Need help with with macbook pro 820-3115-B

[MadMaxDog]

Hi All,

This is my first post. If you read my intro (intro) it will explain why I am here and asking these questions.

The skinny version is this:

- I am new to electronics and smd repairs.

- I have attempted to fix a macbook pro 820-3115-B that has no Vcore, no fan, but has green light on charger (excellent condition, no liquid or any kind of damage whatsoever, and has never been fixed before).

- While probing around, on 3 separate occasions, I touched something I shouldn't with my probes (shorted some connections, I guess - not intentionally) and the fan started spinning. I cannot replicate this, it was in 3 different places, but there must have been a specific sequence to each event, because just going back and touching the same things does not work. I've tried to make it happen again for almost 2 hours and couldn't (don't worry, I didn't go crazy randomly connecting things to each other, I checked the lines, and the voltages, and attempted to discover what was supposed to happen in each case before I decided to bypass certain components. Unfortunately I didn't keep track of those experiments because I was eager to see it working again. Rookie mistake huh?).

- I traced the powerlines, wrote down all the voltages and finally (kind of) realised that everything I chased ended up eventually leading me to the same place: U7941 (which has voltages going in left and right, but no output on pin 4, which is PP3V3S5_EN_L_R)



Here are a few more interesting facts:

I am using a macbook air charger, which is only 45 watts, but I figured since I am only powering the board, and have no screen or keyboard, it should be enough (right?). I got a bit more comfortable with this idea once I saw the fan spin for the first time. It's certainly enough juice to get that going, and that's my goal right now.

If I measure for a short (red probe on ground) at R5410 (which is on PPBUS_S5_HS_OTHER_ISNS), I get a short during 10 seconds after I disconnect the power. then the short goes away. This happens every time like it's some caps discharging.

When I decided U7941 was bad, I injected 3.3 volts from my power station into PP3V3S5_EN_L_R but still no joy (which I think should be created by U7941 on PIN 4 - but hey, sometimes I have no idea what I'm doing, and others I'm quite confident).

Caps C2420 and C2428 both measure as SHORT when power is disconnected (and so do all the many smaller caps along the same line), but I removed both from the board and tested. They are fine so I put them back (yes, in the right order. I know this because I did one first, put it back, then removed the other, so there's no chance I switched them by mistake).

R5070 tests at 8.3K instead of 10K (multimeter is normallyquite accurate)



So I have a few questions before I post the photos:

1. Is the 10 second short on PPBUS_S5_HS_OTHER_ISNS normal? Could that be the problem, maybe some bad cap(s) along the way?

2. Is it normal for U7941 to blow on these boards (maybe it's a well-known issue?) or is it fine and I'm just being a moron and there is not supposed to be any power on pin 4? (I did consider that maybe the power needed to drop on PIN3 for PIN 4 to get an output)

3. Would U7941 just blow like that or would it be the result of some other part going bad first? (I worry about this because I have no spare boards for parts to experiment with, and I don't want to order U7941, wait 3 weeks, then have it blow again due to something else)

4. Does my diagnostic of U7941 being bad fit the events I described and the values in the pictures below? (considering it even worked a few times, and I injected voltage to mimic U7941's output and still nothing happened). Can it do that?, work sometimes and not others?

5. Does the short on C2420 and C2428 mean anything, or are they just very low power lines that can show up as a short due to low resistance?

6. Could R5070's low(ish) reading be a problem?

7. Am I at least on the right track?



and here are the photos:

(Voltages are in red. When a part has a green circle it means I tested it and it checks out. If it says 'no short' it means I tested that too with power disconnected and red probe on ground)















Many thanks for even reading all this.

[MadMaxDog]

Re: Need help with with macbook pro 820-3115-B

Also, I forgot to mention, and it seems important to me: When plugged in, I hear a very faint and constant clicking sound. I have put my ear to it many times but it just sounds even all over the board (on both sides) and impossible to pinpoint (even tried a tube on my ear, cell phone microphone with a sound analizer app). It clicks about twice per second.

[BremiX]

Re: Need help with with macbook pro 820-3115-B

Also, I forgot to mention, and it seems important to me: When plugged in, I hear a very faint and constant clicking sound. I have put my ear to it many times but it just sounds even all over the board (on both sides) and impossible to pinpoint (even tried a tube on my ear, cell phone microphone with a sound analizer app). It clicks about twice per second.

I had this weird sound with a few boards too and couldn't figure out which demon was possessing it....

[spleenharvester]

Re: Need help with with macbook pro 820-3115-B

https://www.rossmanngroup.com/boards...5_en_l-missing

So I'm not a macbook repair specialist or anything, but the above thread seems to suggest that _L voltages are meant to be missing?

[MadMaxDog]

Re: Need help with with macbook pro 820-3115-B

Ah, interesting... So U7941 might be fine after all... So in than case, what power is missing to get this going... SMC_S4_WAKESRC_EN maybe?

... So I see it comes out of U4900... Is this the time when I have to test all U4900 inputs? Or is there a better / smarter way to go?

[MadMaxDog]

Re: Need help with with macbook pro 820-3115-B

Took some more measurements and here's another image... Q7840 on 5V_SUS FET diagram doesn't seem right to me. Any thoughts? or does the 8/5 mean it's closed at 5 volts (4.8 in my case) and open at 8v ?

[piernov]

Re: Need help with with macbook pro 820-3115-B

1. Not sure but could indeed be capacitor discharging and MOSFETs turning off after some time, I wouldn't think much of it about it right now.
2-3-4. No, U7941 is perfectly fine, your issue is elsewhere as PP3V3S5_EN_L is supposed to be low (_L means active low logic, ie. it is on when the voltage is 0V).
5. Measure the resistance across the caps. A low resistance doesn't necessarely means it's shorted.
6. Resistor when measured in circuit are supposed to measure lower or equal readings since the other components will affect the overall resistance of the line so it's fine. A failing resistor will measure higher.

This looks like a missing PM_SLP_S4_L signal so I would check that first.

EDIT: Use an original Apple AC adapter with at least the required power rating. This will save you some headache. It's not only about the throughput of the power supply but also the SMC recognizing it.

[BlueMidnight]

Re: Need help with with macbook pro 820-3115-B

Good job on writing down the voltages on the schematic. That is the best way to do it. I see the voltages around R7270 are wrong, but the circuit is working so it must be a typo/mistake.

You're off to a good start. You aren't 100% familiar with schematics and how all the components work, so you are going down some dead ends. But that's normal when starting out. If you just keep measuring everything, and trying to understand why each voltage is what it is, you'll get better in no time.

1. Is the 10 second short on PPBUS_S5_HS_OTHER_ISNS normal? Could that be the problem, maybe some bad cap(s) along the way?

That is normal. Resistance mode on your meter will send a tiny current through the circuit. It then measures voltage on each probe and uses Ohm's Law to figure out the resistance of the thing you are testing. Never try to measure resistance on a circuit that is powered up. At best, it will give you strange readings. At worst, it can damage your meter.

2-4. U7941...

Those triangles inside U7941 are buffers. A buffer keeps the input and output isolated, but "copies" the high/low state from the input to the output. The little circle on the top one means that the output of it is inverted. It is therefore an inverting buffer (a.k.a. an inverter or a "NOT gate"). Since the inverter is getting a high (3V) signal, it will output a low (0V) signal. So, what you are reading is normal.

Besides voltages, you need to know a couple other things about signals on a schematic. First, you need to know what the signal indicates. For example, a signal might be called "ENABLE_THING". That probably means activating that signal will enable the thing. But secondly, you must know whether a low voltage means enable it (i.e. "active low") or a high voltage means enable it ("active high").

Different schematics have different ways of telling you whether a signal is active low or active high. But there is almost always something added to the name to show if it is active low. If nothing is added to the name, then it is assumed to be active high. Apple likes to indicate active low by adding "_L" to the name. Other schematics will use "#" or "*" or will put a line over the top of the name, etc.

5. Does the short on C2420 and C2428 mean anything, or are they just very low power lines that can show up as a short due to low resistance?

On page 7 of the schematic you can see that PP1V05_S0_PCH_VCCASW is another name for PPCPUVCCIO_S0_REG, which powers many things. One of the things it powers, as the name suggests, is VCCIO. That's the CPU's I/O logic, which uses more than a few amps by itself. Ohm's Law tells us that low voltage and high current must mean a low resistance. A reading well under 1 ohm is normal here.

6. Could R5070's low(ish) reading be a problem?

It's hard to calculate what it should measure in circuit. There may be other components in parallel with it. But it's just a pull-up resistor. It's only job is to keep that signal at 3.3V as long as the power button isn't being pressed, and therefore pulling that signal to 0V. It just has to be large enough to not waste a lot of power when the button is pressed. Even at 8.3K ohms, that signal line will only consume 3.3V / 8300 ohms = 0.0004A. There's nothing worrying about that value.

Q7840 on 5V_SUS FET diagram doesn't seem right to me. Any thoughts? or does the 8/5 mean it's closed at 5 volts (4.8 in my case) and open at 8v ?

They are just indicating that this FET supports a max of 8V between source to drain and 5V between gate and source. Writing it as "8V/5V" is just their own way of keeping track of that information.

I recommend finding some tutorials about transistors (BJT and MOSFET) as soon as possible. Like today. Memorize the different symbols and what conditions will switch them on. Reading a schematic properly will be impossible until you know this.

Q7840 won't switch on unless the gate pin is a lower voltage than the source pin. The gate will go to 0V when Q7822 switches on. And Q7822 will switch on when P5V_3V3_SUS_EN goes high.

7. Am I at least on the right track?

You haven't located the problem, but you've recorded all the information you need to find it.

Apple schematics have one really nice feature. You've obviously noticed that page 3 has a power diagram. The big numbers on that page indicate the sequence in which the power rails turn on.

If you closely examine the voltages you wrote down, you'll see that you are not completing step 10-1 successfully. That is because P5V_3V3_SUS_EN is low, which is a different name for the PM_SUS_EN signal. And you're not getting that because SMC_BATLOW_L is low.

The important stuff is happening around U7940. This chip performs the logic "If (A or B) AND C then set Y HIGH". In other words, set P5V_3V3_SUS_EN high:

1. If PM_SLP_SUS_L is high and if SMC_BATLOW_L is high, or...
2. If PM_SLP_SUS_L is high and P5V_3V3_SUS_EN is already high.

We know that:
- PM_SLP_SUS_L = 3.3V (picture 6)
- SMC_BATLOW_L = 0V (picture 1)
- P5V_3V3_SUS_EN = 0V (picture 1)

That is the correct output for those inputs. Therefore, U7940 is probably working normally. SMC_BATLOW_L is the problem.

Why is SMC_BATLOW_L = 0V? The signal comes from the SMC. The first thing I'd try is what has already been suggested: use the correct wattage AC adapter. There is a data connection (the "onewire" circuit), and the SMC knows exactly which adapter you have plugged in.

[MadMaxDog]

Re: Need help with with macbook pro 820-3115-B

Wow, thanks for that detailed post.

Unfortunately I don't have another charger, nor do I have access to one, so I'm going to have to figure this out using this one. Did I mention it has worked 3 times? All I'm going for right now is a fan spin.

Would it be absolutely silly to inject a volt or two where SMC_BATLOW_L is supposed to be?

As you noticed, I have been struggling a bit with all the abbreviations. And this confuses me more... If the _L's are supposed to be low, why do I need power in SMC_BATLOW_L ? Also, when I see something like SUS I always wonder if it means SUSpend or SUStain... because they're kind of opposites to one another and would influence which path I go down while chasing a problem. If it means SUSpend, then activating something like P5V_3V3_SUS_EN (where I read SUSPEND_ENABLE) would actually be an order to shutdown things and not boot up, right? (I may be saying something incredibly stupid, but since it crossed my mind, I though I would share it).

Apple schematics have one really nice feature. You've obviously noticed that page 3 has a power diagram. The big numbers on that page indicate the sequence in which the power rails turn on.

That's a really cool tip. Thank you.

You were also right about the readings on R7270 being wrong, I have updated the image to the new values.

I just dropped in to read the posts but I haven't had time to mess with the board yet. Tomorrow I will be measuring some more things and trying to follow all the tips here. Then I'll post again.

Thank you so much.

[MadMaxDog]

Re: Need help with with macbook pro 820-3115-B

New Info: I can get the board to start up by briefly shorting R5040 (that in my understanding is sending 3.3v into SMC_BATLOW_L), so I may be getting closer to the problem.

R5040 measures at 100.5 K, so I figure it's alright, but then again, if I have 3.3 volts on a line with a single 100k resistor, shouldn't I also have close to 3.3 on the other side? Could R5040 be the problem?

[MadMaxDog]

Re: Need help with with macbook pro 820-3115-B

Funny how some pages are missing from the schematics pdf. I have found 3 different versions online and none of them have the SMC page.

Anyone got that page for an Apple 820-3115-B ?

I've been searching an old airport board for another 100k resistor to replace R5040 (I don't know, I just think I should try that). Found one, but while tweezer-transporting it to the macboook board, it jumped off into oblivion. That was the only micro-tiny-little-resistor I had (for 100k), the only other one I can find is about double the size... Still tiny, but slightly too big to fit.

In the meantime I thought I would take a look at the SMC Page and try to find out why BATLOW is not being created... But no page.

[piernov]

Re: Need help with with macbook pro 820-3115-B

Don't randomly replace components when they are fine and there are no reasons for them not to be… The only thing that will happen is that it will get worse if you make a mistake (and I'm the first to loose components when I desolder them, so I try not to desolder anything unecessary).
As I said you need to get the correct AC adapter for the machine before messing with the board.

[BlueMidnight]

Re: Need help with with macbook pro 820-3115-B

If the _L's are supposed to be low, why do I need power in SMC_BATLOW_L ?

It just takes a little while to fully understand how this all works. The name doesn't tell you what it is "supposed to be". The signal is just sending information. It is saying some condition is TRUE or FALSE.

A signal will have a high voltage and a low voltage that it can be at. But, does a low voltage = TRUE, or does a high voltage = TRUE? Either way would work. The engineers get to choose. But they will put "_L" or something in the name if a low voltage = TRUE.

For SMC_BATLOW_L, the voltages will be 0V or 3.3V. The "_L" in the name tells us that 0V must mean TRUE, and therefore, 3.3V means FALSE. This signal answers the question "Is the battery (or AC adapter) power low?" If the signal is 0V, then the answer is TRUE. If the signal is 3.3V, then the answer is FALSE. In this case, FALSE is a good thing.

Also, when I see something like SUS I always wonder if it means SUSpend or SUStain... ...activating something like P5V_3V3_SUS_EN (where I read SUSPEND_ENABLE) would actually be an order to shutdown things and not boot up, right?

It does stand for "suspend", but don't worry about it. "SUS" just a name for a certain group of components that are the first to turn on. They are followed by the "S5" stuff, then the "S4" stuff, then the "S3" stuff, etc. Explaining why they have the names they do would just be way too confusing right now.

Funny how some pages are missing from the schematics pdf. I have found 3 different versions online and none of them have the SMC page.

The schematic does have the SMC. It's U4900 and on page 45. If you mean it doesn't have the inside of the SMC, that normal. You aren't going to find that. It's all controlled by firmware anyway.

Would it be absolutely silly to inject a volt or two where SMC_BATLOW_L is supposed to be?
Unfortunately I don't have another charger, nor do I have access to one, so I'm going to have to figure this out using this one.

The board thinks it doesn't have enough power to start. You have a 45W adapter plugged in. The board is correct. Even if there is something else wrong with the board, there is no reason to suspect it. There is nothing more to investigate right now.

However, you might be able to let it sit there with the AC adapter and the battery plugged in and let it charge the battery. Then try to start it up with just the battery (assuming it has a good battery). If that doesn't work, then there is nothing else you can do other than get the correct AC adapter.

[MadMaxDog]

Re: Need help with with macbook pro 820-3115-B

Thank you BlueMidnight, that clears quite a few doubts I had.

It just takes a little while to fully understand how this all works. The name doesn't tell you what it is "supposed to be". The signal is just sending information. It is saying some condition is TRUE or FALSE.

This makes more sense to me now.

For SMC_BATLOW_L, the voltages will be 0V or 3.3V. The "_L" in the name tells us that 0V must mean TRUE, and therefore, 3.3V means FALSE. This signal answers the question "Is the battery (or AC adapter) power low?" If the signal is 0V, then the answer is TRUE. If the signal is 3.3V, then the answer is FALSE. In this case, FALSE is a good thing.

Ah. This is where I was also getting confused about the name. I figured BATLOW *ONLY* referred to the battery, which is not even connected right now so it should always be low, right? (If LOW = TRUE, and the name 'BATLOW' implies "is_BATtery_ LOW?", with no battery connected I would expect the battery to be low, therefore no power on SMC_BATLOW_L makes perfect sense to me. But if SMC_BATLOW_L also refers to power coming from the charger, that changes things and I want it to be FALSE (ie: "meaning: 'BAT' is not low, we're good to go") and it should have around 3.3 volts (I suppose).

What everyone seems to be saying is that I have 0v at SMC_BATLOW because the board thinks my 45 watt charger isn't strong enough to boot?

This would kind of make sense to me... But... If that's the only problem, then getting the correct charger would make this thing work, however, I happen to know it doesn't work with the correct charger, because that's the info that was given to me: Charger was fine, computer would not start.

So I know that getting a new charger won't make this bad boy boot up. It may help me find the problem, but I'll be basically at the same place I am right now, with all power good except at SMC_BATLOW.

It does stand for "suspend", but don't worry about it. "SUS" just a name for a certain group of components that are the first to turn on. They are followed by the "S5" stuff, then the "S4" stuff, then the "S3" stuff, etc. Explaining why they have the names they do would just be way too confusing right now.

Better leave that one for later then, last thing I need is more confusion right now.

The schematic does have the SMC. It's U4900 and on page 45. If you mean it doesn't have the inside of the SMC, that normal. You aren't going to find that. It's all controlled by firmware anyway.

All my versions jump from page 43 to page 45. There is definitely a page missing in between. And it's not the only one, there are more missing pages in the pdfs. Are they missing because they're not important/relevant? I just wonder what's on them.

The board thinks it doesn't have enough power to start. You have a 45W adapter plugged in. The board is correct. Even if there is something else wrong with the board, there is no reason to suspect it. There is nothing more to investigate right now.

Is my theory about the charger stupid? This is what I was thinking: I figure that if the original charger is 60 watts, that's the maximum output it can go to, so that would be needed when the computer is in full running mode, and probably doing quite a lot of processing. Since I don't have anything connected except the fan (no screen, no keyboard, no trackpad, no backlight), there's really no reason for the board to need the full 60 watts. 45 should be more than enough.

However, you might be able to let it sit there with the AC adapter and the battery plugged in and let it charge the battery. Then try to start it up with just the battery (assuming it has a good battery). If that doesn't work, then there is nothing else you can do other than get the correct AC adapter.

I'm definitely going to try this. It had crossed my mind that I wouldn't need the charger if I could find a way to charge the battery. I'm just not sure if it will charge on this board, but I'll give it a shot.

(Just connected battery. Now I'm getting orange light on charger instead of green. Good sign it might be charging the battery. I'll give it 20 minutes then test)

Thanks guys. I'll keep you posted on further developments.

This is what I'm getting on battery connector. Seems to be charging.



P.S. When I connect the charger to the board, I would expect the fan to start spinning, but what about if my power source is the battery? Is just connecting the battery to the board enough, or do I have to mimic a power button press in some way?

[BlueMidnight]

Re: Need help with with macbook pro 820-3115-B

I figured BATLOW *ONLY* referred to the battery.

Understandable. But if you look how U7940 works, that cannot be true. BATLOW always keeps the system from starting because of that circuit. Do MacBook Pros always refuse to start unless the battery is charged up, even if the AC adapter is plugged in? Obviously not. So, that signal must mean something a little bit different.

I happen to know it doesn't work with the correct charger, because that's the info that was given to me

Listen to people when they tell you what they think is wrong with their computer. But don't automatically believe everything. They are often wrong. Try to diagnose the system without making any assumptions.

There may or may not be some other problem with the board. But you are going to have to have the correct AC adapter or a charged battery before you will be able to see the problem and find it.

All my versions jump from page 43 to page 45. There is definitely a page missing in between.

There is a spot in the corner for "page" and also for "sheet". Only the "page" number skips around. No actual information seems to be missing. So, the missing pages are almost certainly for other models/configurations of the board, other circuit boards in the system, notes, etc.

...Since I don't have anything connected except the fan... ...there's really no reason for the board to need the full 60 watts. 45 should be more than enough.

I'm not sure that's enough power or not. But let's assume it is hypothetically. You may know that, and I may know that. But the SMC doesn't know that. It doesn't know what you are planning to do with the computer once it is powered up. All it knows is that someone is trying to switch on the system and this system needs more power than the AC adapter can safely provide. The only reasonable thing for it to do is to prevent that.

This is what I'm getting on battery connector. Seems to be charging.
Yes, it looks like it is trying to at least. Hopefully the battery is good.

Is just connecting the battery to the board enough, or do I have to mimic a power button press in some way?
It might not turn on some components when it's off and only on battery power (the SMC_PM_G2_EN signal may stay at 0V). If you need to, short pin 5 of keyboard connector J5713 to ground for a moment to create a power button press. Or place the board back in the chassis and connect the keyboard and battery there.

[MadMaxDog]

Re: Need help with with macbook pro 820-3115-B

I haven't had too much time to play around with it, but I do have some progress.

I removed 2 resistors that were testing at the wrong values, so I could test them off the board. They checked out so I put them back (R2156 & R5040).

The battery charging worked, so now I have a correct power source for testing. With battery connected (both with and without charger connected) I can get a 20 second fan spin by shorting Pin 5 of keyboard connector to ground (Thanks BlueMidnight). But no chime and fan stops after 20 seconds, so still no CPUVCORE.

I think the heat from replacing the resistors may have influenced something. So I have what I wanted initially, which was a fan spin. Now I want a chime.

Some of the voltages I had measured previously seem to have changed, so I'm going to have to re-measure a bunch of stuff and update the values. In the meantime I thought I would post this new info just in case it's obvious to someone.

P.S. I recently discovered, while trying to gather more info on the situation, that there's a possibility the death was caused by a knock-off charger. In my innocence and newbie-ness I don't imagine that's helpful at this point, since I assume a damage from a knock-off charger would affect the one-wire circuit before anything else, and that circuit is fine.

I'll post again soon.

[piernov]

Re: Need help with with macbook pro 820-3115-B

No CPU VCore implies no chime, but no chime doesn't necessarely means there is no CPU VCore so you have to check that first.

[MadMaxDog]

Re: Need help with with macbook pro 820-3115-B

No CPU VCore implies no chime, but no chime doesn't necessarely means there is no CPU VCore so you have to check that first.

I did. No CPUVCORE.

[BlueMidnight]

Re: Need help with with macbook pro 820-3115-B

That power diagram on page 3 can be very helpful here. Your board has a problem around step 25 or before. So, just concentrate on measuring the signals and power rails displayed on that page, starting at step 25 and working backwards. You can stop when you get to steps that are working correctly. The problem is most likely somewhere in between.

And again, please try to resist the temptation to desolder things unnecessarily.

[MadMaxDog]

Re: Need help with with macbook pro 820-3115-B

My main problem has been more of a logistical one. My girlfriend is constantly requiring the charger so she can watch movies on the projector with my macbook air, so I get stuck here just being able to measure resistors and shorts, and I get bored, so eventually I'm going to take something off then put it back again. It's quite hard to resist (get it? resist... resistor...) :P

I would like to use (or learn to use) the functional test points featured on page 7, but I don't quite understand them. I find it weird that they would list functional test points without the voltage expected, making you go up and down the pdf to figure things out. I also don't know what things like 'need to add 6 grd tp' mean... Seems like "we still need to add 6 ground test points", but that seems like instructions for someone who is still building the board.
The other thing that confuses me is the fact that they all say TRUE... Does this mean every single rail is expected to be in a TRUE state? (voltage on everything except if name has _L?)

I always see Louis Rossmann skip that page as if it's not useful, but if it's a list of all the rails and where all the voltage should be, I think it's worth looking into, or at least understanding better then deciding if I need it or not.