The Reports of My Demise Are Greatly Exaggerated

Alright, Mark Twain said it first, and he definitely said it best, but my last post concerning the demise of my long serving Power Mac G5 Quad may just have been somewhat exaggerated.

As promised, I gingerly moved the machine to a safe location, with a lot of cardboard underneath it, and opened it up, expecting cooling fluid to come spilling out. Instead, I found what looked for all the world like a normal Power Mac G5… Other than a lot of dust everywhere, nothing seemed amiss.

When I removed the main fans to have a better look, it became apparent that there was more than just a little dust afoot however – the main intakes to the CPU unit were nearly fully blocked.

Dusty Grill

I got out the lab vacuum and gently vacuumed all the dust off the intake grill, and off of everywhere else that I could reach, and put the machine back together.

Cleaned Grill

With more than just a little hope at this point I restarted the machine. I was not pleased when it booted and the CPU meter showed a whopping 89C, but to my delight, the reading started to drop quickly, settling down to 54C, before starting to climb again. By the time I was done copying off the Happy Macs Gopher Site, and the few other files I wanted to recover, the temperature meter was reading 63C, but it was stable! The fans were not quite full blast, and the temperature fluctuated between 59C and 63C, with no apparent correlation to what the machine was being asked to do.

I know that the PPC970 chip can run safely at 60C, but prior to this whole incident, the machine idled in the 45C to 48C region, with the fans running at minimum, instead of roaring along at 3600 RPM.

So, after letting it sit for 5 minutes or so, to ensure that it would not recover its normal idle, I shut it down. I am now hopeful that the interior of that CPU compartment is simply choked with dust, and that a good cleaning will recover the machine to full functionality. I will hunt down a service guide for the G5 Quad and see what I can do.

In the meantime, I have restored the Happy Macs Gopher site onto my Power Mac G5 Dual, and it is once more on the air. This post is being composed on that very machine.

Thanks to reader Ty, who responded to my last post with encouragement to soldier on with attempted repairs, instead of just harvesting the useful components and putting the empty hulk of the machine out to permanent pasture. I will be in touch Ty – thanks!

I guess if there is a moral to this story it is that “it is not over until it is over”, and this story is not over yet. These older machines deliver great service, but they will fall victim to their age from time to time, and then they need you to pay back their years of faithful service with the care and attention required to get them back on their feet again.

Stay tuned, I will post more as this progresses.



A Casualty on the Front Lines

No, this isn’t a misplaced war correspondent piece! The title refers to my much loved and long serving Power Mac G5 Quad, which until this week has been on the front lines of the internet as the server for the Happy Macs Gopher site.

G5 Quad

Regrettably I went into the lab mid this week to find its fans running at full throttle and the CPU temperature meter registering well over 80C. Amazingly, the machine was still running and still responding, but of course I shut it down immediately.

Thinking that it might just be a software glitch in the temperature control system, I let it cool off for an hour or so and then restarted it. It started up normally, but to my dismay, the CPU temperature meter started to climb immediately, the fans kicked up in tandem with it, and eventually I was right back to where I had started: fans at full throttle and an unacceptable CPU temperature level. I shut the machine down again to preserve what little run time it would have left at those extreme levels of heat, and pondered what to do next.

I have to conclude that the G5’s elaborate cooling system has failed after many years of faithful service. These machines are well known for this problem, but usually the failure is more catastrophic than the one I have experienced, with cooling fluid spilling out all over and staining floors, carpets and anything else in the near vicinity. I really can’t complain I suppose: 2005 to 2018 is a pretty good run for any computer …and it has not been easy service of late. As mentioned above, this machine has been on the front lines of the internet for several years now, serving up the Happy Macs Gopher site 24x7x365.

I will have a good look at it this weekend, and in the short term at least, I will replace it with my original personal Mac, a Power Mac G5 Dual 2.3 GHz. I will have to fire up the G5 Quad for a limited last run, to transfer the Gopher site off of the HDD and onto a backup drive. This will let me restore it onto the G5 Dual and get it “back on the air”. I can only hope that G5 Quad survives this final service without too much (additional) damage.

Looking forward, G5 Quads are becoming hard to find on eBay these days, and so it may be a while before I can replace my failing friend. I doubt that I can actually repair it – I lack both the time and the skills to do so. I may simply have to “harvest” all the useful components from it (hard drives, RAM, video card, etc.) and put it out to pasture permanently. This weekend’s examination will tell the tale.

RIP my long faithful friend!

Recovering from No Startup Chime, Power Mac G5 – The Saga Continues

Some time back (February 2015), I posted a piece on recovering from no startup chime on a Power Mac G5. Researching on Google, it was clear that this was a fairly common problem, and I was pleased to be able to share a solution for it.

G5 Quad

I recently had this problem occur again, once more on my original personal Mac, a late 2005 2.3 GHz Power Mac G5 Dual. This particular machine seems to get into this state once a year or so. This is surprising, since my other Power Mac G5, a 2.5 GHz G5 Quad (also a late 2005 offering), has never once exhibited this behavior.

At any rate, I smiled knowingly to myself when I unexpectedly encountered the all too familiar “no startup chime” situation on my G5 Dual a few days ago, and of course I followed the solution recipe that I had posted in my earlier piece on this problem – I unplugged the machine and let it sit overnight, the objective of which was to cause the on board SMU to reset. The next day, I plugged the G5 Dual back in and confidently hit the power button, expecting it to chime either right away, or on the second attempt. I was more than just a little disappointed when it didn’t do either. It not only didn’t chime, it didn’t do much of anything else at all, except spin up the hard disk and run the fans, both of which were clearly audible.

Maybe I hadn’t waited long enough with the machine unplugged? Maybe the PRAM battery was dead? Maybe I needed to try using the infamous, and usually totally ineffective, SMU Reset button on the motherboard? I walked through all of these potential solutions, but to no avail. The machine stubbornly refused to chime and boot. Frustrated, I dug back into Google and once again went through everything I could find on this problem and its solution. Happily, a pattern started to emerge in my research, suggesting that it could be a monitor or video card related issue. Countless web authors stated that reseating the video card, or simply attaching a different monitor, would miraculously rouse the sleeping beast and return the G5 to a working state.

With nothing to lose and nowhere else to turn, I decided to try this. I attached a different monitor and hopefully powered on the G5. Nope, no difference. No chime, no boot. This left only one path – reseating the video card. Although the concept made no sense to me in the context of this problem, I decided to try reseating the card, an nVidia GeForce 6600. The location of this card is shown in the image below.

Video Card Location, Power Mac G5.jpg

Reseating the card was a LOT more difficult than it sounds, due in no small part to the fact that removal of the card was impeded by a small plastic tab on the corner of the PCI-E slot into which the card was plugged. This tab steadfastly resisted all efforts to remove the card, and then, when I finally prevailed in that task, proved equally effective at inhibiting the re-insertion of the card. After about half an hour of fuming and fussing with it, I finally accomplished a reseat of the video card, wondering whether I had damaged the card in the process of attempting this seemingly nonsensical “fix”.

With no real expectation of success, I hit the power button and to my complete surprise, I was rewarded with a robust chime and a full and successful boot up of Mac OS X. Had reseating the video card ACTUALLY accomplished this result, or was it just a random co-incidence? I honestly don’t know, and given the difficulty inherent in accomplishing the reseat, I don’t plan to experiment with it again until a “no chime” situation occurs in the future. I can report that the G5 has been successfully booting ever since. The no chime scenario has been once again banished. Who knows when it may raise its head again.

My thoughts turn to defeating this issue once and for all. I have long suspected that there is an incompatibility between the nVidia GeForce 6600 video card in the unit and the monitor to which it is attached (a ViewSonic VP2130b), since the monitor will not always detect the video and display anything after a boot up. The same monitor works flawlessly however when paired with my G5 Quad. So, I am going to head over to eBay and purchase the same model of video card that the G5 Quad has (an nVidia Quadro 4500) and swap out the current card for this higher end part. Hopefully, this will resolve the incompatibility issue once and for all, and perhaps the “no chime” issue at the same time.

Quadro 4500

So, to summarize, simply unplugging a G5 that won’t chime and boot, and leaving it for 24 hours or so, still seems to resolve the situation most of the time. In the event that this does not do the trick however, try reseating the video card. It has worked for me at least once! If and when I gather more information on the solution to the “no chime” problem, I will post it as well.

Happy booting!

Living Life the G5/Tiger Way

Regular readers may have noticed that the pace of postings here at the Happy Macs blog has slowed down quite a bit. There is a reason for that. In January, I started a new job, and moved halfway across the country in the process. The Happy Macs lab was completely dismantled and packed up, and remains largely in that state. We have moved into an apartment for the short term, while we decide where in our new location we would like to buy a home.

I have a small “computer corner” in the new apartment and that is all. I brought two prized computers with me to the apartment and set them up in the computer corner: my Power Mac G5 Quad and my Power Macintosh 7300. Everything else is in storage until we buy and move into a new home (with the exception of an incredible Power Macintosh G3 and its accompanying AppleVision 850AV monitor that I picked up a short while ago, and have yet to begin work on – that will be another post at another time).

Power Mac G3 and AppleVision 850AV

Which brings me to the topic of this post. Note that I did not mention any current day computers. My “daily driver” up until the move was my 2012 27” iMac, a 3.4 GHz Intel machine. An excellent computer, but packed away in a box at this point I’m afraid. Before the move, I backed up all the contents onto an external hard drive and brought that with me. When I set up the computer corner, I restored that backup onto my Power Mac G5 Quad, and for the next six months or so, it will be my daily driver. This post is being composed on it right now.

G5 Quad

So, for the next many months, I will be living life the G5/Tiger way. It is almost like stepping back in time to 2006 when these machines were the shiny new state of the art. Back in 2006, the Power Mac G5 Quad was a kick ass machine. Know what? It still is. Granted, I have accelerated this machine a bit. The boot volume is an SSD, and the main disk is a fairly modern high speed 7200 RPM drive with a whopping 64MB of onboard cache. The computer itself is equipped with 8 GB of RAM, and sports the top end video card of the day in 2006, the nVidia Quadro FX 4500, itself equipped with 512 MB of video RAM.

nVidia Logo

So, the machine packs a punch, but it is still a 2.5 GHz PowerPC G5. By today’s standards, it would be considered pretty low powered I am sure. However, in daily use, I can honestly say that I don’t really notice that. In fact, the opposite is true. The machine feels crisp and fast and I can do everything on it that I was doing on the iMac before (with the exception of managing my most recent iPod, the 160 GB iPod Classic, which I purchased just before Apple discontinued them). As I have often opined, “older” does not equal “obsolete”. This machine is fully up to the challenges of the day, and I am thoroughly enjoying working on it once again.

In the meantime, as we get fully settled in, and I get fully up to speed on my new job, the pace of posting should start to pick up here again. There is lots to do! I am finally in a position to load up my Gopher based vintage Mac software repository and of course there is the Power Mac G3 and AppleVision 850AV to work on … All of this and more will be tracked here in the Happy Macs blog. Stay tuned!

Recovering from No Start Up Chime, PowerMac G5


Synopsis: no start up chime from your PowerMac G5? Suspect the SMU. Unplug your machine, leave it sit overnight and try again the next day. The first restart attempt may fail, but the second should work. Good luck! This post explains the details behind this prescription.

My last two posts have dealt with my attempts to recover bootability (if that is a word!) to my late 2005 Power Mac G5, 2.3 GHz dual core (not dual CPU). The symptoms the machine was exhibiting will be familiar to regular readers of this blog: no start up chime and no signs of intelligent life after powering on the machine. The fans would come on, the disks would spin, but nothing else would happen. Eventually, the fans would ramp up to full speed and stay that way.

All along, I had been assuming that the issue lay with the RAM upgrade I was attempting, but as chronicled in my last post, this was definitely not the case. With that safely out of the way, something told me the issue had to be power related. The behavior was too unrelated to anything else I could determine, and did seem at least temporarily correctable by  simply unplugging the machine, leaving it to sit and then trying again. In fact, I got the machine “back on the air” today after just such a sequence of events. For whatever reason, just like the last time, it took TWO restart attempts to get the machine to “catch” and boot, but just like last time, on the second attempt  it roared back to life and has been stable ever since.

So, what could it be? I recalled that the PowerMac G5, like many of its predecessors, has a power management unit (PMU) on the motherboard, sporting the famous CUDA (Capacitive Unit Discharge ASIC) switch, which allows an externally initiated reset of the PMU. On the late 2005 PowerMac G5, the PMU has been replaced by a newer, but equivalent, microsystem, referred to by Apple as the SMU (the System Management Unit). It still includes the famous motherboard mounted external reset switch:


and Apple helpfully provides full details on how to use it at this page:

A good resource that I have found related to the PMU/SMU and the CUDA reset button is the following page:

The above page contains not just information about the PMU, SMU and CUDA, but also provides links to pages containing instructions for how to reset the PMU or SMU for just about every PowerMac based machine that has one, starting with the eMac and ending with the XServe G5.

Going back to Apple’s page (above), there is a note on it that says that a reset of the SMU may be helpful for many reasons, one of which is the nebulously worded “not starting up” condition. Well, that was definitely the condition I was experiencing, and so this seemed a logical root cause. Somehow, the SMU had to be getting into a bad state and not allowing power to be applied to the CPU after a restart. This fit all the symptoms I was experiencing, and so it was definitely worth digging into.

Regular readers of this blog will recall that I have in fact been faithfully resetting the SMU as I tried various solutions, so why hadn’t this worked? Well, per Apple’s page (above), it turns out that I was doing two things wrong. Firstly, for the late 2005 PowerMac G5, the machine should be plugged in (not running, but plugged in) when you reset the SMU. Secondly, if instead you are going to follow the “unplug the machine” approach, you need to leave it to sit for at least two minutes (…and I am guessing that quite a bit longer is needed!). I was not doing this either. I was opening the machine up, making my changes, pressing the SMU reset button, and immediately closing it up again and applying power. What I wasn’t allowing to happen was a discharge of the capacitor that seems to provide trickle power to the SMU … at least, that is my hunch anyway.

With the above information in hand, it starts to make sense that when I would unplug the machine and leave it overnight, it would accomplish a reset of the SMU, and the machine would then boot again. Why it always seems to take two boot attempts to get it to “catch” and perform a successful boot is a mystery I still haven’t solved, but I think I can live with this relatively small anomoly! 🙂

So, in short, if you have a PowerMac G5 that is powering up but not providing a start up chime, and is not booting, the PMU or SMU (depending on the model of your PowerMac G5) is a likely cause. The field proven approach for my machine has been to unplug it, leave it unplugged overnight, and then reapply power the next day and attempt a restart. If the first restart attempt doesn’t succeed, try a second. If THAT doesn’t succeed, then perhaps this is the wrong solution for YOUR machine’s problem.

This approach has now worked for me twice, and in fact, probably several other times when I have managed to briefly get the machine going again.I am postulating that the root cause is a firmware bug in the SMU, but I have been unable to find any indications on the web that this is a known problem.

In the meantime, my PowerMac G5 is back in the land of the living, and just to make that point, this post is being authored on the self same machine! I know that there seem to be a lot of people out there with similar problems, and I hope that this post may help a few of you to get your PowerMac G5 back on the air.

A Bad Apple Recovers (Briefly)


The saga of the “bad apple in the bunch” continues. Yesterday, after letting the machine sit for a good part of the day, sans both its PRAM battery and its mains power supply (no battery and unplugged) I attempted to restart it. As expected, it failed to restart. Acting on impulse, after reading articles by some people who said that reseating the video card seemed to have an effect on this problem, I unplugged the monitor and attempted the restart again. It succeeded! The machine fired up with no issues at all. I plugged in the monitor as it booted, and the network connection shortly thereafter and all seemed well.

Alrighty… now why had this worked? Was it just a random event, or was there something to the removal of the monitor? I wasn’t sure, but I wanted to do some testing before I tried restarting it again, lest it not grace me with a clean boot the second time around. I pulled up “About This Mac” and was pleased to see the below:

About This Mac, G5 2.3 GHz w 4.5 GB RAM

The full compliment of 4.5 GB of RAM was showing up, so that was a good sign. I ran System Profiler and went to the Memory “tab”, where I could clearly see that all the RAM sticks were of the correct type and that all were working correctly.

Memory Status, G5 Dual

So, the RAM was not the issue after all. It clearly is good. Something else has been bedeviling the boot process, something which I have not yet uncovered. I will now endeavor not to touch those RAM sticks again! 🙂

Well now, since the machine was up and running, I decided to apply a few of the other “solutions” I had read of, hoping that any one of them might solidify the newfound “cleanly booting” status of the machine. I ran Disk Utility and checked the hard drive, which came back perfectly clean, and then repaired permissions, which also went off without a hitch. Finally, I pulled up OnyX and ran all of the maintenance scripts (monthly, weekly and daily), and as a final step, executed all of its “Cleanup” steps, which flush caches and the like. All actions ran to completion successfully.

With the machine in as clean a state as I could bring it to, I gingerly restarted it, expecting… I wasn’t sure what to expect. Happily, I was greeted by the happiest sound in all of Mac-dom, the start up chime. The machine roared back to life again, and continued to do so on repeated restarts. Finally, being the risk-taking sort, I executed the Shut Down command vs. the Restart command that I had been using up to now, and then started the G5 up again. All was well. The machine seemed 100% solid again, for no known reason.

The PRAM battery was still out of the case and so today, I decided it was time to finish this exercise and put it back in. I powered off the machine, unplugged it, opened it up and put the PRAM battery back. This battery, you will recall, is brand new, having just been taken out of the package and inserted into the machine a day or two ago. I checked the Power Mac G5 specs, to be sure I had the correct type of battery, and I can confirm that I do – it is the regulation CR 2032 coin style battery – and that I had inserted it the right way up in the holder.

I closed the machine back up, having done nothing to it but installing the battery, applied power and restarted it. Of course, it would not boot. Of course! 😡 Interesting… OK, I unplugged it, opened it up again, removed the PRAM battery, closed it up again, plugged it back in and retried the boot. This was identically the state the G5 was in when I started today, so it should work again, correct? Nope! It wasn’t having any of that. It would not boot. I repeated the “remove the monitor” step from above and tried again. Nope. No go.

So, here we are again. The machine did recover itself yesterday, after a long unpowered pause, but is now once more cold and silent after the simple step of re-inserting the PRAM battery.

I have put the PRAM battery back in, and left the machine unplugged. Tomorrow, after another suitable powerless pause, I will try it again. If that doesn’t work, I will remove the PRAM battery as well, allow it to slumber for another day, and then try again. I’ll let you know how this goes.

Another Bad Apple in the Bunch

PowerMac G5

The first Mac I bought for myself was a late 2005, 2.3 GHz  Power Mac. I used this machine as my main computer for two years, before replacing it with a 3.2 GHz Mac Pro (which itself has since been replaced with a 3.4 GHz 27″ iMac).

The G5 was delivered with a paltry 512 MB of RAM, in which Mac OS X Tiger somehow managed to limp along. Of course, I had ordered a RAM upgrade along with the machine, and so on the day of delivery I bumped the new G5 up to 2.5 GB, and so it stayed throughout its full time as my main computer. 2.5 GB seemed like a LOT of RAM in those days!

When I started restoring vintage Macs a few years ago, I decided to do something I had always wanted to do with my now retired G5 – add another 2 GB of RAM. The newly revised machine would then be pressed into service as a household file server and would thus gain a second useful life.


Adding another 2 GB of RAM should have been an easy task, but it was not. After I added the two new memory sticks, the machine would not start. The fans turned on and I could hear the disks spinning, but there was no start up chime, and no outward signs that the CPU was running at all. I had taken all the normal ESD precautions, and so I was pretty sure that the RAM was healthy. I was therefore totally perplexed by the sudden and complete absence of any signs of intelligent life from the newly upgraded machine.

As I tried remedy after remedy, at one point I got a start up chime and the machine sprang into life. “About this Mac” showed that it now had a compliment of 4.5 GB of RAM and all seemed well. I was ready to dismiss the whole episode as some form of upgrade weirdness, and assumed that a reset of the CUDA or a reset of the PRAM would clear this up once and for all. Confident in this belief, I powered down the machine to take these steps.

That turned out to be a bad move. Once again, this clearly healthy machine went silent and cold on me. No start up chime, no signs of intelligent life. I did all the usual things one does in circumstances like this, such as backing out the RAM upgrade, changing the PRAM battery, attempting a PRAM reset, resetting the CUDA and so on. NOTHING worked. The machine just sat there, fans and disks turning over, but without any other signs of life.

After a great many unsuccessful efforts, I eventually gave up, deciding that I would have to take the G5 to the Apple Store for repair. At the time I gave up, I had the new 2 GB of RAM out of the machine, and this was how things stayed until a few months ago. My (by now retired) Mac Pro took on the G5’s intended job of household file server, and all was well enough. The G5 sat silently gathering dust in a corner, awaiting its next chance at life.

A few months ago, I decided to have another go at resurrecting it. I plugged it back in, changed the PRAM battery (again!), pressed the start button and sat back, expecting nothing. Much to surprise, it gave out a robust start up chime and took off. Expecting that this was a “one of”, and that it would never restart again, I shut it down and restarted it. Up it came again, full of life and vigor. Again and again this scenario repeated itself. The machine seemed to have cured itself during its long wait, and with some pleasure I retired the Mac Pro once more and put the G5 into its originally intended role of household file server.


Yesterday, I messed that all up again. Those 2 GB of RAM have been sitting out on a desk near the G5 ever since I gave up on my original effort to install them. Reasoning that the machine was now clearly starting very reliably, I decided to finally accomplish what I had set out to do years ago – increase the RAM to 4.5 GB. You can probably guess the rest. I am right back where I started. The machine once again sits cold and silent, resolutely resisting all my efforts to get it to boot.

I have done all the things I can think of to overcome this unfortunate circumstance. I have backed out the RAM upgrade. No joy. I have reseated all the RAM. No joy. I have taken out almost all the RAM out, leaving only the original 512 MB that came with the machine. No joy. I have reset the CUDA. No joy. I have attempted both PRAM and NVRAM resets. No joy. I have tried all nature of start up key combinations to coax the machine into doing something … anything! No joy. I have even checked all the internal connections. No joy. NOTHING will induce the G5 to rejoin the land of the living.

I have searched the web looking for other solutions, but have found none. What I have found is that this is not an uncommon experience. There seem to be a lot of folks out there who have 1.8 GHz, 2.0 GHz and 2.3 GHz Power Mac G5s that just mysteriously stop booting. Some people report that, like me, after some apparently unrelated step, the machine suddenly springs to life again. It is nice to know that I am not alone, but regrettably this is not of great assistance in resolving the matter!

To that end, I am down to repeating the apparently successful prescription I stumbled upon last time: leave the machine sitting dormant for some time, and then try it again. Am I draining a cap somewhere? Who knows what made this work last time, or if it will work again this time. I hope that it may some day. I have a certain fondness for that G5.


In the meantime, if you should happen to know any specific cures for the “G5 no chime syndrome”, I would be delighted if you would comment on this post and suggest them. I am not sure where to turn next.

As I said once before, even Apple can’t get it right all the time. There has to be a bad apple in every bunch and I seem to have one of them!