Repair a water damaged Macbook. Is it worth it?

[amyk]

no, all parts solderd on the board are heatresistant... they have gone through the reflowoven...

With a specific temperature profile, however. They're not designed for anything more than the bare minimum necessary to survive one round of reflow.

I would strongly not suggest going over 100°C, just to bake out the moisture.

[Heribert Hechtersheim]

Hey, I am back!
I decided not to put the logic board again into the oven and startet to measure it. I did what amyk said and looked at the "power system architecture" page of the schematic. Then I found a broken 1 ohm resistor which was protecting the 3.425V "G3Hot" Supply. I shorted it with my multimeter and the current was only about 4 mA so I oversoldered it. I will replace it with a smd resistorbut I don't have them in stock.

Then I was able to start the macbook. I heard the typical mac start sound, but the display was black. The color changes from dark green to black when it starts. I assumed that the backlight is broken. I searched on the pcb and found another broken fuse(F9800; 2A 32V) which was protects the LCD Backlight Driver (MC34845). Again, I shorted the fuse with my cheap multimeter on 10A range but this time I wasn't lucky. The current was higher than 2A so I quickly disconnected the multimeter. How much current should it normally draw? The voltage is 12.6V.
Now I will check if any discrete components in the backlight driver circuitry are damaged or if the ic is the damaged part(How can I proove that?). It could also be possible that the LEDs or something else which is located on the screen is causing the problems.

Thank you all for your previous help. I will appreciate further help, ideas and suggestions.

[fluxcapacitor]

http://forums.macrumors.com/showthread.php?t=972699&page=10

read post # 238

The thread is over 60 pages but your model will be amongst them ,with part numbers ,pinouts,datasheets,pictures, and schematics etc .

[NiHaoMike]

Does it work with an external monitor?

[Heribert Hechtersheim]

http://forums.macrumors.com/showthread.php?t=972699&page=10

read post # 238

The thread is over 60 pages but your model will be amongst them ,with part numbers ,pinouts,datasheets,pictures, and schematics etc .

Thank you! The picture that was posted showes exact the same fuse I found. I measured the current again between the fuse's ends. It was about 5A. But I saw no backlight.

Does it work with an external monitor?

I don't know that at the moment, because I don't have the right adapter for my monitor. But I ordered one which I will receive tomorrow. Till now the only thing I can notice is the mac start sound and hard drive activity.

[Heribert Hechtersheim]

I found the problem. The 50V output from the lcd backlight driver was shorted. I tooke the lid apart and found a partially molten connector:



The missing pad sticks in the connector



Now I will order a new screen and a new cable. It will cost me about 60€.

I shorted the blown fuse and measured the output of the 50v lcd backlight driver. It is only 37V. Will it rise when a load is connected to it?
The circuit is shown on page 67 on the service manual: http://schemalaptop.blogspot.de/2013/03/apple-macbook-a1342-mlb-k87-macbook.html

[ivan747]

So to repair water damage, which might have created shorts because dissolved salts remain after drying, you rinse it with more water with the same salts and then dry it to make sure all dissolved salts get deposited on the board? Distilled water costs less than €1 per liter last time I checked.

I was too lazy to buy distilled water.

So you were too lazy to spend 1€ on that for your €300 repair? I think you have just ruined the board.

Scrap the project. Forget it. The best thing you can do right now is buy 2 broken MacBooks and mix all the working parts together and sell the rest.

Maybe you can use the case of your water damaged MacBook and buy a functional MacBook with a broken case.

Oh and the battery is probably dead as well.


Sent from my iPhone using Tapatalk

[Heribert Hechtersheim]

So to repair water damage, which might have created shorts because dissolved salts remain after drying, you rinse it with more water with the same salts and then dry it to make sure all dissolved salts get deposited on the board? Distilled water costs less than €1 per liter last time I checked.

I was too lazy to buy distilled water.

So you were too lazy to spend 1€ on that for your €300 repair? I think you have just ruined the board.

Scrap the project. Forget it. The best thing you can do right now is buy 2 broken MacBooks and mix all the working parts together and sell the rest.

Maybe you can use the case of your water damaged MacBook and buy a functional MacBook with a broken case.

Oh and the battery is probably dead as well.


Sent from my iPhone using Tapatalk

I am making progress, why should I buy a new pcb?
What do you think got damaged during the washing?
I discharched the caps before I put it into the dish washer.

Tomorrow I will receive my display port adapter so I can hook the macbook up to my external monitor. Then I will see if everything works properly except the internal lcd or not.

[amyk]

Till now the only thing I can notice is the mac start sound and hard drive activity.

It's alive...

color changes from dark green to black when it starts

That's normal, it means the LCD is receiving the right drive signals. You should be able to read the display if you shine a torch at it.

Then I found a broken 1 ohm resistor which was protecting the 3.425V "G3Hot" Supply. I shorted it with my multimeter and the current was only about 4 mA so I oversoldered it.

Which one exactly? It might've been a current sense resistor, and in that case you definitely need to replace it with one of the correct value.

I shorted the blown fuse and measured the output of the 50v lcd backlight driver. It is only 37V.

It's a constant current, not constant voltage driver. Either way the IC can produce up to 60V but it's also PWM'd so I wouldn't expect maximum voltage unless the brightness is also set to max.

(Checking for shorts is something a thermal imaging camera would be very useful for. )

[Heribert Hechtersheim]

Then I found a broken 1 ohm resistor which was protecting the 3.425V "G3Hot" Supply. I shorted it with my multimeter and the current was only about 4 mA so I oversoldered it.

Which one exactly? It might've been a current sense resistor, and in that case you definitely need to replace it with one of the correct value.

I marked the resistor with an arrow. The ~1k vaulue was the value the resistor had before I oversoldered it.





The output voltage is a bit lower then the 3.425V but still higher than the minimum so I think that is ok.

I shorted the blown fuse and measured the output of the 50v lcd backlight driver. It is only 37V.

It's a constant current, not constant voltage driver. Either way the IC can produce up to 60V but it's also PWM'd so I wouldn't expect maximum voltage unless the brightness is also set to max.

Ahh, that makes sense. In the datasheet(http://www.freescale.com/files/analog/doc/data_sheet/MC34845.pdf) they also say that "The 34845 includes fault protection modes for LED short and open, over temperature, over current and over voltage errors." So hopefully nothing got damaged when I shorted the fuse.

(Checking for shorts is something a thermal imaging camera would be very useful for. )

You are right, that would make fault finding a lot easier. And they are usefull for other things too like finding cold spots in the house.
I read this really interesting thread in the morning, but unfortunately I don't have the money to buy an e4.

[cosmos]

The MF marking and circuit position of R6905 makes me think it most likely is a polyfuse.

edit:
I had a recent use of polyfuses in mind when writing that and their part numbers all started with MF-something ... when looking again at the schematic the other resistors also have the same MF-LF naming and they are for sure not polyfuses.  So scrach the above on polyfuse.
R6905 is a simple 1 ohm 1/8W resistor in 0805. 

[ivan747]

That's surprising! The motherboard is working!? We'll congratulations, Truly believed you messed up the motherboard. As someone else said, your backlight is probably broken. Try to get the computer in a very right room and move the display until you have no glare and try to see if the display itself is working, you should see everything. Also try increasing the brightness. Mac's minimum brightness setting turns the backlight off.


Sent from my iPhone using Tapatalk

[Heribert Hechtersheim]

The MF marking and circuit position of R6905 makes me think it most likely is a polyfuse.

Does MF stands for metal film? So the metall should get high-resistive when the current is too high?

That's surprising! The motherboard is working!? We'll congratulations, Truly believed you messed up the motherboard.

Thank you! I did not believe it as well because I hooked it up to an external monitor and got no image, but then I put pressure on the gpu before I started the macbook and voila - I saw the desktop. I tightened the screws of the cooler and put the macbook back together. I don't know how permanent this fix will be. If it fails again I think I will put it into the oven.

As someone else said, your backlight is probably broken. Try to get the computer in a very right room and move the display until you have no glare and try to see if the display itself is working, you should see everything. Also try increasing the brightness. Mac's minimum brightness setting turns the backlight off.

As posted before, the leads for the LEDs at the LVDS connector on the pcb of the LCD panel were shorted and melted the plastic. I ordered a new cable and a new screen, because it has little bubbles inside due to the water.
Everthing else(wLAN antenna, keyboard, touchpad, dvd drive) works fine.

[TheBorg]

 

Really. There's not much here to say.

First of all, when doing repairs like this please put in the time to do it right. Things like buying alcohol and distilled water? Not very hard to do, not all that costly (10 euros? maybe less). Washing the board with tap water introduces the same corrosives and minerals that destroyed the board in the first place. Not good.
And alcohol, unless you buy a whole lot of it at an extremely high purity costs very little.
It's a lot easier to help you get the whole thing working right if you slow down a bit and want to do it right.

Secondly, one thing I learned on fixing these boards is to NEVER short out anything. Shorting out the fuse? Yeah, the problem was likely still there and you just blew the whole backlight circuit and the LVDS connector. Same with oversoldering components. They are there for a reason- its almost always better to wait a week and spend an extra 10 euros to get replacement parts than to try and repair the *entire* circuit later.

That said, I'm surprised you got this far...

Finally, I think the 50v you say the backlight needs is wrong, from what I have read this should be around 32v. Also, you should probably put some more thermal paste on the heatsinks and chips since you have washed the board so many times.

[TheBorg]

[quoteReally. There's not much here to say.]

You said quite a bit.
[/quote]
Ahh...
Seemed less when I wrote it. Plus I was ranting a bit

[amyk]

"smoking out" bad components/shorts is an often-used technique for troubleshooting and it does work.

[TheBorg]

"smoking out" bad components/shorts is an often-used technique for troubleshooting and it does work.

In some instances, yes, this technique would work out quite well. However, I'd have to disagree for doing it here. Shorting out the fuse here with a potential short down the line could've very well had the OP replacing QFNs and green wiring traces. It's a great technique if you want to see if said component is causing the short, but can also cause bigger problems down the line. I've been a part of the macrumors backlight thread a while now, and I can say that shorting out high risk components (fuses especially) almost never works out well. I personally had to replace the backlight fuse in my macbook 3 times before I found the cause of the short.

Then again, I'm not an EE...

[Heribert Hechtersheim]

Really. There's not much here to say.

First of all, when doing repairs like this please put in the time to do it right. Things like buying alcohol and distilled water? Not very hard to do, not all that costly (10 euros? maybe less). Washing the board with tap water introduces the same corrosives and minerals that destroyed the board in the first place. Not good.
And alcohol, unless you buy a whole lot of it at an extremely high purity costs very little.
It's a lot easier to help you get the whole thing working right if you slow down a bit and want to do it right.

I think I would wash dirty boards again in the dish washer. Probably not if they have BGAs on them, because the dirt and moisture will maybe stick under the chips.
If you dry the board after the washing, there is not much time for the metals to corrode.

Secondly, one thing I learned on fixing these boards is to NEVER short out anything. Shorting out the fuse? Yeah, the problem was likely still there and you just blew the whole backlight circuit and the LVDS connector. Same with oversoldering components. They are there for a reason- its almost always better to wait a week and spend an extra 10 euros to get replacement parts than to try and repair the *entire* circuit later.

I cannot check every single component of the circuit when I work with such an overpopulated smd pcb. As a electronics beginner like me, I need very much time to figure out where components of the schematic are on the pcb.
Before I shorted the fuse, I only checked how the components are connected. I must admit that I could have noticed that the LED cathode was shorted. It was a mistake that I didn't checked that.
Now the connector is partially molten and I had to order a new one from Hong Kong which costs me 7€ and what is even worse: time (I have to wait maximal till christmas). I cannot say if the damage to the connector was done before by the water or after by me by shortening the fuse.

Finally, I think the 50v you say the backlight needs is wrong, from what I have read this should be around 32v.

You are right, the 50V they wrote in the schematics would be too high. I disassembled the dirty screen and saw that the led stripe is splittet up into 6 lines which consist of 9 LEDs in series each. So if you say that every LED has a voltage drop of 3V you would need 27V per line to light the LEDs. If they cycle through the LED lines, the voltage of 37V I measured seems to be OK I guess. In addition, the voltage will probably drop if the LEDs are connected.