![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
thermalright 120 ultra- should it move?
- Thread starter KevinG
- Start date
BillParrish
Supreme [H]ardness
- Joined
- Aug 25, 2006
- Messages
- 7,519
IMO no. if you grab it with a "full" hand and twist hard it will move. But if you take a thumb and forefinger and with small or moderate force it should not.
Intel specs for clamping force for the stock heatsink require between 17 and 40 FT Lbs of clamping force. Rougly this means that if you had a 17 pound piece of steel sitting on a smooth table it would resist twisting as much as the thermalright (some friction aspects ignored in this simple example and note 17Lbs is the minimum). Intel specifies the clamping force to apply a "preload" to the cpu and socket to help eliminate shock damage according to their thermal and mechanical design guidelines. That document is availaible from Intel.com desktop processors Core 2 duo (or quad) technical documents if you want to take a look. But it is a fact that clamping force also has an effect on thermal performance mainly in the effect it has on the thermal compound. You want your heatsink firmly attached is all, it should not wiggle easily but on the other hand once firmly attached additional pressure probally does not help much other than to squeeze out excess thermal compound and if the surfaces were uneven, help ensure/make better contact.
It is also reported and thermalright appears to have confirmed that they now manufacture the base of the heatsink with a "bow", a convex buldge in the middle to conform with the typical concave dip in the Intel IHS. I cannot find where I seem to remember seeing this.
Queries to thermalright elect responses that it will wiggle. It is hard to determine if this is a marketing based reply or an engineering based reply.
I suggest, check both the IHS and the base of the thermalright for flatness. Lapp as needed. (if your IHS is indeed concave and the 120 is convex lapping is probally not needed)
If necessary add washer(s) between the spring and the "ears" of the upper clamp so that when scewed down the springs are forced to compress more. Do not compress the springs completely so that the coils touch one another, that is a spacer, not a spring. Finding stronger springs is also an option. If you really want to get into it, mcmaster.com has precision die springs that typicially can be used for this kind of thing if you can determine what you want, all the specs are on the site.
If your operating temps (coretemp) at idle and load are acceptable I would not worry too much about it unless you are as insane as I am (and thats pretty far gone.)
This is just my opinion and worth exactly what you paid for it.
Intel specs for clamping force for the stock heatsink require between 17 and 40 FT Lbs of clamping force. Rougly this means that if you had a 17 pound piece of steel sitting on a smooth table it would resist twisting as much as the thermalright (some friction aspects ignored in this simple example and note 17Lbs is the minimum). Intel specifies the clamping force to apply a "preload" to the cpu and socket to help eliminate shock damage according to their thermal and mechanical design guidelines. That document is availaible from Intel.com desktop processors Core 2 duo (or quad) technical documents if you want to take a look. But it is a fact that clamping force also has an effect on thermal performance mainly in the effect it has on the thermal compound. You want your heatsink firmly attached is all, it should not wiggle easily but on the other hand once firmly attached additional pressure probally does not help much other than to squeeze out excess thermal compound and if the surfaces were uneven, help ensure/make better contact.
It is also reported and thermalright appears to have confirmed that they now manufacture the base of the heatsink with a "bow", a convex buldge in the middle to conform with the typical concave dip in the Intel IHS. I cannot find where I seem to remember seeing this.
Queries to thermalright elect responses that it will wiggle. It is hard to determine if this is a marketing based reply or an engineering based reply.
I suggest, check both the IHS and the base of the thermalright for flatness. Lapp as needed. (if your IHS is indeed concave and the 120 is convex lapping is probally not needed)
If necessary add washer(s) between the spring and the "ears" of the upper clamp so that when scewed down the springs are forced to compress more. Do not compress the springs completely so that the coils touch one another, that is a spacer, not a spring. Finding stronger springs is also an option. If you really want to get into it, mcmaster.com has precision die springs that typicially can be used for this kind of thing if you can determine what you want, all the specs are on the site.
If your operating temps (coretemp) at idle and load are acceptable I would not worry too much about it unless you are as insane as I am (and thats pretty far gone.)
This is just my opinion and worth exactly what you paid for it.
MisterSparkle
[H]ard|Gawd
- Joined
- Feb 9, 2008
- Messages
- 1,077
Yeah, the mounting bracket for the TRUE120 leaves a lot to be desired.
The best fix is the washer mod. You stick a washer underneath the X brace, which will make the heatsink very hard to twist and move around.
I used a slightly wider washer than the one shown here. For height, try not to go above 2mm, as the heatsink will start to get really hard to mount.
The best fix is the washer mod. You stick a washer underneath the X brace, which will make the heatsink very hard to twist and move around.
I used a slightly wider washer than the one shown here. For height, try not to go above 2mm, as the heatsink will start to get really hard to mount.
MisterSparkle
[H]ard|Gawd
- Joined
- Feb 9, 2008
- Messages
- 1,077
There's a screw that pokes down in the middle of the X brace, so a washer with a hole in the center works well 
How thick is the washer you have?
How thick is the washer you have?
MisterSparkle
[H]ard|Gawd
- Joined
- Feb 9, 2008
- Messages
- 1,077
Another option is to make your own spacer. With some decent quality cardboard, cut out a few squares the size of the heatsink base, with a 10mm hole in the middle. Glue these together and put weight on them until they dry.
MisterSparkle
[H]ard|Gawd
- Joined
- Feb 9, 2008
- Messages
- 1,077
Sounds like a plan