RIP: Another one of my conquests.

by **iMac600** » 30 May 2012, 15:15

It seems that every machine I liberate ends up dead. I had my Power Mac G5 with its ultra-warpy board. I had my PowerBook G3 Pismo.

Now I have a MacBook Pro (Early 2008 15-inch) that I acquired from my previous employer. 2.5GHz Core 2 Duo, 4GB Memory, 512MB GeForce 8600M-GT, 15-inch Glossy display. Unfortunately while it was an ultra-healthy machine with a pristine case and a replacement logic board installed last year, the replacement logic board died a very quick and sudden death. Even worse, it's not the nVidia GeForce 8600M GT that so many of these machines fail from, it's a completely different issue entirely. The unit gets power when the power button is pressed, lasts about 1 second and then turns back off again.

Apple denied any coverage on the unit when I spoke to them a few days ago. I can't honestly think of any way to return it to a functional state either. Replacement logic boards cost anywhere between $700 to $1200 from Apple. Even a $499 board from a third party retailer is too much for a board that has an unusually high failure rate.

Another Mac destined for the scrapyard.

Feeling quite sad. :'(

Regards,

Mic.

by **Trash80toHP_Mini** » 30 May 2012, 16:03

How's the availability situation for broken LCD/broken hinged PartsBooks down under? Here in the USA, they're sometimes available at reasonable prices, from those not "in the know" about the value of the undamaged parts.

Sorry about the experience and sadness.

jt

by **uniserver** » 30 May 2012, 16:12

try pulling the battery out, maybe its shorted?

i always like to try simple things first. also check the power adaptor on another machine …
maybe its the power brick? They are known for going bad.

try to stay positive

by **iMac600** » 30 May 2012, 17:01

I've done everything to this machine. Had the logic board out, tried another Left I/O board, been probing around and looking under magnifying glasses for obvious shorts or damaged traces... absolute nada. Battery is good. Power adapter is the same one I use to test 20 or so machines a day on the service bench, so I can guarantee it works correctly.

If I could find another MacBook Pro for parts, or even just the logic board itself, i'd be all over the idea. It astounds me how well kept this machine is, other than the logic board of course. Barely a scratch on it.

(Oh, and for what it's worth, I have not and will not give up on the Pismo mentioned in the first post. I almost have it working. One more part to go.)

by **Theretrogamingroom** » 30 May 2012, 17:23

Just find another parts machine or wait a while for prices to go down. Might be $700+ now, but it will probably be half of that in a couple of years or so.

by **uniserver** » 30 May 2012, 17:43

ok then i blame the BGA , the graphics, flood the bga with flux, and re flow it… i bet there is a little piece of solder in the array causing a short.
its pretty much the only thing with those that goes bad.

that is the one thing that really gets me with apple stuff… the MBP/iMac g5 and intel/Powermac G5, etc They let this stuff get way too hot!
all because they don't want to hear a little fan noise!

by **TheIanMan85** » 30 May 2012, 17:48

Definitely a bummer. I agree with holding onto it until you stumble into the right parts, especially since you mentioned what good shape it is in physically. You never know when you might come across a beat up one with good internals for cheap or free! Plus it doesn't take up much space.

by **iMac600** » 30 May 2012, 23:50

uniserver wrote:ok then i blame the BGA

It's not the GeForce BGA. I reworked it last night. No change in behaviour.

that is the one thing that really gets me with apple stuff… the MBP/iMac g5 and intel/Powermac G5, etc They let this stuff get way too hot!
all because they don't want to hear a little fan noise!

All of those machines you listed are certainly capable of moving some serious air, I wouldn't really be prepared to say cooling fans are the problem. Perhaps the layout of the heatpipe or the fact that these integrated circuits get extremely hot by nature is a possibility, but i'd be more inclined to say it's a combination of the heat output and the results of the Restriction of Hazardous Substances directive that's screwing these machines over.

by **techknight** » 31 May 2012, 03:57

quick and sudden shutdown is the result of a shorted on-board MOSFET.

Find it, replace it, and your good to go.

by **uniserver** » 31 May 2012, 06:47

iMac600 wrote:All of those machines you listed are certainly capable of moving some serious air, I wouldn't really be prepared to say cooling fans are the problem.

my point is not there is a physical flaw in cooling capabilities, because that is not what i am saying...
i am saying, they purposely CONTROL the fan speeds, and over heat components, because they feel "fan noise"
is unappealing... i say kick that dam fan up, keep parts cool, makes them happy!~

by **iMac600** » 31 May 2012, 09:54

techknight wrote:quick and sudden shutdown is the result of a shorted on-board MOSFET.

Find it, replace it, and your good to go.

You could be an absolute legend here techknight - I don't even have to go far to source the MOSFETs. I have a donor logic board at work that does power up, but has a bad SMC.

I'll begin probing into that over the weekend. Hopefully will find some positive results.

Presumably something along these lines on a MacBook Pro board (see below) or do they take alternate forms as well? In addition, how would one test them? I've heard they need to be removed from the board prior to testing, but if that's the case maybe i'll just swap them all with ones from the donor board and call it done.

(That's not an MBP board, but you get the idea)

by **Trash80toHP_Mini** » 31 May 2012, 16:15

From the Wikipedia article, it sounds like the short could be internal or external at the trace/component/solder joint level. Maybe oversize solder globs work fine until they reach a level of thermal expansion that causes the short circuit, turning a four gate implementation into a three gate config . . . or worse? Could you have fixed you initial problem doing re-flow and borked a simple solder joint while you were at it?

Dunno, just a WAG on my part, I do 3-D electron plumbing, not electronics by any stretch of the imagination.

by **iMac600** » 01 Jun 2012, 01:52

The initial problem is exactly the same as the current problem. I don't suspect that the GeForce 8600M GT was ever bad in fact, but considering how common the graphics issue is with these models and the number of ways in which a GPU problem can manifest itself, I performed the BGA reflow to rule it out as a possibility. Based on what i'm now seeing, i'm confident that the issue is in one of the smaller components of the board and a MOSFET certainly isn't out of the question.

by **techknight** » 02 Jun 2012, 03:43

Well if the machine was being used and then just shut off in the middle of something never to come back on, or maybe a short "burst" of it trying to come on but shutting down is the result of a dead short.

99% of the time it is a MOSFET somewhere, as ive seen cases where these things have actually caught fire before because of a shorted mosfet exploding. Its been a few years, but ive seen some pictures and cases online where some of those machines had popped and caught fire in a MOSFET cluster area.

Maybe the same FET failed, but less destructively. What you want to do is locate all your power rail clusters and read them to ground. the CPU v-core will read low ohms, this is normal. but it should NEVER read dead short. if any power rail reads dead short, you have found the faulty circuit.

Also pay attention to every semiconductor component on the motherboard. MOSFETs are in several different package types on the motherboard, some are in the higher power package types as you picture, but ive seen lower power MOSFETs hiding in your standard SOIC 8-pin or 14-pin style chips in totem-pole configurations. so it looks like a regular old integrated circuit chip but is actually a transistor. You can usually pick these apart because the most common MOSFET IC chips start with 4XXX like 4420 or 4430 etc being its only markings.

the 14-pin style are mostly FDSXXXX series MOSFET totem-pole ICs. There are shitloads of different ones, those are examples of what to look for. These little buggers drive me nuts becuase they are hard to locate and once you do, its smooth sailing. Ive fixed many of motherboards in my day, and some LCD TV mainboards with those little buggers shorted.

its rare, but another thing to check too once you do locate the shorted mosfet, is sometimes it can damage the gate drive resistor and speed-up shaping capacitor/diode. at worse, it could also damage the Maxim PWM control IC as well. But this is extremely rare.

Dont know about that mac, but i know on PC mobos they use an ENE chip that runs the entire power management and is the boss over all the buck/boost converters on the entire board. So i always start from there and work my way out.