So what's with the little rubber/cloth looking pad between the heatsink and the CPU on the 1400? Wouldn't that insulate the CPU from the heatsink?
I've been thinking about machining a whole new heatsink out of copper to improve my chances of overclocking this puppy. What about just replacing that pad with a bit of copper plate and some thermal goop?
You could probably do a better heatsink than that pad.
Watch out for changing the shape of the heat shield. I had a lot of trouble with the shield on my former 1400 shorting out on the memory expansion card.
Thermal pads suck, but they are also often gap fillers.
when i had my 1400 i kinda put a load of thermal paste on it...
That's not a good alternative as it dries out and shrinks.
Yeah, but that pad looks to be at least 3mm thick. Ima try that copper thing.
While we're on the subject, anyone got a spare 1400 mobo I can dick around with?
decided to use the obviously inadequate thermal pad. I imagine you can easily correct their lousy design decision with a manually carved-out chunk of copper.
Err, or not.
Is there some reason to think the thermal pad isn't up to its task? Have you tried overclocking it and find it gets too hot for stability? Seems like you are putting the cart before the horse in this case.
Thermal pads are pretty efficient at transferring heat from the processor to the heatsink, primarily because they can maintain complete contact between the two surfaces. The thermal pad which you intend to replace was chosen for any number of excellent reasons and that's just one of them.
(rant)
As usual with this-sort-of-thing, I'll argue that Apple's engineers generally really do know what they are doing!
Maybe a solid copper heat sink would do a better(?) job, but unless there's some reason to believe the original solution isn't good enough, why do folks think that their half-thought-out home-built solution will be superior? I'm not necessarily referring to the topic of this thread in particular, but rather to re-jigging well-thought-out engineering with half-baked kludgey solutions.
I'm always interested to see 'puter cooling mods, but almost invariably the perpetrators of such hacks appear to have little idea of the complexity of such systems. It's relatively easy to direct cooling to one component or another, but unless one has the knowledge and training to analyze the whole picture these mods can seriously detract from overall cooling functions.
An excellent example (and sorry to jump sort of OT) is the PM G4 Quicksilver's processor cooling fan. I've read any number of complaints about how it 'recirculates' hot air and thusly fails to provide optimal cooling. What those folks apparently fail to notice is the design ensures a supply of cooling air even if the air intake gets clogged or blocked. Not a trivial design decision by any means and a critical part of the overall cooling design which would prevent the processor frying itself in the event.
(/rant)
Ahh, heck, I've done tons of similar crap, just . . . well, just because. So would a solid copper heatsink do a better job? Perhaps, if properly done. However, I'm just not convinced in gratuitously modding cooling functions, especially before the original has proved inadequate.
Dan K
Thankyou for your well-thought out answer Dan.
Originally this wasn't about oc'ing, but about running lid-closed in the cabin of my air-con-less truck as an MP3/CD deck. Australian summers are mean.
The ideas I've had so far include; a complete new copper heatsink matching the dimensions, screw holes etc of the existing aluminium one, leaving the thermo pad in place; and the above thoughts about replacing the thermo pad with copper. Less drastic thoughts include adding vent holes to the video port cover and/or the PCMCIA doors - with dust filters of course.
Then there's the more exotic (insane?) ideas, like extending the heatsink out the video port to a finned section, adding a tiny fan (from a dead P-III laptop), liquid cooling, heatpipes, BookCover heatsink etc etc etc.
I was always under the impression that getting copper (or silver or gold
) closest to the heatsource was a Good Thing. I can see from your answer that incomplete surface mating is at least as big a problem as what metal you're using.
If I was to attempt to replace the thermo pad, I would use copper sheet of an appropriate thickness, not just any old chunk (my housemate suggested grinding down a copper penny, but even I
sensed a flaw in that reasoning). Lap both surfaces and use thermo paste both sides. But after reading your answer I am more inclined to leave the thermo pad in place whatever else I might do.
Would I be right in thinking that the thermo pad is a bit more forgiving of mechanical shocks to the CPU than direct metal contact might be?
Is there a simple way of getting accurate *relative* CPU temperature readings out of a 603e?
Are there other components in the 1400 that I should take note of for heat issues?
Greetz, all - I am new here.
It's not so much a question of whether or not Apple engineers knew what they were doing, many 1400 users run with CPUs which were never even imagined by the engineers.
I find this topic interesting, as this week I have been considering making changes to the heatsink as well. My 1400 runs a NewerTech G3 215 (or so). After an hour of use the laptop is about 100 degrees. After two hours about 120 degress… It still runs great, but that's damn hot. Some kind of heat dissapation is definately necessary!
Metrophage
i put it around the pad.