Tim0n Posted January 29, 2012 Share Posted January 29, 2012 Hey I just bought a H80 cooler and stupid as I am i bought aftermakret thermal paste too and removed the pre-applied paste. So I mounted the radiator did a nice stright line on the cpu and installed the H80. ( It didn't got so good because I seem to have a very tight chassi and motherborad so I had a ahrd time making the h80 fit with the radiator on.) And then I got round 50 celcius idle which I didnt think was so good, So today I tested to reinstall the cooler and this time I did a small dot on the cpu instead after a while I also got the cooler to fit and started the comp and now I got round 68-70 celsius idle on core 1! So now I have to re-attach it again and wounder if there are any other techniques that I can use? Like credit card or something. Or should I remove the radiator from the case then install the h80 and then the radiator? Link to comment Share on other sites More sharing options...
garetjax Posted January 30, 2012 Share Posted January 30, 2012 Well, depending on the TIM you purchased, each manufacturer has their own method that their paste should be applied. For example, Arctic Silver 5 (AS5) uses the straight line method on many Intel processors, while ICD7 uses a pea-sized gob on the exact center of the IHS. Check with the manufacturer of your TIM and find out the proper way to apply their grease. The temperature differences, as you have noted, can be quite substantial when applying it their way versus your own. For the record, I removed the stock grease that came on my H80 and replaced it with ICD7 using a pea-sized dollop on the center of my Intel 1155 processor IHS, per their instructions. Link to comment Share on other sites More sharing options...
parsec Posted January 30, 2012 Share Posted January 30, 2012 I doubt very highly that the TIM you used, or your application method is the cause of your high temp issue. What did the TIM imprint on the CPU and pump's heat plate look like when you removed it? Complete and even coverage? We don't know what your hardware is, which could help. First question when tracking down issues with high temps, what is the pump's speed? Link to comment Share on other sites More sharing options...
Tim0n Posted January 30, 2012 Author Share Posted January 30, 2012 I doubt very highly that the TIM you used, or your application method is the cause of your high temp issue. What did the TIM imprint on the CPU and pump's heat plate look like when you removed it? Complete and even coverage? We don't know what your hardware is, which could help. First question when tracking down issues with high temps, what is the pump's speed? Okay I got a Q9450 cpu, a Fractal design r3 case, mx-4 thermal compound, gtx 260, motherboard: Gigabyte Ga-p35-ds3. And the fan speed on the pump are fine. Link to comment Share on other sites More sharing options...
parsec Posted January 30, 2012 Share Posted January 30, 2012 Well, your mother board might be one of those that has parts very close to the CPU socket, and they contact the bottom of the pump, causing a less than optimal contact between the CPU and the pump's heat plate. We've seen that before with socket 775 boards. That need so to be checked before anything else. BTW, your pump speed is... Link to comment Share on other sites More sharing options...
Tim0n Posted January 30, 2012 Author Share Posted January 30, 2012 Well, your mother board might be one of those that has parts very close to the CPU socket, and they contact the bottom of the pump, causing a less than optimal contact between the CPU and the pump's heat plate. We've seen that before with socket 775 boards. That need so to be checked before anything else. BTW, your pump speed is... Well I reseatted the cpu.. again and this time did the line method and now it seems to give better temps Core 0: 40C Core 1: 39c Core 2: 37 c Core 3: 38 C and how do I check pump speed is that the speed of the fans on the pump or what do you mean? Link to comment Share on other sites More sharing options...
parsec Posted January 30, 2012 Share Posted January 30, 2012 If you have the single wire/three pin fan connector from the H80's pump connected to the CPU Fan header (recommended) or another fan header on your board, the "fan speed" reading of that header is really the pumps speed. If you have the H80's fans connected to it's fan controller (the usual method) you can't monitor the fan speed, since they are not connected to the board. So with the normal H80 configuration, you can only monitor the pumps speed, which should be constant, with a small +/- variation. Link to comment Share on other sites More sharing options...
Tim0n Posted January 30, 2012 Author Share Posted January 30, 2012 If you have the single wire/three pin fan connector from the H80's pump connected to the CPU Fan header (recommended) or another fan header on your board, the "fan speed" reading of that header is really the pumps speed. If you have the H80's fans connected to it's fan controller (the usual method) you can't monitor the fan speed, since they are not connected to the board. So with the normal H80 configuration, you can only monitor the pumps speed, which should be constant, with a small +/- variation. Okay well I guess it is the fan 1 speed: wich on maximum h80 speed (the third part, perfomance mod) is varying between 2027-2096rpm Link to comment Share on other sites More sharing options...
parsec Posted February 1, 2012 Share Posted February 1, 2012 Your H80 has two basic parts, the pump and heat plate that take heat from the CPU and transfer it to the coolant which is then pumped to the radiator, and the fan speed controller. The pump runs at a constant speed all the time, the fan speed selection button has no affect on it. The pump has a part that spins quickly to circulate the liquid for cooling, and that part's speed can be measured and displayed like a fan's speed. The pump should be at ~2000 RPM, so your pump speed is fine. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.