E2160 overclocking, something wrong?

[$0m#0n#]

lo, ve playin about a week with my e2160

Box cooled, highest oc at default (1.3V load) 3150Mhz

idle-34C in coretemp 0.94/speedfan, 15 more in TAT
load-71C in coretemp 0.94/speedfan, 15 more in TAT

the heatsink was just a bit warm on touch (i m sure it was mounted right, cose i mounted it 3 times)

today I bought a thermaltake ultra 120 and a 2000rpm 12cm fan (78CFM)

3150Mhz@1.3V

idle-22C in coretemp 0.94/speedfan, 15 more in TAT
load-58 in coretemp 0.94/speedfan, 15 more in TAT

dont know whats wrong.. i guess the heatsink is mounted right cose of the temp in idle.. bit i dont like the 35 increase of temp in load, i ve been lookin at a lot of reviews and the increase was muuuch lower, when i touch the sink whereever i want its cold as it was in idle

so the question is.. why does it show me, such high temps..

thanks a lot

[$0m#0n#]

after 1.5h is 62C, the heatsink is still cold (in the meanin, that it has the same temp as it was in idle/or off)

also.. can some1 tell me, what temp measuring prog should i trust the most?

[3oh6]

i found the exact same thing with my E2160...hottest running C2D i have ever had. i was seeing 60C~70C under load at anything over 1.36v but had no problem clocking to 3300MHz at 1.36v~1.38v anything higher and temps just got too hot. this is under an Ultra-120 in 22C~23C ambient.

my E6600 L629B is ice cold in comparison to this little bugger with the same hardware used.

BTW, the 15C lower temps in CoreTemp 1.94 are because the tjunction is off by 15C. just add 15C to CoreTemp 1.94 or use CoreTemp 1.95 if it doesn't blue screen your system. TAT is apparently reading the correct tjunction so you can just use that as well...TAT and CoreTemp 1.95 should show the same temps. HTH

[$0m#0n#]

i found the exact same thing with my E2160...hottest running C2D i have ever had. i was seeing 60C~70C under load at anything over 1.36v but had no problem clocking to 3300MHz at 1.36v~1.38v anything higher and temps just got too hot. this is under an Ultra-120 in 22C~23C ambient.

my E6600 L629B is ice cold in comparison to this little bugger with the same hardware used.

BTW, the 15C lower temps in CoreTemp 1.94 are because the tjunction is off by 15C. just add 15C to CoreTemp 1.94 or use CoreTemp 1.95 if it doesn't blue screen your system. TAT is apparently reading the correct tjunction so you can just use that as well...TAT and CoreTemp 1.95 should show the same temps. HTH

the weird thing is.. that i dont believe i get 10C lower temp with thermaltake ultra 120 than with box.. the difference i saw in reviews was like 25C in load

[$0m#0n#]

dont u think there something wrong, if the heatsink is still "cold" i even touched it at the "base" above the cpu.. and its like 30C i guess

[3oh6]

that is exactly what i found too...very little temp being transferred to the heatsink. i thought it was the mount but after many remounts and perfect contact patches each time, i gave up. i am assuming it is the contact between IHS and the die itself. i would have attempted an IHS removal but had a system for a family member that i used it in instead.

[billdavis]

just break out your bargeld and get a better proc and cooling

[$0m#0n#]

just break out your bargeld and get a better proc and cooling

better proc? whats wrong with e2160?

better cooling? whats wrong with the second best air cooler atm?

[Bail_w]

it is the contact between IHS and the die itself.

Same case with my E2140 and Xeon 3210 Quad. I am positive (100&#37 the heatsink and IHS is making great contact but the heat just not transferring to the heatsink. E2140 @ 3.2 ghz with 1.35v (full loaded 60 degree), Xeon 3210 @ 3.4ghz with 1.37v (full loaded 65 - 70 degree).

[$0m#0n#]

Same case with my E2140 and Xeon 3210 Quad. I am positive (100%) the heatsink and IHS is making great contact but the heat just not transferring to the heatsink. E2140 @ 3.2 ghz with 1.35v (full loaded 60 degree), Xeon 3210 @ 3.4ghz with 1.37v (full loaded 65 - 70 degree).

look at it the other way.. it is stable.. do u think.. my comp should be stable running 94C at core?

[Bail_w]

look at it the other way.. it is stable.. do u think.. my comp should be stable running 94C at core?

Well, it is stable, but i am looking to overclock the hell out of it and heat becomes an issues if i use too much voltages.

No, your computer will not be stable at 94C, unless you pass 8+ hrs orthos but your computer will crash before that happens.

[$0m#0n#]

Well, it is stable, but i am looking to overclock the hell out of it and heat becomes an issues if i use too much voltages.

No, your computer will not be stable at 94C, unless you pass 8+ hrs orthos but your computer will crash before that happens.

thats what i m talkin about.. i think.. the mobo is just reporting weird numbers.. cose its highly impossible to have 10C lower temp with thermaltake ultra 120 than with box..

[Lestat]

first of all take off the U120
i can almost guarentee that its not flat.
take a razor and check to see if its flat.
if its not RMA it. or sand it,, its your choice.

mine had a HUGE hump in it.
i ran out of sand paper trying to sand it down.
i was REALLLLLYYYY pissed.

next 90c you can kiss that cpu goodbuy.
90c is beyond any max heat spec of any cpu ever made.
if you havent damaged it by now you will if you keep running it that hot.

you need to contact Intel and tell them whats going on.

also check for flatness on the cpu IHS
if its not flat RETURN IT!
or sand it, its your choice. but sanding the IHS will void the warranty.

[$0m#0n#]

first of all take off the U120
i can almost guarentee that its not flat.
take a razor and check to see if its flat.
if its not RMA it. or sand it,, its your choice.

mine had a HUGE hump in it.
i ran out of sand paper trying to sand it down.
i was REALLLLLYYYY pissed.

next 90c you can kiss that cpu goodbuy.
90c is beyond any max heat spec of any cpu ever made.
if you havent damaged it by now you will if you keep running it that hot.

you need to contact Intel and tell them whats going on.

also check for flatness on the cpu IHS
if its not flat RETURN IT!
or sand it, its your choice. but sanding the IHS will void the warranty.

so both the cooler and cpu had a small "bump" (like the letter U )

so i was sanding it a few hours now.. the temp is like 2C lower

Can some1 tell me, what to trust in? coretemp 0.94 or TAT, cose in bios i ve got 21C, in windows idle 20-21C

[Lestat]

you also have to consider that the 2160 and the mobo you have dontplay nice. in that it cant read the temps right and the temps are totally wrong.
same thing for tat, it may not be able to read the temps correctly.

i went thru and looked at the tat readme and of course it says it only supports Pentium M cpu's.
so..... it very well may not work right.
but those temps are bad, really bad.

if you got any friends with a different baord you might try seeing how their system handles it.

when you sand sand both the cpu and heatsink until they both are a flat copper.
dont sand them and leave any of the silver color showing.
just keep right on sanding them down until they show nothing but copper.

now, the last thing, the intel I"HS is held on by solder. and i have seen tons of cpu's where people removed the IHS and only 1/2 the cpu had solder on it.
what does that do?
it doesnt allow the other core or X % of the core(s) to touch the IHS properly. thus making temps insanely high.

as the conroe chips aged in stepping, ie; my original B1 S5 vs the F steppiong cpus. the F stepping were a much cooler cpu. and i believe that to be intel refining their method of installingthe IHS.

intel chips are infamous for not being flat where as, i have NEVER had to sand any AMD cpu.

i had mentioned RMA'ing the cpu because you now voided your warranty by sanding it and Intel wont rma it. (not that i have heard of)

tell you what, do this:

run it at default speed. and set the voltage to the lowest voltage possible.
my 6600 ES ran 2.4ghz @ 0.975v where stock was 1.23v.
i tried lowering the cpu speed and dropping the vcore but it didnt matter cuz most boards dont go below 0.80v so..
anyways the point is, drop the cpu speed and the voltage as far as you can, and see if temps remain overly high.

also how hot is your room?
do you have a fan on the NB chipset? if not get one. those chipsets get hot.

and again, personally this is just me, but i think the u120 is a pile if crap.
i have tried mine on several cpus and not a single one runs cool.
my 6600 ran hot enough i couldnt overclock it very far because load temps were above 60c and i dont run my cpu that hot.

and how were you sanding? im curious.
the best way is to use the glass from a picture frame and put the sand paper on that then take the cpu and go at it until its down to an even copper finish.

[$0m#0n#]

you also have to consider that the 2160 and the mobo you have dontplay nice. in that it cant read the temps right and the temps are totally wrong.
same thing for tat, it may not be able to read the temps correctly.

i went thru and looked at the tat readme and of course it says it only supports Pentium M cpu's.
so..... it very well may not work right.
but those temps are bad, really bad.

if you got any friends with a different baord you might try seeing how their system handles it.

when you sand sand both the cpu and heatsink until they both are a flat copper.
dont sand them and leave any of the silver color showing.
just keep right on sanding them down until they show nothing but copper.

now, the last thing, the intel I"HS is held on by solder. and i have seen tons of cpu's where people removed the IHS and only 1/2 the cpu had solder on it.
what does that do?
it doesnt allow the other core or X % of the core(s) to touch the IHS properly. thus making temps insanely high.

as the conroe chips aged in stepping, ie; my original B1 S5 vs the F steppiong cpus. the F stepping were a much cooler cpu. and i believe that to be intel refining their method of installingthe IHS.

intel chips are infamous for not being flat where as, i have NEVER had to sand any AMD cpu.

i had mentioned RMA'ing the cpu because you now voided your warranty by sanding it and Intel wont rma it. (not that i have heard of)

tell you what, do this:

run it at default speed. and set the voltage to the lowest voltage possible.
my 6600 ES ran 2.4ghz @ 0.975v where stock was 1.23v.
i tried lowering the cpu speed and dropping the vcore but it didnt matter cuz most boards dont go below 0.80v so..
anyways the point is, drop the cpu speed and the voltage as far as you can, and see if temps remain overly high.

also how hot is your room?
do you have a fan on the NB chipset? if not get one. those chipsets get hot.

and again, personally this is just me, but i think the u120 is a pile if crap.
i have tried mine on several cpus and not a single one runs cool.
my 6600 ran hot enough i couldnt overclock it very far because load temps were above 60c and i dont run my cpu that hot.

and how were you sanding? im curious.
the best way is to use the glass from a picture frame and put the sand paper on that then take the cpu and go at it until its down to an even copper finish.

ve got a 8cm fan@1800rpm on the NB

sanding-was sanding on glass as i read it in faq on vr-zone i think, i bought 2x 600, 1x 1000, 2x 1200, 2x 1500, 2x 2000, when i ended with the 1000, i saw just copper, so i guess thats allright, was sending in the number (8) movement, when i was finishing started with 1500 i was moving it all the time in one direction

[tiborrr]

I have the same problem, guys. My E2160 L736Axxx M0 stepping has about 30+ degC deltaT between idle and load on Scythe Infinity @ 3300MHz, 1.44V (no EIST or C1E). IHS is perferctly even, whole base has a perfect contact with the cooler. I am considering removing an IHS. I have heard the IHS on E4300 is not glued nor soldered to the core (http://forums.overclockersclub.com/?showtopic=71021), there is just a thick slush of thermal paste between the core and IHS. I am wondering if the IHS of E2160 week L736A is soldered, glued or perhaps just held on the substrate by black rubber?

[tiborrr]

Anyone?