Is this lapping job adequate?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

DaPoets

Junior Member
Aug 22, 2007
22
1
66
I ended up lapping both my Q6600 and my Thermalright Ultra 120 Extreme tonight... yeah I'm tired... It took me about 3.5 hours total and I went through 220, 400, 800, 1000, 1500 grits on both of them. It was crazy how warped my heatsink was and my Q6600 was a mountain range with a surrounding moat...

Running at 3.0 ghz Before Lapping:
Idle Load
43c 69c
40c 63c
43c 69c
39c 63c

After Lapping both Q6600 + Thermalright Ultra 120 Extreme
Idle Load
42c 62c
38c 58c
42c 62c
37c 58c

http://img208.imageshack.us/im.../image080resizeww0.jpg

http://img473.imageshack.us/im.../image092resizevz2.jpg

http://img515.imageshack.us/im.../image093resizeio4.jpg

http://img296.imageshack.us/im.../image098resizehh4.jpg

http://img204.imageshack.us/im.../image100resizedh3.jpg

http://img529.imageshack.us/im.../image101resizeqa6.jpg

http://img513.imageshack.us/im.../image104resizevi1.jpg
 

BonzaiDuck

Lifer
Jun 30, 2004
16,287
1,866
126
Well, that's good -- good enough.

I didn't lap my processor cap, and I have you beat by a considerable margin.

I've been away from the forums for several days, as I refined my ducting mod for my Q6600 (B3), improving the seal of the duct-parts to the case and fans, and adding a small 80x15mm fan to the VGA duct extension.

I also (finally) got all my fans connected to the motherboard plugs. The StrikerExtreme has eight fan plugs, and provides BIOS thermal control to four of them excluding the CPU_FAN. The thermal control requires thermal sensors, so there are two-pin plugs for sensors each associated with three of the fans just mentioned. And -- whooppee-doo-da -- I am really, really pleased with how the Striker controls these fans. Apparently, this was finally corrected or improved in BIOS version 1301.

@ 3.0 Ghz, my peak, core load temperature (highest among the four) is 57C with room ambient of 77F.

I'd be interested in the room ambient that prevailed during your load tests. This is not a matter of competition -- I was joking when I said "I have you beat."

I need to know if lapping my processor cap is going to give me better temps, and you might want to know if it was really worth it. So the key for comparison purposes is the room-ambient. I follow my own convention of reporting room temperature in Fahrenheit, but if you give it to me in Celsius, I won't be inconvenienced to do the conversion. Of course, the loose joint in this is how flat your processor cap was before lapping -- or how flat mine is without lapping.
 

DaPoets

Junior Member
Aug 22, 2007
22
1
66
Originally posted by: BonzaiDuck


I'd be interested in the room ambient that prevailed during your load tests. This is not a matter of competition -- I was joking when I said "I have you beat."

I need to know if lapping my processor cap is going to give me better temps, and you might want to know if it was really worth it. So the key for comparison purposes is the room-ambient. I follow my own convention of reporting room temperature in Fahrenheit, but if you give it to me in Celsius, I won't be inconvenienced to do the conversion. Of course, the loose joint in this is how flat your processor cap was before lapping -- or how flat mine is without lapping.

My house is brick and I had all the lights on in my office, windows open but there was no breeze last night, so I would say it was about 82 F. My house just stays warm unless I put on the Central Air. My house thermostat said 86 F last night but it didn't seem that warm so I'm knocking it down a notch...
 

BonzaiDuck

Lifer
Jun 30, 2004
16,287
1,866
126
By the looks of things, and as I review your peak temperatures before and after lapping the heatspreader cap, you got a 7C degree improvement -- which I have to say is stunning.

In my situation, I had purchased the "custom-lapped" TR Ultra-120-Extreme from SVC, so I was half-way there.

My motherboard duct-design includes a duct for the U-120-Extreme, but the cooler is so efficient that ducting the cooler itself just allows me to avoid hanging a fan directly on the cooler, and probably results in a negligible temperature drop on the CPU by itself.

The main advantage of the duct is in the cooling of motherboard components, like the chipsets and mosfets. When I built and tested the duct, I hadn't put a thermal sensor on the northbridge cooler, but I did just that yesterday. I now find under various load tests that the northbridge temperature measured with the sensor at the NB heatsink base and at the junction with its copper heatpipe is 40C at idle and 40C at load. Considering that the temperature measured the same way on my old Prescott system's NB was 40C at load and this new NB employs the 680i chipset which consumes significantly more power, I'd say the ducting is a real advantage. On the old system, the temperature would actually vary between idle and load by about 4 to 6C. To get the "real" NB temperature using these tape-on sensors, one would have to add several degrees to the measured temperature, maybe in the order of magnitude with the difference between TCase and TJunction measured on our Intel CPU's.

It looks as though lapping the processor cap may be useful. Another member here, clocked to 3.0 Ghz using a very good and well-ventilated case at a 78F room-ambient, was still 3 to 4C higher at load than mine at 3.2Ghz, but he'd only lapped the cooler heatsink base for the U-120-Extreme.

 

BonzaiDuck

Lifer
Jun 30, 2004
16,287
1,866
126
. . . . One more observation.

For various reasons over issues I had with our thermostat thermometer and another analog thermometer I was using, I bought a digital thermometer. But it, too, was inaccurate, and I had to calibrate it with a bowl of ice-water. It would read 80.1F, but the calibrated value was 77.0F.

So it's probably "iffy" to compare systems, even with reported room-ambients, unless the same thermometers were used. It's still "a loose joint."
 

BonzaiDuck

Lifer
Jun 30, 2004
16,287
1,866
126
. . . and another . . .

Even if room-ambients weren't "loose-jointed," in comparing your temperatures to mine, I forgot the conversion between F and C. Even a 7F-degree difference in room-ambient would result in only a difference of about 3.8C degrees.

I'm just not convinced at this point that lapping the heatspreader is going to get me that much. I think that lapping the Extreme's base was worth 4 or 5C to our fellow member.

AigoMorla will appreciate this: Ducting and lapping may be worth a considerable margin, but the improvements are still "marginal" by comparison with water-cooling.
 

MrDudeMan

Lifer
Jan 15, 2001
15,069
94
91
People here aren't grasping the long-term effects of lapping to shave off a few degrees C. There is a reason the heat sinks are convex. Did you ever consider that?

During the initial thermal cycling, the substrate (which is soldered to the outermost metal layer) undergoes considerable expansion. The warm-up from idle to full load is again a considerable expansion. Herein lies the first point: the coefficient of expansion, which dictates how much stress and strain is put on the substrate, CPU, and solder, is a constant and can not be changed. To counter-act this expansive force, a resin based material is injected to fill the gaps and even out the pressure. The convexity of the heatsink is then designed to put immense pressure on this area to resist the expansion of the CPU, thus relieving the stress and strain. When you decrease the amount of relief (i.e. lapping), you increase the range of motion of all of the parts involved. Eventually the solder and surrounding connections will become inelastic and shear.

The second point: it is very much a myth that heat is the primary cause of a CPUs inevitable death. What people don't realize is that their precious CPUs are subjected to several hours of tests at 120C before they get it. Silicon doesn't start to change properties until several degrees above that, which is why 120C is used as an upper boundary. Safe operating temperatures for longevity are below 80-90C, but not because heat magically kills your CPU - it's because of electromigration. The higher the energy of the electrons, the higher chance you have of causing electromigration, or breakdown in the poly causing an eventual dead transistor. The increase in electromigration is far greater with increased voltage, however, so increased temperature really can only claim it as a second order effect.

Moral of the story: lapping your CPU will gain you a few degrees C in the short run, but very much limit the lifespan of your CPU. For some people this is a non-issue due to a constant upgrade path, but for others it could very much be something to investigate. If you plan on keeping your rig for 3-4 years, lapping your CPU would be a bad idea. If you plan to swap out your CPU once a year then it doesn't really matter.
 

BonzaiDuck

Lifer
Jun 30, 2004
16,287
1,866
126
Thanks, Dudeman --

Only as a footnote on the matter of CPU longevity and temperature, there are contacts between metal leads and the silicon. At higher temperatures, these connections, over time, can separate.

There may be some CPUs out there with warped heatspreaders, but ignoring the very perceptive discussion you've made about the expansion of parts, I've been concerned that lapping the heatspreader might weaken it. Any company, even if going through growing pains, staff turnover, misplaced priorities or other issues, would not release a product to the public without a careful determination of the force or pressure it exerts on components like the heatspreader. They don't make assumptions about the lapping of heatspreaders, and so I would worry about lapping it.

The role of temperature and its interplay with voltage and resistance bears most heavily on our concern with temperature and over-clocking. Nobody on the forums or anywhere else has managed to quantify this relationship, although I'm sure it can be done -- somehow, by some-one. Like I said before in another thread, you can reduce temperatures by any means, but it can also lead to false security as regards voltage settings. Our meager attempts to reduce temperatures under an air-cooling regime has some marginal benefits. For instance, after ducting my rig, I was able to get the same over-clock settings and stability after reducing the voltage by one or two notches. To nutcases such as ourselves, this may seem marvelous, but it is still marginal.
 

mjrtoo

Member
Jul 25, 2007
120
0
0
That's interesting, thanks for the different point of view. Maybe I'll pick up another Ultima 90 and use it stock and see what happens.

Anyone 'want' a lapped Ultima 90?
 

Noubourne

Senior member
Dec 15, 2003
751
0
76
Well a lapped heatsink wouldn't affect the lifespan of the CPU. He was talking about the IHS from what I can tell.

IBM seems to think that some amount of "microchannels" are good for cooling and assist in TIM spreading when mounting their IHS.
 

MrDudeMan

Lifer
Jan 15, 2001
15,069
94
91
Originally posted by: BonzaiDuck
Thanks, Dudeman --

Only as a footnote on the matter of CPU longevity and temperature, there are contacts between metal leads and the silicon. At higher temperatures, these connections, over time, can separate.

There may be some CPUs out there with warped heatspreaders, but ignoring the very perceptive discussion you've made about the expansion of parts, I've been concerned that lapping the heatspreader might weaken it. Any company, even if going through growing pains, staff turnover, misplaced priorities or other issues, would not release a product to the public without a careful determination of the force or pressure it exerts on components like the heatspreader. They don't make assumptions about the lapping of heatspreaders, and so I would worry about lapping it.

The role of temperature and its interplay with voltage and resistance bears most heavily on our concern with temperature and over-clocking. Nobody on the forums or anywhere else has managed to quantify this relationship, although I'm sure it can be done -- somehow, by some-one. Like I said before in another thread, you can reduce temperatures by any means, but it can also lead to false security as regards voltage settings. Our meager attempts to reduce temperatures under an air-cooling regime has some marginal benefits. For instance, after ducting my rig, I was able to get the same over-clock settings and stability after reducing the voltage by one or two notches. To nutcases such as ourselves, this may seem marvelous, but it is still marginal.

The contacts between the metal leads and silicon are not subject to any stress. The contact area between these two layers would tolerate barely any fatigue before breaking. The metal layers (7-8 in consumer, 9-10 in enterprise) aren't subject to any stress either. As I mentioned in my previous post, the problem exists between the outermost metal layer and the pin taps. The resin is injected to keep them in place and to spread pressure. Lapping the convexity off of the IHS or heatsink reduces the effectiveness of the resin. The pressure should be around 40-50lbs/sq. in., which is almost completely a function of the distance the springs must stretch to clamp down the CPU. When you lap, you reduce the distance, thus reducing the force since the spring constant is just that, a constant. F = kx. Force is the result of the spring constant multiplied by the distance, and in this case it is pretty much linear, so the equation works. If you reduce x, you reduce F. When you reduce F, you reduce the lifespan of the CPU (significantly, not just in an academic sense). Again, whether or not this matters to you depends on how long you plan to keep the CPU.

Lastly, and this is just an aside, your ability to overclock is largely not bounded by the temperature. Warmer transistors tend to switch faster than cooler transistors as a matter of fact. The crux of the issue is where the part was binned (the clock frequency) due to how well it passed screening tests. I actually write these tests for a living, so this is an area I am fairly knowledgable about. The common programs used to test stability don't fully exercise both the functional and electrical capabilities of the CPU. Obviously I can't distribute the tests I write/use, but most overclocked CPUs would fail miserably when you were sure they were stable if they were to execute a verified statistically maximal stressor. Think of it this way - if the CPUs could be sold reliably at a higher frequency, they would be. The reason you think your chip is stable is because you may not run across the particular instruction combination or data set that showed a failure, but just because you won't doesn't mean no one else will either. Thus the clock frequency is lowered by a statistical amount to "guarantee" valid answers with any functional or electrical combination of data patterns.

Basically what I'm saying is your ability to overclock hinges on where your particular die was on the wafer during manufacturing (diffraction patterns influence overclockability) and luck.

Originally posted by: Noubourne
Well a lapped heatsink wouldn't affect the lifespan of the CPU. He was talking about the IHS from what I can tell.

IBM seems to think that some amount of "microchannels" are good for cooling and assist in TIM spreading when mounting their IHS.

I was talking about anything to do with applying pressure on the CPU. This includes the heatsink as well as the IHS.

 

BonzaiDuck

Lifer
Jun 30, 2004
16,287
1,866
126
Good to have your input on these matters, Dudeman, so stick around (and with 10,000 posts under your belt, I don't see you going anywhere unless you punch out from natural causes to be found with your face mashing down on your keyboard.)

"Metal layer and outermost pin-taps" -- I think this is what I was referring to -- having got it from Dr. [Somebody] in a THG article.

So -- on the general level -- increased resistance in metals reduces electrical efficiency, requiring more voltage? Increases in heat arise from voltage increases? And some semiconductor materials actually display less resistance at higher temperatures and therefore better conductivity?

So reductions in temperature -- even through some radical phase-change method -- might improve electrical efficiency in some components and not in others?

See, with what you've said, and what I gleaned from the THG articles, it would seem that reducing temperatures has a questionable effect on over-clocking results, since some components improve in performance and others degrade. We see improvements of 300 Mhz or so in water-cooling versus air-cooled systems nevertheless, but this is because the over-clocker is simply boosting voltage while keeping the heat reduced enough to avoid the thermal limitations of the processor. In the long run, he is reducing the life of the CPU from electromigration through the voltage boosts.

So people reduce temperatures through water-cooling, then pump up their VCOREs way too far beyond the manufacturer spec -- assuming that the resulting temperature regime means they're not going to damage anything. But -- they do anyway -- for the voltage increase.

Here's an observation about the ThermalRight Ultra-120 (and Extreme) cooler, and other coolers built with that design. SVC told me they shave off less than a millimeter of heatsink base to flatten it. That's really a thin sliver of copper. If one were worried about the reduced pressure on the heatspreader, you could simply add (nylon or metal) washers to the bottom of the springs in the mounting assembly to compensate.

Of course, you're saying that this pressure is concentrated along the ridge of the convex surface of the heatsink base, and that removing that ridge sets up an entirely different "pressure regime." But if that were so, wouldn't TR recommend against mounting the cooler in one direction as opposed to another?

Your first-hand remarks about spec limits and statistics confirm second-hand assumptions or deductions I have made as to how spec limits are set in the first place, and ultimately with the goal reducing RMAs-under-warranty to near-zero.
 

MrDudeMan

Lifer
Jan 15, 2001
15,069
94
91
Originally posted by: BonzaiDuck
Good to have your input on these matters, Dudeman, so stick around (and with 10,000 posts under your belt, I don't see you going anywhere unless you punch out from natural causes to be found with your face mashing down on your keyboard.)

"Metal layer and outermost pin-taps" -- I think this is what I was referring to -- having got it from Dr. [Somebody] in a THG article.

The reason I clarified is because you said:

there are contacts between metal leads and the silicon. At higher temperatures, these connections, over time, can separate.

which isn't the case. The outer metal layer is a metal to metal connection, not metal to silicon.


So -- on the general level -- increased resistance in metals reduces electrical efficiency, requiring more voltage? Increases in heat arise from voltage increases?

Not quite. Increased resistance in any of the materials means you need more voltage, but voltage and current have a very complex and intertwined relationship. The current is what causes the heat, not the voltage. You can think of voltage like potential energy, and current like kinetic energy. The rush of current through any material is a result of the differing potentials applied to it.


And some semiconductor materials actually display less resistance at higher temperatures and therefore better conductivity?

Resistance and heat are again a complex relationship. Light bulbs for example gain massive resistance as they heat up. When you first turn on an incandescent light, thousands of amps are surging through it. As it gets hotter, the resistance increases exponentially and it stabilizes.

MOSFETs are also bound by this thermal runaway limit. The hotter a MOSFET gets, the smaller the available channel for the electrons/holes (depending on the doping) to pass through. When I said transistors switch faster, that doesn't mean the circuit will work at a faster frequency. Unlike integrated circuit MOSFETs, Power MOSFETs are at risk of thermal runaway. Thermal noise injection and junction temperature play a complicated role in all of this.


So reductions in temperature -- even through some radical phase-change method -- might improve electrical efficiency in some components and not in others?

I'm not totally sure how to answer this, but lower temperature usually means lower resistance if that is what you are asking. The mechanisms which dictate the rate of decrease change as you hit certain boundaries (such as the Debye temperature), but the resistance is a net decrease.


See, with what you've said, and what I gleaned from the THG articles, it would seem that reducing temperatures has a questionable effect on over-clocking results, since some components improve in performance and others degrade. We see improvements of 300 Mhz or so in water-cooling versus air-cooled systems nevertheless, but this is because the over-clocker is simply boosting voltage while keeping the heat reduced enough to avoid the thermal limitations of the processor. In the long run, he is reducing the life of the CPU from electromigration through the voltage boosts.

Pretty much. Be careful thinking that higher temperatures are a good idea. I simply said the transistors can switch faster, but the circuit will operate slower. The mobility of the electrons/holes has a lot to do with this.

The electromigration effect isn't noticeable immediately, but increasing the voltage is a definite way to cause it. The higher energy the electrons are, the more atoms are ejected from the channels where contact is made. If enough atoms are removed from the channel, the transistor will become a high impedance open circuit instead of a semiconductor, meaning it is always off.


So people reduce temperatures through water-cooling, then pump up their VCOREs way too far beyond the manufacturer spec -- assuming that the resulting temperature regime means they're not going to damage anything. But -- they do anyway -- for the voltage increase.

Right. People are way too worried about temperature. Running at a higher voltage is far more detrimental. The Q6600 is cooled very much within spec using the stock heatsink. There is no need for water blocks just to run it at stock speed, which I have heard many people say.


Here's an observation about the ThermalRight Ultra-120 (and Extreme) cooler, and other coolers built with that design. SVC told me they shave off less than a millimeter of heatsink base to flatten it. That's really a thin sliver of copper. If one were worried about the reduced pressure on the heatspreader, you could simply add (nylon or metal) washers to the bottom of the springs in the mounting assembly to compensate.

The extra force would be deflected onto the frame instead of a relatively small area above the die. I understand what you are saying, and it would probably add some marginal increase in pressure, but nothing like what the convexity of the bottom would have added.


Of course, you're saying that this pressure is concentrated along the ridge of the convex surface of the heatsink base, and that removing that ridge sets up an entirely different "pressure regime." But if that were so, wouldn't TR recommend against mounting the cooler in one direction as opposed to another?

The convexity of the bottom is supposed to be circular, meaning that it can be mounted in any configuration as it should be uniform.
 

BonzaiDuck

Lifer
Jun 30, 2004
16,287
1,866
126
I'll need to read over this some more. You know your stuff. It's too easy for me to simply discuss things in terms of general negative and positive linear relationships -- like an economist. It's been 40 years since college physics and almost as long for college thermodynamics.

You would also think that boosting the clock speed, even at the same voltage, requires more power in watts, and therefore one is increasing the current in amps, and I think that is implicit on one of your statements.

I'm sure we all accept the decreased life-expectancy of these processors. I've got mine (a B3 stepping) volted at 1.3V for a 3.0 Ghz / 1,333 Mhz FSB speed, but it's set to 1.4188V for a 3.2 Ghz speed. The first setting is comfortably below the Intel retail box "Maximum" spec, and of course the second setting is 9% above that spec. You'd figure that they choose the spec to avoid RMAs under warranty, and that the probability distribution of failure is positive-exponential with voltage. Over time, 9% may take a significant toll, but then again maybe not.

But the clock speed increases power usage and thermal wattage in a linear relationship, anyway, so it would increase current -- I would think -- a source of eventual degradation and damage.

CONVEXITY

ThermalRight's design is deliberate, but it is not a conical convexity -- it is, as someone put it -- "cylindrical" convexity. There is an even and precisely molded ridge that goes across the heatsink base -- the same height at one edge as at the middle or the other, opposite edge. So flipping the cooler around so the broad face points upward as opposed to pointing toward the rear exhaust port would change the line of even pressure completely, and the area contacted would be completely different. Yet TR seems to imply that installation in either position depends on the user's needs.

What I heard from ThermalRight was that they designed it to cross over the cores of a C2D, but this would result in less complete coverage using a C2Q. I'm still speculating that they might have done this to add flexibility to the spring-loaded installation -- meaning that the cooler could rock slightly from side to side without losing contact with the heatspreader, but your argument holds equal weight -- I'm just not sure that it is all that significant. I would think a flat and more pervasive contact with the heatspreader cap would be just as good if the pressure overall was sufficient -- it would be better for more efficient heat-transfer.

The other thing inconsistent with what you said about the "convex ridge" is that the original U-120, which I have, was certified to work with C2D and Presler cores, but was designed with a flat base. And the original stock cooler from Intel, although circular copper the size of a quarter or 50-cent piece (smaller than that, I think) -- is perfectly flat. Yeah -- I just checked it -- sitting on the coffee-table for "disposition" -- it is the size of a 50-cent-piece, and it is perfectly, perfectly flat. If that's the case, then one could simply compensate the loss of material lapped off by shimming the springs accordingly.
 

MrDudeMan

Lifer
Jan 15, 2001
15,069
94
91
Originally posted by: BonzaiDuck
I'll need to read over this some more. You know your stuff. It's too easy for me to simply discuss things in terms of general negative and positive linear relationships -- like an economist. It's been 40 years since college physics and almost as long for college thermodynamics.

You would also think that boosting the clock speed, even at the same voltage, requires more power in watts, and therefore one is increasing the current in amps, and I think that is implicit on one of your statements.

It only increases the average power used, which is measured in seconds. It doesn't increase the amount of power per cycle. The result is more heat is generated per second, but not per cycle.


I'm sure we all accept the decreased life-expectancy of these processors. I've got mine (a B3 stepping) volted at 1.3V for a 3.0 Ghz / 1,333 Mhz FSB speed, but it's set to 1.4188V for a 3.2 Ghz speed. The first setting is comfortably below the Intel retail box "Maximum" spec, and of course the second setting is 9% above that spec. You'd figure that they choose the spec to avoid RMAs under warranty, and that the probability distribution of failure is positive-exponential with voltage. Over time, 9% may take a significant toll, but then again maybe not.

I agree. Anyone overclocking needs to accept that they are doing damage to the CPU if they raise the voltage. There really isn't much else to say about that.


But the clock speed increases power usage and thermal wattage in a linear relationship, anyway, so it would increase current -- I would think -- a source of eventual degradation and damage.

It's not linear. Also, the amount of current isn't the cause of electromigration; it's the energy of the electrons, which is a function of the voltage. There are several different voltage domains across the CPU, so when you increase the Vcore, you are actually helping in some ways and hurting in others. For example, even though people think CPUs are strictly digital devices, they are not. The guts of a CPU are very analog in nature as the capacitance involved in the switching rounds out the waves to look almost like bad sine waves, not square waves. By raising the voltage, you have increased some of the noise immunity between different domains, such as the cache and the core because they run at different voltages, but you also have caused a rapidly increased breakdown between the interacting boundaries of those domains.


CONVEXITY

ThermalRight's design is deliberate, but it is not a conical convexity -- it is, as someone put it -- "cylindrical" convexity. There is an even and precisely molded ridge that goes across the heatsink base -- the same height at one edge as at the middle or the other, opposite edge. So flipping the cooler around so the broad face points upward as opposed to pointing toward the rear exhaust port would change the line of even pressure completely, and the area contacted would be completely different. Yet TR seems to imply that installation in either position depends on the user's needs.

What I heard from ThermalRight was that they designed it to cross over the cores of a C2D, but this would result in less complete coverage using a C2Q. I'm still speculating that they might have done this to add flexibility to the spring-loaded installation -- meaning that the cooler could rock slightly from side to side without losing contact with the heatspreader, but your argument holds equal weight -- I'm just not sure that it is all that significant. I would think a flat and more pervasive contact with the heatspreader cap would be just as good if the pressure overall was sufficient -- it would be better for more efficient heat-transfer.

The other thing inconsistent with what you said about the "convex ridge" is that the original U-120, which I have, was certified to work with C2D and Presler cores, but was designed with a flat base. And the original stock cooler from Intel, although circular copper the size of a quarter or 50-cent piece (smaller than that, I think) -- is perfectly flat. Yeah -- I just checked it -- sitting on the coffee-table for "disposition" -- it is the size of a 50-cent-piece, and it is perfectly, perfectly flat. If that's the case, then one could simply compensate the loss of material lapped off by shimming the springs accordingly.

I'm not in disagreement with how they designed their heatsinks. My initial purpose was to tell people if they are lapping a heatsink, which has been designed the be convex in some way, they are reducing the spring tension and causing some long term harm to their CPU.

ThermalRight designed their heatsink within spec I'm sure, so what can be extracted from this is that they met the height requirement for the heatsink to maintain pressure.

Finally, the reason there is supposed to be a convex bottom is because the amount of pressure designed to be put on the CPU actually causes it to go somewhat concave, although ever so slightly. The resulting concavity is more of a fit with a convex heatsink than a flat heatsink. These effects are extremely slight, but designed nevertheless.

The actual heat generated in a CPU is in very localized hot spots. It is most definitely not uniform across the surface of the die. The majority of the contact with of the IHS and heatsink is meant to be approximately over the hot spots. Most of the heat in the cache, for example, is diffused heat generated by the core or I/O traffic.

Shimming a spring is an extremely imprecise way to compensate for missing material. I wouldn't even bother with it anyway since most likely a person who is willing to lap their heatsink or IHS will have a new CPU before they see any negative effects introduced by the lapping.
 

BonzaiDuck

Lifer
Jun 30, 2004
16,287
1,866
126
Agreed. And over the last four years, it's cost me more than I'd like, but not for the processors burning out -- but more as a matter of "keeping up."

This stuff is addictive, and I've changed my poison from software to hardware since early in this decade.

I should really be saving my money for a fiberglass camper-trailer, so I can be somewhere nice when I'm the one who is going to "punch out for natural causes!"
 

BonzaiDuck

Lifer
Jun 30, 2004
16,287
1,866
126
MrDudeMan

I thought I saw an equation where boosting the speed in Mhz caused linear increase in heat, while boosting the voltage increased heat by its square. However, just as I write this, I realize you're speaking about the electron-migration.

This has been a very interesting discussion.
 

WoodButcher

Platinum Member
Mar 10, 2001
2,158
0
76
Originally posted by: BonzaiDuck
MrDudeMan
This has been a very interesting discussion.

Wow, I'll say. Most of it over my head so I'll spend a few hours doing my homework. Dude, Duck, thanks for the lessons. Your not charging tuition, are you?


 

MrDudeMan

Lifer
Jan 15, 2001
15,069
94
91
Originally posted by: BonzaiDuck
MrDudeMan

I thought I saw an equation where boosting the speed in Mhz caused linear increase in heat, while boosting the voltage increased heat by its square. However, just as I write this, I realize you're speaking about the electron-migration.

This has been a very interesting discussion.

I meant the relationship between power and heat isn't linear. I guess it could be depending on the design of the heatsink, but usually it isn't.

A higher frequency wouldn't yield a net increase in instantaneous power simply due that change alone - it would depend on what instructions were being executed. What I'm saying is turning up the frequency only increases the average power, or joules per second if you want to think of it that way instead of just watts. When the clock rate increases, there is less time between cycles which means there is less time for the heatsink to cool off.

Increasing the voltage increases both the instantaneous power due to higher currents and the average power since each cycle is using more power than at the lower voltage. This is compounded by higher temperature causing more leakage thus creating a vicious cycle, which is only controlled by the heatsink's ability to extract heat or sensors which will throttle the frequency. I am admittedly not very knowledgeable in this area (thermodynamics I mean), but I think this is right.
 

WoodButcher

Platinum Member
Mar 10, 2001
2,158
0
76
Good then I can afford to ask a question of you all,
If in the case of a mirror finish on both surfaces as many seem to prefer wouldn't something thin or more fluid- like be better for thermal transfer than a paste that would not squeeze out? maybe half a drop of mineral oil mixed with that diamond dust of yours Duck?
 

BonzaiDuck

Lifer
Jun 30, 2004
16,287
1,866
126
I wouldn't know the thermal properties of mineral oil, but maybe it's worth trying. Remember my "Silver Eagle" experiment: the worst that happened was the CPU showing an idle temperature of 55C, so I shut down and removed it. The fact was that I'd attempted to embed diamond powder in the silver by pounding the coin between two polished, flat steel plates, but this left air pockets -- it was a bad idea.

I haven't read anything that is really definite about how thin should be the TIM layer, but too much or too little does degrade performance. Arctic Silver tests their paste with a layer that is 0.001" thick.

As long as there's enough pressure to flatten the paste evenly, and as long as the medium eliminates air pockets, I'd think it wouldn't matter -- thick or watery. With the diamond stuff, 10% content makes it spread easily, but only matches AS5 in performance. With the IC Diamond, they tried to drive content above 90%, and it's hard to spread. But they also used a medium which partially evaporates, so it spreads easier to begin with, and then leaves an even layer of diamond particles with some medium still in it.

But you can't have the stuff so powdery and dry that it makes an uneven layer or contains air-pockets. "Silver Eagle" Experiment #1 proved that to me.

AigoMorla said joking to someone that a diamond disk would be preferred. But think about it: like the silver, it would require two thermal interfaces -- top and bottom, and you'd need some sort of TIM just for the disc against heatspreader and heatsink base.
 

firewolfsm

Golden Member
Oct 16, 2005
1,848
29
91
For best results sand down to 1000 grit, do not wet the last sheet.

I sound like a bottle'a tropicana orange juiiccceeee

but no one here will get that, even though it sucked
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |