If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#11
|
|||
|
|||
Peter Hucker wrote:
What I am suspecting is the heatsink gunk. They supplied that silicone grease. I've never liked that stuff, it's nowhere near as effieient (probably as half of it oozes out under the pressure when you clamp the heatsink on) as the thermnal tape stuff that comes on boxed athlons. I might have to see if I can buy some of that tape stuff. It's like a sort of bluetack consistency. Thermal grease (if good quality), will out perform the bubblegum that you are about to change to, by quite a big margin. Saying that half of it oozes out points to incorrect application on your part. For an athlon-sized core use a tiny amount (half a grain of rice at most) and spread it paper thin onto the core using a suitable spreader. I suggest cutting your credit card in half length ways and using it to spread the thermal compound, this should allow you to get an even application with the added bonus that you won't be able to buy any more crap kit. On a serious note ensure that both core and heatsink base are very clean, using Isopropyl alcohol if possible with a lint-free cloth. Rub a small amount of the thermal compound into the heatsink base and clean it off with a dry lint-free cloth, this will fill any microscopic scratches, you should not be able to see any compound after cleaning. Apply compound to the core as described above and fit the heatsink. HTH -- Ian |
#12
|
|||
|
|||
On Tue, 29 Jun 2004 09:11:49 +0100, Apollo wrote:
Peter Hucker wrote: What I am suspecting is the heatsink gunk. They supplied that silicone grease. I've never liked that stuff, it's nowhere near as effieient (probably as half of it oozes out under the pressure when you clamp the heatsink on) as the thermnal tape stuff that comes on boxed athlons. I might have to see if I can buy some of that tape stuff. It's like a sort of bluetack consistency. Thermal grease (if good quality), will out perform the bubblegum that you are about to change to, by quite a big margin. Saying that half of it oozes out points to incorrect application on your part. For an athlon-sized core use a tiny amount (half a grain of rice at most) and spread it paper thin onto the core using a suitable spreader. Um surely too much will just cause the excess to ooze out lkeaving the correct amount? Anyway I may be ****e at it, so I'll stick to what works best for me - a pad. I suggest cutting your credit card in half length ways and using it to spread the thermal compound, this should allow you to get an even application with the added bonus that you won't be able to buy any more crap kit. ROTFPMSL! But I'll use one of the other 8! On a serious note ensure that both core and heatsink base are very clean, using Isopropyl alcohol if possible with a lint-free cloth. Rub a small amount of the thermal compound into the heatsink base and clean it off with a dry lint-free cloth, this will fill any microscopic scratches, you should not be able to see any compound after cleaning. Apply compound to the core as described above and fit the heatsink. Hmmm never thought of CLEANING the thing. I'll try that. -- 13 parrots and rising http://www.petersparrots.com 93 silly video clips http://www.insanevideoclips.com 1259 digital photos http://www.petersphotos.com Served from a watercooled dual 3.2GHz silent Athlon with 512GB of watercooled Raid. Why do they call it a TV set when you only get one? |
#13
|
|||
|
|||
Peter Hucker wrote:
ROTFPMSL! But I'll use one of the other 8! ;-) same as me then. On a serious note ensure that both core and heatsink base are very clean, using Isopropyl alcohol if possible with a lint-free cloth. Rub a small amount of the thermal compound into the heatsink base and clean it off with a dry lint-free cloth, this will fill any microscopic scratches, you should not be able to see any compound after cleaning. Apply compound to the core as described above and fit the heatsink. Hmmm never thought of CLEANING the thing. I'll try that. Cleanliness is very very important, never use your fingers to spread the stuff, you contaminate it with greases / skin cells etc. The aim is to get an extremely thin layer of thermal compound on the core, metal to metal conducts heat best and the thin layer is only there to fill the tiny scratches (think peaks and valleys). Metal to thermal-blutack-stuff to metal gives relatively poor thermal conduction compared to properly applied thermal grease. Give it another go with the grease supplied (or go buy some arctic silver) it does take some practice to get the layer thin and even, but it's worth it. I did one the other day and had to wipe it clean twice and start again before I was happy with it. IMHO it seems a shame to spend that kind of money on watercooling and install the waterblocks with bubblegum crap. -- Ian |
#14
|
|||
|
|||
On Tue, 29 Jun 2004 13:58:59 +0100, Apollo wrote:
Peter Hucker wrote: ROTFPMSL! But I'll use one of the other 8! ;-) same as me then. I got 8 so I could pick and choose. The offers keep moving about, and some do stupid things like charging me fees, which I phone them about, and if they don't refund them, the card is cut into pieces and returned to the manager. On a serious note ensure that both core and heatsink base are very clean, using Isopropyl alcohol if possible with a lint-free cloth. Rub a small amount of the thermal compound into the heatsink base and clean it off with a dry lint-free cloth, this will fill any microscopic scratches, you should not be able to see any compound after cleaning. Apply compound to the core as described above and fit the heatsink. Hmmm never thought of CLEANING the thing. I'll try that. Cleanliness is very very important, never use your fingers to spread the stuff, you contaminate it with greases / skin cells etc. The aim is to get an extremely thin layer of thermal compound on the core, metal to metal conducts heat best and the thin layer is only there to fill the tiny scratches (think peaks and valleys). Metal to thermal-blutack-stuff to metal gives relatively poor thermal conduction compared to properly applied thermal grease. Give it another go with the grease supplied (or go buy some arctic silver) it does take some practice to get the layer thin and even, but it's worth it. I did one the other day and had to wipe it clean twice and start again before I was happy with it. IMHO it seems a shame to spend that kind of money on watercooling and install the waterblocks with bubblegum crap. It's not really to make it ultra cool. As long as it's well within spec. My aim is to make it run silently, while not overheating. I've got CPUs at 50C, water at 35C, and room at 25C. Does this sound reasonable? -- 13 parrots and rising http://www.petersparrots.com 93 silly video clips http://www.insanevideoclips.com 1259 digital photos http://www.petersphotos.com Served from a watercooled dual 3.2GHz silent Athlon with 512GB of watercooled Raid. They say confuscious does his crosswords with a pen. |
#15
|
|||
|
|||
Peter Hucker wrote:
It's not really to make it ultra cool. As long as it's well within spec. My aim is to make it run silently, while not overheating. I've got CPUs at 50C, water at 35C, and room at 25C. Does this sound reasonable? What mobo is it? what prog are you using to measure the cpu temp? I ask because some mobos cannot measure the cpu diode temp only the cpu socket temp. The socket usually reads ~10C cooler than the cpu core. If 50C is the core diode temp and not the socket temp then it's reasonable. I get room 22C, water 29C and socket 37C (~47C core, mobo can't measure cpu diode) on mobile athlon 2400 (35w) o/c to 2.6 GHz, 1.8vcore, cpu, gpu & northbridge watercooled. -- Ian |
#16
|
|||
|
|||
On Tue, 29 Jun 2004 19:02:45 +0100, Apollo wrote:
Peter Hucker wrote: It's not really to make it ultra cool. As long as it's well within spec. My aim is to make it run silently, while not overheating. I've got CPUs at 50C, water at 35C, and room at 25C. Does this sound reasonable? What mobo is it? MSI K7D Master with two Athlon MP 2800+ what prog are you using to measure the cpu temp? BIOS, and the supplied Fuzzy logic 4 and PC Alert 3. I ask because some mobos cannot measure the cpu diode temp only the cpu socket temp. The socket usually reads ~10C cooler than the cpu core. If 50C is the core diode temp and not the socket temp then it's reasonable. Well it appears to crash around 70 - which would that suggest? I get room 22C, water 29C and socket 37C (~47C core, mobo can't measure cpu diode) on mobile athlon 2400 (35w) o/c to 2.6 GHz, 1.8vcore, cpu, gpu & northbridge watercooled. Sounds not bad then. -- 13 parrots and rising http://www.petersparrots.com 93 silly video clips http://www.insanevideoclips.com 1259 digital photos http://www.petersphotos.com Served from a watercooled dual 3.2GHz silent Athlon with 512GB of watercooled Raid. WinError: Unable to exit Windows. Try the door. |
#17
|
|||
|
|||
Peter Hucker wrote:
On Tue, 29 Jun 2004 19:02:45 +0100, Apollo wrote: What mobo is it? MSI K7D Master with two Athlon MP 2800+ what prog are you using to measure the cpu temp? BIOS, and the supplied Fuzzy logic 4 and PC Alert 3. I don't know anything about that mobo or software, sorry. I use Motherboard Monitor, you can pick the sensors to display, so if you pick the diode and the mobo can't measure it, you get some bizzare reading, I get -46C for my diode ;-) I'd suggest trying it, takes a while to set up but can display and log; all voltages, fan speeds (if connected to mobo headers), temperatures and lots of other useful stuff. http://mbm.livewiredev.com/ If you click on Motherboard list (on the same site), find your mobo and it tells you what sensor chips are used for what, it doesn't show 'diode' next to the sensors for your mobo, maybe it is socket temp, someone here should know I imagine. I ask because some mobos cannot measure the cpu diode temp only the cpu socket temp. The socket usually reads ~10C cooler than the cpu core. If 50C is the core diode temp and not the socket temp then it's reasonable. Well it appears to crash around 70 - which would that suggest? Again this could point to the socket being read, Athlons IIRC are rated for 85C, although I panic if mine hit 55C socket temp. The 50C you mentioned before is under full load isn't it? If you're getting 50C at idle that's too hot. I get room 22C, water 29C and socket 37C (~47C core, mobo can't measure cpu diode) on mobile athlon 2400 (35w) o/c to 2.6 GHz, 1.8vcore, cpu, gpu & northbridge watercooled. Sounds not bad then. I must admit that I'm addicted to overclocking ;o) -- Ian |
#18
|
|||
|
|||
On Tue, 29 Jun 2004 21:02:53 +0100, Apollo wrote:
Peter Hucker wrote: On Tue, 29 Jun 2004 19:02:45 +0100, Apollo wrote: What mobo is it? MSI K7D Master with two Athlon MP 2800+ what prog are you using to measure the cpu temp? BIOS, and the supplied Fuzzy logic 4 and PC Alert 3. I don't know anything about that mobo or software, sorry. I use Motherboard Monitor, you can pick the sensors to display, so if you pick the diode and the mobo can't measure it, you get some bizzare reading, I get -46C for my diode ;-) I'd suggest trying it, takes a while to set up but can display and log; all voltages, fan speeds (if connected to mobo headers), temperatures and lots of other useful stuff. http://mbm.livewiredev.com/ If you click on Motherboard list (on the same site), find your mobo and it tells you what sensor chips are used for what, it doesn't show 'diode' next to the sensors for your mobo, maybe it is socket temp, someone here should know I imagine. I have motherboard monitor, used to use it. I'll give it a shot again.... I ask because some mobos cannot measure the cpu diode temp only the cpu socket temp. The socket usually reads ~10C cooler than the cpu core. If 50C is the core diode temp and not the socket temp then it's reasonable. Well it appears to crash around 70 - which would that suggest? Again this could point to the socket being read, Athlons IIRC are rated for 85C, although I panic if mine hit 55C socket temp. The 50C you mentioned before is under full load isn't it? If you're getting 50C at idle that's too hot. I get room 22C, water 29C and socket 37C (~47C core, mobo can't measure cpu diode) on mobile athlon 2400 (35w) o/c to 2.6 GHz, 1.8vcore, cpu, gpu & northbridge watercooled. Sounds not bad then. I must admit that I'm addicted to overclocking ;o) I'd be happy to get this stable UNDERclocked! Know anything about STOP errors? IRQ not less than or equal. Something or other about a yield was attempted. -- 13 parrots and rising http://www.petersparrots.com 93 silly video clips http://www.insanevideoclips.com 1259 digital photos http://www.petersphotos.com Served from a watercooled dual 3.2GHz silent Athlon with 512GB of watercooled Raid. When shagging a goat you are best taking it to the edge of a cliff because they push back harder. -- Billy Connely |
#19
|
|||
|
|||
hi
i have a watercooled setup also, and your temp. is quite hi (cpu), suggest bad aligment of block maybe. I've got CPUs at 50C, water at 35C, and room at 25C. Does this sound reasonable? to small radiator and maybe to little flow also. watertemp should only be a couple of degrees over airtemp. my numbers: cpu 42°c, water 25°c and air 22°c. cpu is 2500 @ 2300mhz vcore 1.7v 100% load (prime). i got two hardware labs extreme II with two 120mm fans eatch in semiparallell setup. (cpu temp is diod, motherboard said to report higher temp than real, i dont know but mb cpu-backside is cool) IRQ not less than or equal. can be a lot of things, first to come is drivers, may be hardware overheating/oc to much. coco "Peter Hucker" skrev i meddelandet newspsadkyrf5aiowgp@blue... On Tue, 29 Jun 2004 21:02:53 +0100, Apollo wrote: Peter Hucker wrote: On Tue, 29 Jun 2004 19:02:45 +0100, Apollo wrote: What mobo is it? MSI K7D Master with two Athlon MP 2800+ what prog are you using to measure the cpu temp? BIOS, and the supplied Fuzzy logic 4 and PC Alert 3. I don't know anything about that mobo or software, sorry. I use Motherboard Monitor, you can pick the sensors to display, so if you pick the diode and the mobo can't measure it, you get some bizzare reading, I get -46C for my diode ;-) I'd suggest trying it, takes a while to set up but can display and log; all voltages, fan speeds (if connected to mobo headers), temperatures and lots of other useful stuff. http://mbm.livewiredev.com/ If you click on Motherboard list (on the same site), find your mobo and it tells you what sensor chips are used for what, it doesn't show 'diode' next to the sensors for your mobo, maybe it is socket temp, someone here should know I imagine. I have motherboard monitor, used to use it. I'll give it a shot again.... I ask because some mobos cannot measure the cpu diode temp only the cpu socket temp. The socket usually reads ~10C cooler than the cpu core. If 50C is the core diode temp and not the socket temp then it's reasonable. Well it appears to crash around 70 - which would that suggest? Again this could point to the socket being read, Athlons IIRC are rated for 85C, although I panic if mine hit 55C socket temp. The 50C you mentioned before is under full load isn't it? If you're getting 50C at idle that's too hot. I get room 22C, water 29C and socket 37C (~47C core, mobo can't measure cpu diode) on mobile athlon 2400 (35w) o/c to 2.6 GHz, 1.8vcore, cpu, gpu & northbridge watercooled. Sounds not bad then. I must admit that I'm addicted to overclocking ;o) I'd be happy to get this stable UNDERclocked! Know anything about STOP errors? IRQ not less than or equal. Something or other about a yield was attempted. -- 13 parrots and rising http://www.petersparrots.com 93 silly video clips http://www.insanevideoclips.com 1259 digital photos http://www.petersphotos.com Served from a watercooled dual 3.2GHz silent Athlon with 512GB of watercooled Raid. When shagging a goat you are best taking it to the edge of a cliff because they push back harder. -- Billy Connely |
#20
|
|||
|
|||
Peter Hucker wrote:
I'd be happy to get this stable UNDERclocked! Know anything about STOP errors? IRQ not less than or equal. Something or other about a yield was attempted. Can I suggest that you start a new thread, this one no longer relates to your original question, and it could go on for a while. I've suffered with that stop error, do you get a chance to write the error down? or does it just restart? Stop the restart by going into System Properties - Advanced - Startup & Recovery Settings - uncheck Automatically Restart. Could be; Memory - occasional fault, run memtest86 from boot-cd, for a few hours. Bios - update to latest version. Temperature - possible from what you've said already. Drivers - I upgraded my ati drivers two weeks ago and started getting this stop again, downgrade and it stopped the error. Probably not an ati driver problem as no other reports of this error. This stop seems quite common for windows versions using virtual irqs, I have never found a solution, only the suggestions given above. Try what I suggest in that order, I can cause the error by overclocking my memory too much and memtest does indeed report errors at the that fsb speed even with the multiplyer lowered to give a low overall cpu speed. If memtest gives you errors, try relaxing the timings and retesting. -- Ian |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Cooper or gold heatsink? | Peter Hucker | Overclocking | 45 | July 8th 04 02:11 PM |
Took off Heatsink and no thermal compound left any reason this would happen ? | We Live For The One We Die For The One | Overclocking AMD Processors | 6 | May 4th 04 09:26 AM |
Artic Silver 5 vs Artic Ceramique | Nil Einne | Overclocking | 50 | January 7th 04 04:39 AM |
Artic Silver 5 vs Artic Ceramique | Nil Einne | Overclocking AMD Processors | 50 | January 7th 04 04:39 AM |
Faulty fans: dust or bearings? | S.Boardman | General | 37 | October 16th 03 10:34 AM |