820 2375 No Chime / No backlight

[JorgePayar]

Hi, I have a Macbook air that had the usual problem, it needs to replace the graphic chip.

I did the oven thing a couple times, long time ago, and it was working well, but now...has green light, fan spinning but doesn't chime and has no backlight.

I´ve measured u7100 pin 16 and I have 0.0V, I put some hot air in BGA...nothing...I put some hot air in memory area...dang!! and then...about 30 seconds later the screen was crazy, vertical lines and the computer was screaming like a rat.

Do I have to change the u7100 or could be another thing?

 

[dukefawks]

Bad RAM probably. My advice on these boards is to scrap them. The first generation MacBook Airs are just GARBAGE.

 

[JorgePayar]

Thanks Duke for the advice, but now Im learning so I would like to follow the rabbit hole to hell probably
Did you fix any bad ram memory, did any reball, changing it or something like that in the past? Any temperature advise for changing memory slots?

 

[dukefawks]

Memory slots? The problem is here you never know which RAM chip is bad. My advice is pull them all and replace with NEW ones. As for soldering theses, just do the same as you did with the MCP79U.

 

[Crizz]

I remember a video where a bad tantalum cap prevented the machine to boot.
If you give these broken caps some heat they will work again for a while.
I don't know which section you really heated up and what temperature you used BUT maybe you should try reheating the caps first with under melting temperature before you swab the RAM.
It is worth to give it a shot

 

[JorgePayar]

Thank you both, I´m gonna try it now and I´ll let you know

 

[JorgePayar]

No Title

OK, now the board is trolling me. Now I have a dim green light on charger and no fan spinning. Im learning so I know for Louis videos that first thing to measure is ppbus_g3h but there was no time for that, I plugged it in and the r6905 blew away...there is a short there, my index finger was burned so I think there is a short there
So I need to replace it but I have another board with a similar resistor that is 47 5% 1/16w 402 and this one is 47 5% 1/8w 805, can I use that one? and what would it be next?
Another thing that I find quite strange is that in the schematics I see d6900 when I put this on the board view it shows the place where you can see in the picture that there is no d6900!! ,this MB was working before and I didn't remove the d6900 am I missing anything?

 

[dukefawks]

You plugged it in without the bottom cover AND used a Chinese junk charger. If you plug it in not aligned perfectly it will short out the one-wire pin to 18V. Now you one-wire circuit is fried too probably. D6900 is "no stuff" on the diagram so correct. R6905 fried because of a short on PPDCIN_G3H, D6901 is fried too probably. Also R6905 is 47 Ohm, not 4.7K

 

[JorgePayar]

You plugged it in without the bottom cover AND used a Chinese junk charger. If you plug it in not aligned perfectly it will short out the one-wire pin to 18V. Now you one-wire circuit is fried too probably. D6900 is "no stuff" on the diagram so correct. R6905 fried because of a short on PPDCIN_G3H, D6901 is fried too probably. Also R6905 is 47 Ohm, not 4.7K

I use this charger, I didn't know about the alignment, thanks. You are right about the Ohms sorry I edited already.
Ok I will work on it. Thanks again Duke.

PS: I didn't see the no stuff on the diagram, I will check twice next time sorry

 

[Crizz]

I had once a similar accident with a naked DC in board and a aftermarket charger!
If you don't have the alignment of the case and you don't be careful enough you put 18V to the middle pin of the DC-in board witch is the dataline!
This fries your one wire circuit as duke mentioned!
In my case I was kind of lucky since I saw the magic smoke coming out of U6990 and I manage to fix the board by replacing this IC.
Maybe your U6990 is fried too.
Now I've the LM397 on stock and I try to plug the charger first to the DC-in header then to the wall.

To your case:
You're trying to replace R6905 with one smaller resistor wich has also a lower power rating, I wouldn't recommend this!
Look for a bigger resistor
Is the machine running of the battery?
Furthermore you will still have the RAM issue after fixing this problem ...
This board is a headache!

 

[JorgePayar]

I had once a similar accident with a naked DC in board and a aftermarket charger!
If you don't have the alignment of the case and you don't be careful enough you put 18V to the middle pin of the DC-in board witch is the dataline!
This fries your one wire circuit as duke mentioned!
In my case I was kind of lucky since I saw the magic smoke coming out of U6990 and I manage to fix the board by replacing this IC.
Maybe your U6990 is fried too.
Now I've the LM397 on stock and I try to plug the charger first to the DC-in header then to the wall.

To your case:
You're trying to replace R6905 with one smaller resistor wich has also a lower power rating, I wouldn't recommend this!
Look for a bigger resistor
Is the machine running of the battery?
Furthermore you will still have the RAM issue after fixing this problem ...
This board is a headache!

Lol, I know Crizz, but I think this is the way to really learn something
Ok I will try with a better resistor.
I tested tantalum caps before as you told me and many of them are shorted, could be because of mcp79u b2?

 

[Crizz]

Be careful with your measurements.
On the board there are a lot of caps across CPU/GPU and GND!
CPUs and GPUs appear as a very low resistance so it looks like the cap is shorted!
You have to desolder the CAPs to get proper measurements.

AGAIN headache

 

[JorgePayar]

Ok I have solved this board, it was all because a 470 resistor that you can see in the pictures and the memory, I applied flux under the memory, then put heat (430C) from distance couple of minutes and now it is working perfect, I've been testing it for a few days and works great!!
Thank you all.