IMac G5 Dies When Waking From Sleep

Question

I have a Rev A iMac G5 I bought in 2004. Recently it started doing some weird stuff when waking from sleep. After looking at some of the other posts, I think this is going to be repetitive, but I thought I would see if anyone had any suggestions
After either putting the computer to sleep or letting it go to sleep, it will frequently die after the mouse has been clicked to wake it up. I have tried resetting both the SMU and PRAM as well as repairing disk permissions and repairing the disk. None of this has seemed to help. Occasionally, I get lucky and am able to wake it from sleep normally, but this doesn't happen often. After it dies following the mouse click, I can unplug it and then plug it back in and it will start up normally. Interestingly, it will start up from shutdown normally as well as restart normally. Additionally, as long as it doesn't go to sleep, I think you can typically use it as long as you want and it won't quit.
Today, I took the back cover off and messed around with it for awhile. The diagnostic LED behavior seems odd to me. LED #1 comes on when it is plugged into the outlet. When following Apple's instructions for troubleshooting a computer that won't start up, mine does fine because as I said it will start up normally if you unplug it and then plug it back in. Because of this, I set my computer to go to sleep quickly with inactivity. Although it did wake normally at least once, it died numerous times. The irregularity seems to occur with LED #2. One time it "died", I believe LED #2 actually remained lit and something seemed to still be working based on noise. However, the green light for the "caps lock" button wouldn't come on. Other times LED #2 would go out when it died after pushing the mouse with it asleep. During these times, I could push the internal power button and LED #2 would briefly light up, but nothing more than that would happen. On more than one occasion I would then push the SMU reset button and then push the internal power button again at which point LED #2 would fail to come on at all. After pushing the power button (and occasionally pushing the SMU button again) what seemed to be a random number of times, the computer would all the sudden start up normally again.
My top 3 guesses are a failed power supply, failed logic board, or exhausted PRAM battery (I have never changed it). The first two may be a bit to pricey for me to consider replacing. If they are the problem, I may just settle for setting the computer to never sleep. Does this sound like it could be caused by the PRAM battery going out?
Any advice would be greatly appreciated!
Message was edited by: Joshua213

Answer 1

My bad on not reporting back sooner. Indeed, replacing the PRAM battery did not fix the problem. The situation seems odd to me as I was able to sometimes wake the computer from sleep without it quitting while others times it dies. It does fine as long as it doesn't go to sleep. I can start it up and shut it down with no problems; additionally, it appears to do fine even when left on for long periods of time as long as it doesn't go to sleep. To be correct it doesn't seem to so much be the state of "sleeping", but the act of trying to wake it up from sleep. I don't think ventilation is a problem.
What I have done is just set it to not sleep. We just shut it down once in awhile. Rather than forking out the money, I think I will just use this work around as long as it continues to work.