iMac Heat Issues

robertleeking wrote:
Firstly, I make my living doing both programming and board level electronics.
I have worked in this industry for 3 decades and know a great deal more about both electronics and software than I would venture 99% of the populous of this forum could ever dream of.

You may have worked for 3 decades, but your posts aren't doing a good job of backing that up. You keep making comments that show you're either (1) not entirely sure what we're talking about or (2) not reading closely at all. (That, or I guess you could just be a troll, but that fact you have some points from helpful responses tends to rule this out). Again, no one here said anything about the more fragile parts of our Macs running at 80C, so you can go ahead and drop your "fails in seconds," "ambient temperature" and "reference only" replies. We've said the power supply pushes 80 (which, as I said in my first response, seems to be supported by my measured body temperatures around 110F and exhaust port temperatures exceeding 130F). So again, you're either not reading or you just don't get it. I'll leave it to you to clarify, but until you start making statements that show a bit more sense, your "3 decades" of experience don't mean squat to me.

Threads like this are born as we try to determine if Apple is favoring quiet system over cool system by allowing components to push their maximum operating temperatures before implementing appropriate cooling techniques. I have 20+ years in various areas of IT, so I'm aware of the "temperature monitors aren't accurate" argument and it's always put forth by the guys (and now girls) that don't understand the basic electronics behind the reading. But just to humor you, I'll admit my power supply may not be running at exactly 80.00000000000C. But is it within, say, 5%, allowing for some error in the shoftware that reports the temperature? Absolutely. Is a power supply running at 76C (or possibly 84C) something to worry about? And, as you said (" if you've a power supply operating at such temperatures, you've a power supply without any cooling at all"), that's something to be concerned over.

Talk about your "3 decades" of experience all you want, but may personal take is that you're in programming, because you're talking like someone who knows software pretty well, but you aren't exactly "up to speed" on the hardware side (and "board level electronics" can mean anything, so barring further clarification, I'm not particularly impressed). Also the fact that you think your experience exceeds 99% of this forum population shows you're completely out of touch with reality. You're running with some pretty big dogs here, so you might be careful before you start implying that you're the Top 1% here. Your responses to this thread aren't particularly insightful, so I'm just gonna get back to the responses that actually have merit and let you respond however you need to satisfy your own "ego's urgings."

Mechanic man wrote:
So, overheating is certainly a serious topic and the fact that some forum members could convince Apple delivering them a new computer should be enough reason for iMac owners to keep their eyes wide open for such issues. Odd temperatures reported by iStat or Temperature Monitor, even if they are not that precise, could be a good indication something is wrong.

Excellent summary of the thread. Good job. 🙂

Xian,

Let me clarify it for you. I still work as an electronics tech, repairing both Apple and non-Apple hardware, printers, data networks etc. I perform daily, board level repairs, as in replacement of individual capacitors, diodes, ICs, etc on laptops, desktops and printers as needed.

I have seen overheating issues on all platforms, some more severe than others but do not use third party software EVER to make that determination.

Experience has taught me that those software products are unreliable.

Attempting to write my own, has also proven problematic since to determine
the accuracy of the measurements requires both software and direct physical measurement which more often than not is simply not possible physically.

Further, as there are no published specs listing port locations, addressing methods, data structures of the returned data nor even the bit order of that returned data, there is no way to code a single program that can accurately read this data on multiple systems. This is why Apple has system specific tools.

If your machine is actually running at such high temperatures, look for obstructions. Look for power settings that force the machine to run the fans less in favor of less noise and so on. But measure these temperatures with accurate physical tools or Apples own tools if you have access to them, not with third party software that, may or may not, be even close to accurate.

Several have mention fan control software, again, these third party products are not reliable and can lead to failures themselves because, again, the information they act on is questionable at best.

Running a fan when it isn't needed shortens the fan and the power supply's lifespan through drawing more power than the system is designed to deliver at sustained levels and, the fan which was designed for a much lighter duty cycle, overheats itself, wearing out bushings/bearings, lubrications and armatures due to increase heat within the fan itself.

Modern computers are a very delicate balance of tradeoffs. Some extra heat is allowed for in favor of lower energy consumption. Quiet in some platforms trumps performance and oft times component longevity.

Also, in some designs, cooling is approached from an entirely different viewpoint.

In the aluminum unibody Macs, Macbook, Macbook Pro, Macbook Air, iMac 17, 20 and 24 inch models, the entire case is used wether intentionally or not, as a second level heat sink. This is due to the fact that everything that generates heat is connected to this sheet of aluminum and naturally transfers heat quite nicely to it by contact transfer.

Since heat rises, the tops of these systems are naturally warm and even hot to the touch as the case may be. It is illogical to expect otherwise, metals transfer heat, and without something dispersing it, retain that heat longer than plastics do.

This is the very purpose of the fans within the computer, to cool the metals used as heatsinks by dispersing their heat buildup as efficiently as possible.
They aren't there to keep the entire system cool, only those components that must not be allowed to exceed a given temperature metric.

Again, if your system is actually running as hot as you claim, by all means, take it to your nearest service center and have it checked out. Note in advance how long it takes to reach these temperatures, which peripherals are in use and any given software that appears to consistently drive these temperatures up. Supplied with this information, the technicians will be better able to evaluate the situation and determine wether or not there is a faulty component or, if these temperatures are considered normal for that system.

Computer and peripheral service is a time critical endeavor. Both for the end user and the technician. The end user wants/needs their equipment working and returned to them as quickly as possible. The technician is on the clock and must work as efficiently as possible to resolve any issues or lose customers, his/her position or both.

In the case of warranty matters, every manufacturer has specific times alloted to a given repair. Exceeding that time, costs the shop money though the end user will likely never know about it and certainly would not care. But in a very real sense, failing to meet those established re-imbursement specifications can make or break any given service center and like it or not, the bottom line, is far more important than most would want their customers to know.

This is why many manufacturers simply replace entire machines when in fact it may more likely be a simple issue that requires more time to repair than is financially feasable. Labor is still, much more expensive than hardware, at least in 1st world countries.

This is also why most Windows platform manufacturers turn first to factory restoration procedures with software issues. It's simply too costly to support an unknown. Have a problem with your Windows PC, restore it to factory or shipping condition. Problem gone? Good, issue solved. But the issue isn't solved at all, only the circumstances have changed.

I could go on but I'm sure you see the point I'm making here.

Xian,

On the experience issue, I believe I said, I'd venture to say. Meaning, considering the posts, most here are end users not technicians nor programmers. There may be a few of both but the majority, are in the end user category.

Further, if you've never done solder to board level work nor bit level programming for device control etc. even the "technician" side of that group is much more end user experience level as a whole. I don't say this to denegrate anyone, only to point out that many who consider themselves technically adept, are nothing of the sort. Despite their job titles.