XP-90 not cooling as well as it should, any suggestions?

[sabrewolf732]

BS you go stand in the cold and see how cold you get, then go stand in the cold with 30mph wind, what's colder? the wind obviously.

no...., it just cuz the wind is taking away the heat you produce faster. If you die (all living matter in and around you) and take away all the heat at 20c, your body temp will be 20c. If there is a 1000MPH wind blowing over you, your body temp will still be 20c.

[scrandman]

Agreed. With high winds its how cold you feel as opposed to how cold you actually are...

[mcoleg]

Thermalright's heatsink bases are very flat. Hard to say what the deal is in the other thread. You can check your heatsink with a straight edge and a back light. Place a quality straight edge like a 6" machinist scale on the clean heatsink base and with a light behind the scale you can see any high or low spots. You can place the scale diagonally corner to corner and in the middle in both directions. Based on any number of XP-90 and XP-120 reviews these heatsinks have a flat surface and generally quite smooth. They are not lapped and for some folks they view this as criminal... In reality the difference in temp on the XP-90/120 before and after lapping is 1 degree C unless of course you have some really exceptionally warped heatsink, which would be very rare. FWIW I have a new XP-90 here and it is quite smooth and flat as was the case in this review and others.

http://www.silentpcreview.com/article194-page2.html

just passing the info. it's an emperical evidence, i'm afraid; some are just not 100% flat according to user experiences. i am not tr hater, btw, got several of their sinks myself and i like them.

[RATIFIED]

I too am having a similiar problem with my 3500+ but on a Abit AV8. I have seen a few posts saying it is actually the cpu giving false readings on the temps. I have also seen a post saying that its the mobo manufactors fault because 2 pins are used to give temps and the mobos use only one!!! I am using a XP-90 with a medium flow Panaflo 92mm fan with AS5. My idle temps are only 4-5C above my case temp, but load temps soar to 30-35C above cae temps!!!!!

[Beenthere]

Guess Ricki Mate doesn't quite understand "wind chill" but it only applies to the "feel" a person gets when standing in the cold wind. There is no such thing as "wind chill" for non-living things.

Well as I said it's easy enough to check a heatsink for flatness, which I would think anyone would do before installing it. If it needs a few swipes across some silicon carbide sand paper then do it. The point was ALL of the Thermalright heatsinks I've ever seen were flat and had a resonably smooth finish as they are milled. I've never seen a review where they claimed the Thermalright heatsink was warped. Some folks do like to lap the milling machine marks off and that's just fine. The only portion of the heatsink that needs to be flat anyway is the part that actually contacts the CPU die or heat spreader. Anything else is out in the free air and has no contact with the CPU.

The Mobo thermister or diode and software are not the most accurate means to determine CPU temps. They are there for relative temps. If you really want to know the CPU temp then you need an embedded thermocouple and calibrated circuit readout. For those who are interested AMD has details in their Tech Docs on exactly how and where to place the thermocouple. As I said originally when this person said the case temp was 27C and that was cooler than the ambient room temp - IT AIN'T POSSIBLE so the temps are suspect.

[mcoleg]

i guess it varies; m/b they make them at different places or something. was news to me when i herd of it - as i said i have several tr sinks and all of them are pretty darn flat, never had to sand them. of course, the newest i've got is sp97; that's pretty old now i figure...

[Beenthere]

Well it's pretty difficult to mill a heatsink un-flat with a sweep cutter as used on the Thermalright heatsinks. You can generally see the faint mill marks in a new Thermalright heatsink so I gotta take the horror stories with a large grain of salt until I see these warped ones with my own eyes. But even if it was slightly warped a few swipes on sandpaper should make it perfect. That clearly isn't the issue with this fellas PC from what I can see.

[mcoleg]

i really not sure what's going on there or why. i know the guy though, so i take his word for it. i think he did get it flat with some sanding.

[Vihsadas]

Guess Ricki Mate doesn't quite understand "wind chill" but it only applies to the "feel" a person gets when standing in the cold wind. There is no such thing as "wind chill" for non-living things.

Where did you hear that? Of course there is wind chill for non-living things. The cool air accelerating past any object will draw heat from that object. That's wind chill.

[Beenthere]

No as explained above there is no such thing as "wind chill" for inanimate objects. Check physics 101 if you have any doubt.

Thanks for playing along though !

[Vihsadas]

Actually, I am a physics major at Mcgill University, and I'm graduating in one year. How about you?

NO, this is physically IMPOSSSIBLE. The temp inside the case can NOT be cooler than the ambient room air unless you are running a vapor chill system. You can have a thousand fans blowing in and around the case but the lowest the case or CPU temp can ever be is whatever the cooling medium temp is and that's the ambient room temp in an air-cooled system.

What you said here is true. You cannot cool case temps lower than ambient temps, unless you have some object in the case that is cooler than ambient temps. This is correct.

However this, which is the only thing that I responded to:

Guess Ricki Mate doesn't quite understand "wind chill" but it only applies to the "feel" a person gets when standing in the cold wind. There is no such thing as "wind chill" for non-living things.

is not correct.
Wind chill is a phenomenon that makes it seem colder than it actually is. Wind chill will act upon ANY object that is warmer than ambient temps. Consider the two situations:

1) Air is stagnant surrounding a heated object
2) Air is moving past a heated object.

In the first situation, the heated object will lose air slower than in the second. This is because in the first, the air particles closer to the heated object will warm up first. BUT, as the "air that's closer to the heated object"'s temperature rises, it gains heat from the object SLOWER. That means that the other air in the area will gain heat even slower, (because heat transfer is based on difference in temperature, not absolute temperature).

However, in the second situation, air is flowing by a heated object, and since newer, colder air is continusously contacting the object, the difference in temps between the object and the flowing air is CONSTANT, and doesn't get smaller (like in the first situation). This is the phenomenon that we call "wind chill". It is an accelerated cooling of a heated object. More specifically, it has to do with any object that is continuously heated, and maintaining temperature. (like a human).
But to say that there is only windchill for livings things is inaccurate. There is windchill for every heated object. Hope this furthers your understanding.

Thanks for playing along though!

[Crovax33]

So do you guys think I should just say **** it and not worry about it? Maybe my temps are abnormally high because my temp reader is a little whack, or maybe my chip just runs hot?

Because other than lapping, I've tried pretty much everything in this thread, and nothing has helped get the temps any lower.

[mcoleg]

well, just on an outside chance, check it for flatnes as been said...

the sensors are off though on that board. tried new beta bioses?

[Lord Vetinari]

BS you go stand in the cold and see how cold you get, then go stand in the cold with 30mph wind, what's colder? the wind obviously.

As stated above windchill only applies to living or other objects that have significant surface moisture. The forced evaporation of the moisture cools the object. So unless your regularly spray your heatsink with moisture (not advised) wind chill does not affect CPU temps.
Obviously raising the speed of your fan to give more airflow will cool the HS more simply because you are carrying the heat away faster- but that is not windchill.

[Beenthere]

Amazing that some people just refuse to accept reality...

Vihsadas, I would suggest you ask your physics professor if "wind chill" applies to inanimate objects. When he stops laughing, ask him not to flunk you out of the course.

As several of us have noted... unless the relative humidity is about 100% there is no "wind chill" in an air cooled fan/heatsink system. Now if your PC is located in a room that has a London Fog all the time, well maybe then there is a "wind chill". My suggestion would be to remove your PC from that environment cause the electronics are not rates for 100% relative humidity. Honest ! Ask your prof.




Here: I did your homework for ya just to put an end to this silliness:


"Definition of Wind chill

Wind chill is the apparent temperature felt on the exposed human (or animal) body due to the combination of air temperature and wind speed. The wind chill temperature is always lower than the air temperature, even when the air is hotter than the body, because the wind increases the rate at which moisture evaporates from the skin and carries heat away from the body. The phase change of water (in sweat) from liquid to vapour requires that the molecules reach a higher energy state. That energy is acquired by absorbing heat from surrounding tissue by conduction. (See heat transfer)

Air movement increases the rate at which the temperature of an object reaches the temperature of the ambient air. Humans perceive or 'feel' this increased rate of heat transfer as wind chill.

The concept of wind chill is of particular significance in very cold climates such as the Arctic and Antarctic, at high altitude, at high speeds, or in very high winds. It is of great importance to the survival of humans and animals, and can even affect machinery and heating systems.

The official definition of windchill in meteorology was originally based on measurements taken at a distance above the ground. The exact definition of windchill has been controversial because it is a composite index, because animate and inanimate bodies behave differently, and because windchill reports have a major impact on winter tourism.

The first wind chill formulae and tables were developed by the United States military during World War II, initially by Siple and Passel working in the Antarctic, and were made available by the National Weather Service by the 1970s. In 2001 the formulae were revised to reflect more accurate theories and testing than those done by the military. These formulae are designed specifically for the human body, or even more specifically for the human face. Wind chill also affects animals, and wet, inanimate objects, but different formulae apply to them.

It has generally been conceded that the original model for windchill was not necessarily the best possible for all purposes. The physical basis for the calculation of windchill is now the relationship between the temperature, volume and pressure of a fluid. Moving air reduces air pressure and increases the cooling effect. Still air can actually insulate, which is why windchill was measured a number of metres above the ground rather than at ground level.

The new wind chill index used by the US and Canadian weather services is calculated from the following formula:

<math>T_{wc}=35.74+0.6215 T_a-35.75 V^{0.16}+0.4275 T_a V^{0.16}<math>
where Twc is the wind chill temperature in °F, Ta is the air temperature, and V is the air speed in mph.

As the air temperature falls, the chilling effect of any wind that is present increases; that is to say, a 10-mph wind will lower the apparent temperature by a wider margin at an air temperature of -4 °F (-20 °C), than a wind of the same speed would if the air temperature was 14 °F (-10 °C). Winds of stronger than 40 mph are assumed to have no significant additional chilling effect beyond the effect felt at that velocity, and the wind-chill phenomenon is thought to be absent altogether at air temperatures above approximately 68 °F (20 °C).

The US and Canadian formulae are best suited to extremely cold climates. Other formulae such as the Steadman wind-chill index (developed by Australian environmental scientist Robert Steadman) have been developed for temperate climates, but are less well known. Some wind-chill indices also take humidity into account."

http://www.wordiq.com/definition/Wind_chill_factor

THE END !