A1708 820-00840-A Needs SMC reset to start macbook after lid_close

[jhk753]

Hello,

I'm new to macbook repair but after spilling some water on my macbook I did an ultrasonic bath of the motherboard already.
Everything works fine except after closing lid.
From the schematic, I see that there is 2 hall sensors, but playing with magnets, I can put the mac to sleep only with 1 magnet triggering the sensor on the left.

I've been wanting to dig in a bit by playing with TP for sensor and lid_open but I'm not sure how to do that cause the test point are not visible when the MB is fixed to the case. Should I just try to plug the battery to the motherboard and nothing else so I can play with hall_sensor TP and Open_Lid TP?

Also just to make sure I'm on the right track, the way I understand it:
CLAMSHELL should be detected as closed if I have LOW signal on both hall_sensor_right and hall_sensor_left, correct? Meaning if no magnet is applied, those hall_sensors should yield "high" voltage?
CLAMSHELL should be detected as OPEN if hall_sensor_right OR hall_sensor_left is HIGH so that LID_OPEN is high?

The fact that I can trigger sleep mode with 1 magnet on the left makes me think that somehow SMC_LID_RIGHT is already low... but what I don't get is, when I remove the magnet, HALL_SENSOR_LEFT HIGH should be enough for the OR gate to yield LID_OPEN.
(Just want to make sure I'm understanding correctly how this should work)

Additional information: I ran ioreg -r -k AppleClamshellState with an external monitor plugged in and it confirms.
Everything starts with "AppleClamshellState" = No
Then if I close the macbook "AppleClamshellState" = Yes
Then if I re-open the lid, it stays "AppleClamshellState" = Yes

Final question, is it possible that the problem doesn't come at all from the hall_sensors or the logics around it? (One repair store told me I needed to change my screen cause the video card might be faulty... But I fail to see how a faulty screen could explain what I'm experiencing with magnets, etc)

Thank you

 

[2informaticos]

First of all, welcome to the forum!

As per schematic, right sensor is placed on audio flex and I bet was affected by liquid spill.

Seems that you've understood how U5255 works.
One high level sensor is enough to get high level output.
It goes low only if both inputs are low too; check associated note.

Remove R5257 and tie together U5255 inputs.
If machines works good, leave it as is.

 

[jhk753]

Thanks. I'll see how I can try that. (don't have much experience with micro-soldering etc). If there is anyway to check/test the audioflex sensor, it might be less complex for me without micro-soldering to just change the audioflex completely?

In the meantime if I look at schematic, I'm not sure of one thing though. If my left sensor is working correctly, it should be "enough" to wake the machine right?

Because if I'm understanding correctly what you're trying to suggest, you want to remove signal coming from hall_sensor_right and have SMC_LID_RIGHT = SMC_LID_LEFT_R

So I'm not sure exactly what is the reasoning behind it, if you care to give me more insights
Cause I also tried to unplug the audio flex and nothing happened. (probably rightfully so, cause if the expected behavior is to have current in SMC_LID_RIGHT when OPEN, unplugging the audio flex would not change that.

I guess my question is, do we need SMC_LID_RIGHT to be "high" for the computer to "WAKE"? ==> if SMC_LID_LEFT_R is high and LID_OPEN is HIGH (thanks to the or logic gate), it's not enough to wake the machine?

 

[2informaticos]

"I guess my question is, do we need SMC_LID_RIGHT to be "high" for the computer to "WAKE"? ==> if SMC_LID_LEFT_R is high and LID_OPEN is HIGH (thanks to the or logic gate), it's not enough to wake the machine?"
Both hall sensors are connected to SMC too; not just to U5255.

"Cause I also tried to unplug the audio flex and nothing happened. (probably rightfully so, cause if the expected behavior is to have current in SMC_LID_RIGHT when OPEN, unplugging the audio flex would not change that."
The pull-up resistors are on the sensors side, not on the MLB.

Of course, correct repair requires changing bad hall sensor...

Write directly on the reply box, instead of clicking Reply button.
I don't see any reason to quote an entire post, visible few centimeters above.
Please, maintain forum aesthetic.

 

[jhk753]

Sorry for the esthetics.
"Both hall sensors are connected to SMC too" => Ok so you are saying that it's expected that we need 2 hall sensors with high signal to wake the computer right?
Can I try the fix above without removing anything and just connecting TP SMC_LID_RIGHT with TP SMC_LID_LEFT? If it works does it mean I can probably just change the audio flex (the problem would be isolated to it?)

Thank you

 

[2informaticos]

Corrosion on right hall sensor will kill then HALL_SENSOR_LEFT signal too.
You must remove R5257 to prevent this.

 

[jhk753]

Ok thanks I'll see what I can do

 

[jhk753]

Cause I don't have experience with micro soldering and don't want to mess it up, can I actually unplug the corroded right sensor by unpluging the audio connector on J6600 and then connecting the 2 TP TP SMC_LID_RIGHT with TP SMC_LID_LEFT ? Can this work to isolate the problem?

 

[2informaticos]

I already explained you, pull-up resistor for HALL_SENSOR_RIGHT signal is placed on the sensor side.
If you disconnect it, need to solder such resistor on the MLB, to 3V3_G3H; similar with R5253, used for HALL_SENSOR_LEFT.

Don't need special soldering experience to desolder R5257; maybe to put it back.

 

[jhk753]

Hi,
So update on this. I have tried with a brand new audio flex thinking the right sensor was faulty. But no improvement at all. I still can put my mac to sleep only with left sensor… as if I haven’t changed anything.

I did a few measure and I do have 3.3v on TP LID_OPEN and TP SMC_LID_LEFT.
With old and new audio flex (only the flex and the board connected with nothing else), I get 0.03v on TP HALL_SENSOR_RIGHT.

There is still a chance that the new audio flex is broken as well, but that led me to think I'm not sure I understand what voltage I'm supposed to get from this sensor.
Basically when there is nothing connected but the board and the audio flex (+power of course), am I supposed to get high voltage on TP hall_sensor_right (similar to what I'm getting from hall_sensor_left) if it's working properly?

By the way R5254 is supposed to be 1M Ohm but I measure only 200 to 300k... weird. Not sure it impacts anything but I’d rather say in case

PS : I understand the fix about desoldering R5257 but as I'm looking for the long term solution with both sensor playing their part, I'm trying to understand what is wrong with this line first.

 

[2informaticos]

Did you locate the right hall sensor on that flex?
It receives 3V3_G3H voltage and has a pull-up resistor there too.

If you changed the audio flex and issue persists, check J6600 solder on pin 25.

 

[jhk753]

Ok so I do locate the sensor on both flex, but it's covered/protected in some sort of silicon. so I can't really access it, without removing it. I'll try to do that on the old one just to see if I can measure stuff there directly.
Other than that I did measure voltage on J6600 pin 27 and pin 25 doesn't seem broken or anything from the look of it.

 

[2informaticos]

A pull-up resistor (mounted on the sensor), must bring 3V at pin 25; powered from 3V3_G3H at pin 27.
If you get 3V on both pins, sensor should be good.
You can check its reaction aproaching a magnet; pin 25 should drop to 0V.
This signal goes to U5255 and SMC via R5257.
If this resistor is broken, or sensor corroded, it will be always 0V.

 

[jhk753]

any tips to measure those voltage when the audio flex is connected? Cause there is no tp and the pins are not easily accessible…

when the audio flex is not connected, I do get 3v on pin 27.
Regarding pin 25, why measuring from it rather than TP hall_sensor_right should be different? There is 0 ohm between them (measured when audio flex not connected)

i find it very weird that I just bought a broken audio flex. Is there any remote way that the pb can be caused by something else elsewhere?

ps : I also made sure there is no short by measuring resistance from tp_hall_sensor_right and ground base. Doesn’t seem like a short

 

[2informaticos]

I mentioned pin 25, to know what signal to check on MLB; doesn't matter, where, if is the same line.
You can test the board alone on the desk, just audio cable connected; and charger, of course.

Pull-up resistor is high value (100K-1M), similar to R5253.
Even slight corrosion to ground, will kill the signal.

 

[jhk753]

Ok that’s what I thought. Do we agree that if there was corrosion to ground I would measure 0 Ohm between hall_sensor_right and ground? If that’s not the case we can safely assume it’s not a short, correct?
Or is ground not the same circuit accross the MLB?

 

[2informaticos]

"Do we agree that if there was corrosion to ground I would measure 0 Ohm between hall_sensor_right and ground?"
NO!

If pull-up resistor is 100K, parasitic 10K resistance to ground (caused by corrosion), will drop voltage to almost 0V.
Even 50K will drop signal to aprox 1.5V, which is prohibited level; 2.1V is low limit for high level.
If pull-up resistor is open (corroded), no voltage at all there; without detecting any short, or low resistance to ground.

 

[jhk753]

Ok but I’m measuring 300k ohm… (even though I should measure 1meg based on the schematic)
And I tried linking tp hall_sensor_right + tp hall_sensor_left and the signal doesn’t die and stays at 3v on both tp (Even with audio flex plugged in and touching no other resistors)
So what do we conclude? Do I have corrosion? Or just 2 non functioning sensor?

 

[2informaticos]

300K between pins 25 & 27?
Check value on both directions.
If pull-up resistor is 1M, the reading could be good.

I already explained you how to check if original right sensor works.
Why don't check???

 

[jhk753]

"I already explained you how to check if original right sensor works.
Why don't check???"

===> I'm under the impression that I've checked everything you shared already. please tell me, what did I miss? What haven't I tested already?

By the way I tried with 2 other audio flex (so 3 new in total) and they hall yield the same problem. No voltage on TP sensor_right. But strangely if I bootup my computer mic is not working on any of the new audio flex but is working on the old supposedly broken one.

So either the 3 new audio flex are broken OR there is something weird happening with the Motherboard where with a working audio flex, nothing works...