Slashdot Mirror
Apple G4 Power Supply Woes?
Predius asks: "I, like many others based on forum hopping, have an Apple G4 desktop that has decided it no longer wishes to power up. I've replaced PRAM batteries, verified power supply voltages, pressed the CUDA switch, and done combinations of the above. Trying to start it just results in a blip of the powersupply fan and the power led on the motherboard flashes once. Every now and then it will actually boot, and when it does it's 100% rock solid, will reboot fine, etc. If I power it off, I get to play games again trying to power it up. Does anyone know how to manually start an Apple Gigabyte or later rev power supply? Standard ATX power supplies can be fired up by shorting the green PWR ON lead to ground, but Apple decided to be different. I await the collective's wisdom."
Did you check the blinkerfluid?
I have the same problem with my pc. I've had it since july, and i think it is based on humidity. when i crank the AC up in my room and close the door, it will take about 20-30 mins before it will start. it's not the temperature because it can be 60 degrees in the room, but will still not start if the humidity is too high.
Sounds like this may be a inrush current problem... maybe something is taking extra juice to start up, or the supply can't deliver the initial kick. A multichanel scope would help a lot to see that all voltages are getting to where they need to be in a reasonable time. Otherwise, try removing everything that you can and see if it comes on (drives, ALL ram, usb, pci cards, modem, etc.) - maybe eliminate it that way. Remember, a bad card (one that's drawing too much current) may not cause the system to fail until another good card is inserted and pushes the power supply over the threshold.
HIV Crosses Species Barrier... into Muppets
Have you tried resetting the PMU?
Pop that sucker open (unplug first!) and there should be a little black button on the lower front of the logic board. "Reset pmu" will be printed on the board somewhere near the button. Push it for one second and only press it once. Let the box sit for roughly 5 minutes unplugged and then turn it on.
"My mother never saw the irony in calling me a son-of-a-bitch." - Jack Nicholson
Especially if you are out of warranty. Apple describes it in support article 95039. The most frustrating part is that replacing the power supply is the last step -- they try to replace the CPU before checking the power supply. This means that you have to commit to purchasing a CPU if it is the problem child. My CPU (500Mhz G4) was ~USD900 last time I checked -- more than the value of the the computer at the time. All this is after you pay their diagnostic fee.
Don't get me wrong, Apple Care is wonderful, but once it expires you can be in for some expensive repairs...
I hate to ask the obvious, but since your machine is clearly working improperly in a manner that cannot be due to software, OS installation, disk formatting, etc....
Have you considered taking it in for service?
The friendly folks at the nearest Apple Store will be happy to take a look at it. If it's still under warrantee, or if you've invested in AppleCare, it won't even cost you anything.
"I've replaced PRAM batteries, verified power supply voltages, pressed the CUDA switch, and done combinations of the above."
Look, there are many ways a powersupply can fail and still appear to have the correct output voltages. For instance, the powersupply has a power good signal going to the motherboard. The PS determines whether it's within its specifications, and when it is (some mS after actual turn on) it asserts the signal, and the motherboard comes out of reset. It could be that either the powersupply is not asserting the signal correctly, or that it is still out of spec.
But it could be a dozen other problems. Until you swap it for a known good PS, you cannot know for sure that the PS is good. (other than, perhaps, putting it in another good system and powering it up there)
Here is the pinout of a dual G4 powersupply. I suspect that the pinout is substantially similar. You may be able to force the PS on, but you'd have to find and assert the power good signal (I doubt that the supply doesn't have one - chances are they made this pinout with a multimeter, and there's no easy way to find the power good without a scope or analyzer)
But I suspect that you'll spend less money and time by simply getting a new PS and replacing it.
Please note that powersupplies contain many parts that fail over time. Aside from the obvious fan, the electrolytic caps eventually dry out, causing out of spec voltages, out of spec ripple, and out of spec regulation. You won't be able to see the ripple on the average multimeter.
Also note that computer hardware troubleshooting is a black art, and should only be practised by mages trained in this specialty.
-Adam
I have a G4 Dual 400 (mirrored drive door) and had a similar problem. After numerous games, and since it was still under warranty, I brought it in. Four visits to my local Apple store later, it works. They replaced the power supply, memory, ram, processor, and finally after I demanded it, the motherboard (or "main logic board" in Apple lingo. It was the motherboard.
0 38
Turns out there's a known Apple TechNote on the mirrored drive door boxes that produces a power on, but unexpected results type problem. Here's the technote (iirc) or at least one similar: http://docs.info.apple.com/article.html?artnum=95
I HIGHLY recommend making a list of technotes that directly refer to and explain the problem in detail. Do what you can yourself. Go to an Apple store (make it a road trip if there isn't one near you - I'm in NY, US so it wasn't a problem), walk up to the "Genius Bar" and make them work for their famed support reputation. I've owned macs since sometime prior to the Quadra and when it doubt, Apple *will* save you should you ask the right way. Be nice. Beg. Your warranty is an intangible and can be extended by the manufacturer if it is a known problem and you plead your case correctly.
This is usually related to a bit more than the power supply and is absolutely a known issue.
Good luck. Chin up... and beg if you're out of warranty.
It's not so simple to just drop a new power supply into a Mac. The supply on the G4 Powermac is not, as far as I can tell, identical to or compatible with the standard ATX supplies that x86 machines use -- the part in question is both proprietary and expensive. Your advice would be perfectly reasonable if we were talking about an x86 box, but we aren't, and there's the rub.
I had a G4 that refused to boot, and when I brought it to Microcenter to be repaired they informed me that if it was a bad power supply, the part would be something like $200. When I asked if the $30 ATX supplies they have could do they job, the answer was a clear "no" -- which even they seemed disappointed by. In the end, the problem turned out to be the motherboard, not the power, and replacing that would have been about $500. I ended up getting a new G4 instead of paying that much to keep an old machine alive.
Unfortunately, my new Mac has exactly the symptoms that are described in this discussion: normally the machine runs just fine, but every now and then when I reboot it it will just refuse to turn back on. The power light will throb briefly, the power supply fan will spin up, and the machine makes a few other chirps & squeaks -- but it never gets as far as spinning up the hard drives, and the "Mac chimes" never get played. At that point, the fan willl keep spinning until you turn the machine off, but nothing further will happen. However, if it gets past that point -- if the hard drives turn on, and the Mac chimes sound off -- then the machine comes right up and there are, as far as I can tell, no other issues at all.
The only "solution" I've found so far, aside from not turning the machine off if I can help it, is to just be patient: Hit the button, wait for it to start. If it doesn't, turn it back off & wait some more. Try again. if it still doesn't work, wait longer. It seems to help a little bit if I make sure that the machine is completely stopped (no fan noise, etc) before trying to turn it back on again, but that's no guarantee -- it just helps the odds a bit. It's not just the quick turnaround though: once the machine sat off overnight & still wouldn't start up right away in the morning without going through the try, fail, retry nonsense for five or ten minutes.
Elsewhere in this dicussion it was suggested that it may help to hit the Power Management Unit button on the motherboard while the machine is turned off & unplugged. It's worth a try I suppose, but I'm a little unclear as to how this is supposed to help, or if it will help. Another poster points out Apple support document 95039, which seems closer to the mark and definitely sounds promising. I'll have to try those steps & see if it helps...
DO NOT LEAVE IT IS NOT REAL
:-)
You are, of course, exactly right.
Still, it's not as bad as it used to be. As far as I can tell, the only really proprietary parts on a modern Mac are the CPU, motehrboard, the power supply, and maybe the video card. In the case of the video card, it's only "proprietary" because the systems are Open Firmware based (like Sun, SGI, and maybe some others), and the video card has to be able to talk to the rest of the system -- but it uses the same AGP bus that x86 uses, and my impression is that there's no reason that the "Mac" cards wouldn't work just fine if (say) AMD ever designed an Open Firmware based x86 platform.
Beyond that though, it's all standard hardware. Same RAM, same hard drives, same PCI bus for expansion, same peripherals on the USB & Firewire ports. And so on.
It used to be much worse than that, in the days up to the blue & white G3, with the Nubus architecture, "standard but exotic" SCSI bus for hard drives, etc. Moreover, above the hardware level, they've gotten very good about using standard protocols & formats for nearly all of the system, and have offered up their new developments to be used as open standards (e.g. Rendezvous).
While it's annoying that they can get away with charging so much for the proprietary parts that they still control, at least these parts are generally pretty reliablee & most people don't have any problems. And the rest of the system is just so much nicer than anything that could be run on x86 that I for one am willing to take a chance on what Apple has to offer, knowing that if anything does go drastically wrong, the fix can be ridiculously expensive compared to the same repair on the other site.
Everything is a tradeoff, ya know?
DO NOT LEAVE IT IS NOT REAL
...technical support queries. Pay for your own Apple Care and leave /. for the quality reporting that it provides.
Doesn't it make you feel good to know that our freedoms are protected by politicans, lawyers and journalists.
There was a revision of the G4 in August, 2000. The first ones off the line had power switch problems, giving almost the same symptoms. We had them all replaced under warranty.
One work around, find an older Mac USB keyboard that has a power button on it. Plug it in and see if the machine will start then. If so, you will need to get a replacement power switch.
More info available from ZDNet.
What, me worry?
My Power Mac G4 is on it's second power supply. The problem went away! And it's a Power Mac Gigabit, not Gigabyte.
Don't know about G4s or desktops, but I'm now on my third dual-USB iBook external power supply. First replacement was covered by the warranty, but I just shelled out ~$85 for the second replacement today at the local Apple store. And the main logic board had to be replaced three months in. AND the little twist lock on the underside that releases the battery is broken and held in place with scotch tape. I love this laptop, but the build quality is not good.
just push it downhill, press the power button, and pop the clutch.
After submitting the question, I resumed scouring the web and got a bit braver with my parts...
o nv ersion/G4_AGP_to_ATX_case_pg2.htm has the mods that get you 99% of the way there, you just need to supply 28v to finish the job on any ADC capible G4s. Once I figured this out I had the system booting off a PC power supply (and the 28v trickle from the G4 supply). I bench tested the G4 supply, and sure enough, shorting the power good, which should have fired it up just bliped the power. Based on the symptom set, it seems pretty likely this is whats affecting others. The test is simple, use a paperclip or something similar, short the green power up line to a black ground line, and see if the power supply fires up. If it does, it's the mobo, if it doesn't, its the power supply. No more throwing batteries sacrifical virgins at the problem while performing dark rituals around the CUDA switch.
After reading up on some G4 to PC ATX case conversions, and doing a lil guessing, I determined that the G4 power supply should use the same power up sequence as an ATX. What threw me off is earlier PC ATX power supply to mac conversions had to invert the power on signal, and I incorrectly assumed Apple did the same with the G4.
http://www.xlr8yourmac.com/systems/ATX_G4_AGP_c
And for those curious, the machine I'm building is a FrankenMac, assembled from random bits on eBay and my parts closet. I've TRIED talking to apple stores, but the instant they figure out its not a boxed apple, they tell me to pound sand. I've got $75 into the power supply, so I wanted to be ABSOLUTELY certain it was the culprit before canning it. Unfortunatly, while Apple will offer some simple diagnostic guidelines, they won't give you all you need to do this sort of work outside of an Apple authorized repair depot.