2012 13"Mbp 820-3115 sudden death

[jurgenkoppen]

I am still very much a newbie , so please forgive my lack of knowledge, I am working on acquiring it, by doing, my best way of learning. I have a 820-3115 board which went from doing great to a sudden death, albeit with light on mag charger still on. I checked the power rails up to PP5V_S3, which is not present. One of my laptop repair gurus suggested to pour alcohol on the audio section and induce voltage into PP5V_S3 to trace the short in that line. I was wondering if I could not just measure the resistance or lack of resistance over the various caps and resistors in that line. If a cap or other component is shortened, would it not beep in diode mode over that component? I have a power supply, but I have never done any inducing power yet, and I am nervous that I might destroy something...

[piernov]

Re: 2012 13"Mbp 820-3115 sudden death

You have to confirm there is actually a short to ground before injecting voltage. Measuring resistance mode to ground on that rail first.
It can be 100 other things causing your problem.
First is to remove the board from the machine and inspect for liquid damage (those rarely die suddenly for no reason). Then disconnect everything except fan and DC-in board, and plug Magsafe in. It should turn on automatically with fan spinning.

[jurgenkoppen]

Re: 2012 13"Mbp 820-3115 sudden death

Thank you for your reply Piernov. I did as you suggested and took out the board. I would like to add that this board came from an Ebay reseller of logic boards, and it did well for the first boot up, in which I tested the GPU by running "Heaven", which it passed. The next time I tried to boot it was dead. The condition of the board is actually quite fine, no corrosion. It looks like there had been some work done by the seller, around U1800 and U2600, judging from the flux residue. I checked the resistance to ground in diode mode on some components on the missing PP5V_S3, such as C6608 and C6607, and they read 1559.However, on the 200 ohm setting on my meter it fluctuates wildly, and then drops to OL. ( I must be missing a bit of info here on what the difference is between diode mode and regular ohm measuring) Can you fill me in on what this might mean? And I have no fan spin....yet. Maybe later with some help...Also, should I have a RAM chip in place to allow for fan spin?

[KvnTM]

Re: 2012 13"Mbp 820-3115 sudden death

Just as an explanation for your measurement:
When you measure a power rail you inject a little voltage from your multimeter into it which charges the cap's on that rail. Capacitors have a very low resistance when discharged which rises while charging. This is why you see fluctuations on your multimeter. OL just means overload which indicates that the value is out of range. >200 ohm in your case. This is a completely normal and expected behavior.

[jurgenkoppen]

Re: 2012 13"Mbp 820-3115 sudden death

Thank you KvnTM for shedding some light on this aspect for for me, all the way from Bayern. So in doide mode then the diode prevents the discharging and I get a more stable value of 1559? Did I get this one right?

[KvnTM]

Re: 2012 13"Mbp 820-3115 sudden death

Bayern is great. Don't think about lederhosen, Bier and Brezeln though. This is not how we look like haha.

Anyways..
In diode mode you measure the voltage drop across your probes. Pretty straight forward.
Measuring a power rail in diode mode doesn't really make sense in my opinion.

One use case for diode mode:
Imagine you have an unknown Diode and you want to know how much voltage it drops. On a normal diode your multimeter would read something around 600-700mV.
The same can be done if you want to know the forward voltage of an unknown LED.

Those are just examples and i don't really know any practical reason to test a power rail in diode mode besides the "beep" when something is shorted. It's just easier to find shorts I guess... Maybe someone else can elaborate further..

[jurgenkoppen]

Re: 2012 13"Mbp 820-3115 sudden death

Ja ich war schon mal da in Bayern, ich bin selber vom Kohlenpott...Duisburg.Und wir haben auch nicht alle Kohlestaub im Gesicht.
To the point:
Oh Boy, I really have to back up and read up on diode mode versus resistance measurements before messing about any more with trying to understand logic boards. Louis Rossmann had recommended using diode mode for quick measurements. So my reading in diode mode is the voltage difference in mV , and not related to ohms at all! To many gaps still in my overall understanding of electronic circuits. Back to the basics electronics study...
Thank for helping me see where to start...

[KvnTM]

Re: 2012 13"Mbp 820-3115 sudden death

Ja ich war schon mal da in Bayern, ich bin selber vom Kohlenpott...Duisburg.Und wir haben auch nicht alle Kohlestaub im Gesicht.
To the point:
Oh Boy, I really have to back up and read up on diode mode versus resistance measurements before messing about any more with trying to understand logic boards. Louis Rossmann had recommended using diode mode for quick measurements. So my reading in diode mode is the voltage difference in mV , and not related to ohms at all! To many gaps still in my overall understanding of electronic circuits. Back to the basics electronics study...
Thank for helping me see where to start...

Ich komme ursprünglich aus Hamm. Ist nicht weit von Duisburg
----

Yes, diode mode is indeed a quick way to find shorts because the meter beeps on a short so you don't have to move your eyes from board to the meter all the time. As you mentioned it has nothing to do with resistance. Thats what the forum is for.. Sharing knowledge and helping others in need. Im pretty sure someone is able to help you further. Im not into MacBook logic boards that much I only have basic knowledge about their power rails. In fact im writing this on a mid 2012 MacBook with the same logic board. Those mid 2012 MacBooks are one of the most reliable ones with only one major known failure point which is the sata cable. I can tell it is really reliable. I carry it every day and abuse it a lot with high cpu loads/temps. Never let me down once.