Yes, that's one of the most common problems. You should measure anyway the motherboard standby internal power supplies.Hardcoro escribió:so what you are saying is basically that the problem when reballing this board is the PCB warping, aside the Maxim and cpu chain reaction of course. I use a vanilla Achi IR6000 and i've yet to notice PCB warping but it's not like i'm paying very close attention to it. I have the same problem with a board exactly the same, it loops turning on and off and i have to remove the battery to make it stop. what would you recommend to do first?Elect escribió:
Well, this type of motherboard and chip are quite not easy to repair because of some main causes.
The main problem will be the chip itself, most because the common nVIDIA "design bug" especially the date coded previous to 2009.
The difficulty of reworking this chips is not because the size or the 0,5mm BGA as many says. Its more the kind of machine you use to rework it and the operator skills/experience to deal with most PCBs. Most rework machines have a good support platform for amorphous PCBs which prevent warp that may damage the motherboard and preclude a good "landing" of all the balls in each individual pad.
Thank you so much for your help
But, about 90% of that symptom is solved reballing the MCP. If the date code its 2007 you might consider replacing it for a recent one.
The warp maybe too little to notice but might be enough to cause bridging on the chip corners or even not welded balls in the center.
As you may know, the pads in this motherboard aren't bounded and are very small, its more difficult to set a good aliment.
If you use a good profile or in first case a manual set up temperature you may reduce the warp to minimum, but the better and professional choice would be using some jig support.
For your machine you might consider(if you don't have already) using something like this: http://www.diy-buy.com/bga-reballing-ji ... p7454.html
Or you can make yourself ones like I did for my first machine as the ones for my machine were out of stock... (I build most of my tools, but time is money...)
Its not too hard. You can find parts like this from outdoor clearings of the roof eaves and adapt them.
My recommendation for your repair, if you didn't try anything like a reball yet, is to do a step by step clearing for failure.
If possible, try with a known working CPU and RAM. Don't need to attach the fan but keep the heatsink attached while testing, this AMD get hot quickly and may cause temperature shutdown.
If it doesn't work disassemble the laptop, take the power button flat of the chassi too and be careful with its motherboard connector (I receive a lot with broken connectors).
Now with RAM and CPU, connect the power button flat to the motherboard, a external monitor to VGA out and the D/C power adapter.
Apply some pressure over the MCP heatsink zone and press the power button. But be careful with the amount of pressure or you may damage the chip.
This technic its not 100% safe or flawless but when the BGA fault is little, like when there's artifacts on the image, you can get some fault return from it and see changes.
Alternative you can make a gentle reflow not passing the 200ºC but if there's a cold join with oxidation in a pad you'll have no better result.
If you get image on screen after that your problem will be solved with a reball to the MCP. Use some supports to prevent wrap but don't block the direct zone under the chip placement.
Be careful with this chip. It handles temperature very bad.
I get a lot of this laptops from other companies with shorted MCPs due to overheating.
Try placing the PCB with the best alignment between the chip and the Top Heater so there's a good match in the heating area.
Its good to protect the crystals around the chip with aluminium foil but in my opinion its very important to check your temperature probe accuracy and keep one probe in touch with the top of the chip core while removing and soldering back the chip.
Don't pass over 230ºC peak temperature in this chip, specially if the date code is below 2009. I usually go up to 220ºc peak and let it cold down until 190ºC keeping it for 40 seconds mean while. (T.A.L.)
I have no returns yet but I use 3 probes to control the temperature of all chip area to make sure that's really the temperature all over the chip area.
Regards,
FSimões