
Thekwango said:
IC, quick question, just got a tube of IC Diamond and IC Perihelion today so gonna start testing tonight with a few different oc's-probably stock 2.66ghz, 3.0 3.2 and 3.4 and if time allows 3.6&3.8

but what kind of times are folk generally letting their prime95 run for on this test, is there an average length of time or a fixed minimum we must use?

too lazy to read the entire thread to find out if my question has already been answered, sorry :(

cheers for the TIM btw
Click to expand...

It's up to you to decide for how long you want the prime to run for. Could be 10,20,30mins, 1hour etc.

However you must keep the same prime settings for each thermal compound you are testing. The testing must include your older thermal compound aswell since that is the one against whom you are comparing IC products.

All your test settings including bios settings must remain same in all the tests for each thermal compound.

You also need to record ambient temperature aswell before loading your cpu in each test :)
 
WingZero30 said:
It's up to you to decide for how long you want the prime to run for. Could be 10,20,30mins, 1hour etc.

However you must keep the same prime settings for each thermal compound you are testing. The testing must include your older thermal compound aswell since that is the one against whom you are comparing IC products.

All your test settings including bios settings must remain same in all the tests for each thermal compound.

You also need to record ambient temperature aswell before loading you cpu in each test :)
Click to expand...

Ahhh cheers WingZero.

I was just going run the blend test each time leaving the settings at default
i.e min fft (in k) @ 8
max fft @ 4096
mem to use (in MB) @ 1600

and for bios the oc's would be using the same multiplyer/bclk settings each time with the different compounds

just wasn't sure if there was a set time we needed to run it for.

cheers for the info

edit: just need to find a thermometer to get my ambient temp
 
IC Diamond said:
run test 2 hours min after install to let the compound settle
Click to expand...

Lol sorry I meant after the thermal compound has settle down for few hours:p.

@ Thekwango

Once you apply the thermal compound, you can run prime95 for 2 hours to heatup the TIM in order for it to settle down and then commence testing.

Alternatively after you have applied TIM, let it settle down for 2-3 days by making general use of computer and then commencing testing
 
You must log in or register to reply here.
Back
Top Bottom