Q: Late 2006 iMac, Lines, Kernel Panic, Freeze, Reboot, Restart, Serial W8 ?

The last PowerPC G5 was built in 2005, all of the white models built in 2006 are Intel iMac's not G5's.

It is the "How to Identify your iMac" support articles that changed in 2007 with the introduction of the Aluminum models.

http://support.apple.com/kb/HT1463

http://support.apple.com/kb/HT1758

Yes, sorry, I was taking apart a G5 today and had my brain stuck in G5 land.

My November, 2006 20" iMac is dead. The logic board to blame. I've considered it a lemon all along. I've been a "mac" for ever . . . first one the 512 (thats k). All my macs have worked well until this one. It was in the shop every few months. I feel I got the run around from mac, and now the repairman says they're all dying. Are we just going to accept this?! I feel like a friend has lied to me. I wonder if I should go buy another brand that will cost half as much and last twice as long.
Is Apple listening?

… and now the repairman says they're all dying.

FWIW, as a former electronics repair shop owner, I suggest you not take that comment too literally. Techs don't see many units that don't fail & usually have no idea how many have been sold, so they often have a very distorted idea about the real failure rate of popular products.

It is a lot like what sometimes happens in Discussions & similar sites: users see reports of issues with some model or software product & assume it is representative of what all users of those products see. But without considering how many units are in use, it isn't a safe assumption. It is sometimes called the hospital effect -- if you try to gauge the general health of the population from what you see in hospitals, you are likely to conclude everybody is sick.

Pfff. Seems to me there are more trees falling. But you can always look the other way. And there are still trees standing. But when you see those trees falling. You are too late to solve the problem!

If 2 out 10 people decide not to buy the expensive tree you eventually you lose more profit. And their will be an other brand tree in the market that fills this empty space of those 2 out 10 people. And when those 8 people see that those other brand trees are standing longer they won't think twice!

@R C-R You're doing your pretty d... best to talk something good out of this. Like there are many sold that works great... what do I have on those words, _mine isn't_! Like many other iMacs over here.

@R C-R Do you have a solution to solve mine issue with my iMac?

You're correct, older computers fail at a higher rate then newer ones. However, most people don't expect to throw away their (relatively) expensive iMacs or incur $900+ repairs within three to four years.

Ironically, two of my own iMacs AND my neighbor's started exhibiting this exact problem over the last two months (all purchased within four months of each other). You still don't believe this is either a design or manufacturing defect? All these iMacs just happen to be failing at about the same time with exactly the same issues? And if we assume it is "normal failure rate", is that the message Apple wants to convey to the market? "Please be aware that you should expect a major and costly failure of your new iMac within three to four years." I expect to replace a hard-drive from time to time, but I do not expect to replace a motherboard in three to four years because the GPU is soldered directly to it. Let me put it this way, if the ten PC's I built three years ago as a render-farm all had exactly the same motherboard component fail within a few months of each other - I consider that a defect in one way or another.

For all the hundreds that are reporting this problem there are many thousands who don't get on discussion boards and do the research to find out if this is an isolated problem - clearly it is not.

All I'm saying is you can't just look at just the failed ones to decide if a design or manufacturing defect is involved. At the very least, everyone reporting a failure here needs to report the first 5 (& only the first 5!) characters of the serial number. That encodes the place & week of manufacture of the iMac, as described much earlier in this topic. With that, it may be possible to trace the problem to a factory and/or time of manufacture.

However, you still have to consider that at least a hundred thousand late 2006 iMacs were manufactured, so a few hundred reports still may be too small a sample to show anything significant.

I hear what you're saying, but when two iMac's I own (bought at the same time) and one a few houses down (bought within a couple of months of mine) all exhibit exactly the same problem at approx the same time, that's an extremely significant failure rate and a large enough sample given one street in a small neighborhood. It's not a big stretch to extrapolate that out to a much larger sample.

If the product we were talking about was a car instead of an iMac, and all three had the rear axle fall off at the approximately same time, would we be asking if this was a design defect?

And I understand the point about the serial numbers, but that doesn't preclude a design defect as opposed to a localized manufacturing defect.

1. It is a huge stretch to extrapolate three units as representative of a much larger sample!

2. A design defect would affect every unit made using that design. A manufacturing defect would affect only those units manufactured with that defect, whatever it might be.

If this much is not self-evident, there is no point discussing this further.

RCR & I have come down on opposing sides of the iMac operating temps issue, both of us presenting reasonably relevant info to back our positions IMO, for what can only be speculation due to the lack of hard data WRT failure rates. One could see some of this by searching for threads to which I've posted this year. *

There are at minimum hundreds of reports across the WWW of graphics system failures for the IMacs we are talking about, with new reports coming in almost daily, and these numbers are surely not representative of all such failures.
Given this, I'd be curious of the answers to these questions:

*1) Is it clear the graphics subsystem is the most likely to fail / weakest link in these units?*

*2) Is it a good bet that the designed operating temps, often just within component's max ratings, have contributed to such failures (as most report lowered symptoms through raised fan speeds/lowered temps)?* *

*3) If you knew someone w/one of these units, who often stressed the graphics subsystem (e.g. video games, watching movies as encouraged by Apple), would you recommend they observe their temps (e.g. iStat Pro) and consider installing 3rd party util to lower temps if found high?*

_If your answer is "Yes" to any of these questions_, esp. the last one, and its hard to imagine one familiar w/these reports not to answer so, _then doesn't that support the position that Apple has not designed these units (e.g. firmware's default fan speeds in relation to temps) to cope properly with their intended purpose?_
Where "cope properly" = fully expecting greater than a 4-year service life (as mac owners have come to expect of past models)?

Note there is no recourse for owners of 17" and 20" of these models but cost-prohibitive logic board replacement. Going against Apple's previously earned customer loyalty & any "green" policies they say they support.

* http://discussions.apple.com/thread.jspa?messageID=11912045&#11912045
http://discussions.apple.com/thread.jspa?messageID=11911516&#11911516
http://discussions.apple.com/thread.jspa?messageID=11946131&#11946131
http://discussions.apple.com/thread.jspa?messageID=11627340&#11627340
http://discussions.apple.com/thread.jspa?messageID=11912045&#11912045

Just to be clear about it, I'm not on any side of this issue. That's because there isn't enough data to do anything other than guess about it. You can speculate all you want about how many unreported failures there are, how representative the reported ones are, & how many are still going strong, but all that is pointless without the data to back it up.

It should be obvious that if the cause is a design defect then all units of about the same age will fail after about the same amount of use. That isn't much, but it is the only thing we can assume with reasonable certainty. The numbers don't seem to support this right now but maybe over time they will. Likewise, if enough users report the first 5 characters of their S/N's, maybe a pattern will emerge that is suggestive of one or more manufacturing defects affecting units made in different places at different times. But there aren't enough users reporting that to say if that's the cause or not.

If you are an affected user or know one, the most constructive thing you can do is post that S/N data. Otherwise, nothing will change no matter what questions you pose.

R C-R wrote:
It should be obvious that if the cause is a design defect then all units of about the same age will fail after about the same amount of use. That isn't much, but it is the only thing we can assume with reasonable certainty.

Not really... You can't say a design defect will cause ALL units to fail. The nature of this problem seems to be heat related. The heat environment experienced by all units, from Green Bay to Phoenix, isn't the same, nor are the GPU stresses based on how the iMacs were used. Even if you could establish identical use and environment histories, it's still not a scientific certainty that ALL units would fail within 6 months or a year of each other.

A design defect may ONLY suggest that ALL units are potentially at risk of a failure, not that they ALL WILL fail.

RCR - To me your posts related to this issue of graphics system failures seem to counter, question or undermine any/all hypotheses concerning what their root cause may be. Or how their frequency of occurrence may be/may have been reduced (e.g. perhaps through more aggressive fan-speed reaction to heat). Sorry that this lead me to believe you'd chosen "a side".

Not surprising, but still disappointing, that you choose not to express your opinion/best guesses through answer of any of the questions I posed.
Contrary to your opinion that nothing would change, I think many here recognize your technical background & would value your answers, esp. if given to the last question, on what their best immediate course of action might be.

My best guess, from the many reports that boosting fan speeds lessens symptoms (like reports I've linked to & posts in this very thread), is that doing so could very well extend the useful life of these systems. I'd not want to place one more Mac user's unit in jeopardy, if using it the way I described, by not making any recommendation WRT temperatures in such cases (e.g. effectively recommending inaction even for units out of warranty showing 1st signs of artifacts?)

We know that trying to collect the potential thousands of ser no roots for affected units, while logistically unlikely to happen (affected units discarded, many go up on ebay, many go unreported), is also unlikely to change anything. As has come up, we are usually talking about units > 3 yrs old here & the case can be made that we expect such technology to be replaced/disposed of in that time-frame. Perhaps the best that can be hoped for is that eventually, after many more of these units become affected, a large list of affected units ser nos could be forwarded to Apple and maybe they would do something nice. Quite unlikely too as if Apple was to do something "nice", like $100 discount on a replacement unit, it would be akin to admitting a defect.

Having given the GPU-crippled unit to someone else to use, I'm awaiting the ser no for it so I can post it prefix here. That unit has elevated fan-speed reaction to heat through use of Fan Control & that's the only thing keeping it usable.

The first five characters of my late 2006 iMac are QP723.

Pixel lines appear across the screen from time to time, several stalls per day as a rule requiring restart, apps refusing to open with icon bouncing in the dock for 60 seconds yet not opening requiring restart because force quit is ineffective at times, keyboard viewer doesn't load and by hovering over far right side of the menu bar causes spinning color wheel to appear requiring restart.

I trust this helps.

Not really... You can't say a design defect will cause ALL units to fail.

True, but one would expect the majority of them to fail within a similar time frame, at least for units used in typical environments. But the point is that even this, the most reasonable of assumptions, is not all that certain, & all the others even less so. Without much data to work with, that won't change.