Announcement

**BillyJ** · 11-01-2021, 08:48 AM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

Trying to figure out why the S3 rail is dead.

U7400 is not switching the FETs
Makes sense given PP5V_S3 is 0V

Looks like PP5V_S3 is created from PP5V_S4 and the PP5VS3_EN signal
PP5V_S4 = 0V

Looks like PP5V_S4 is created from U7501 using PPVIN_S5_HS_COMPUTING_ISNS
PPVIN_S5_HS_COMPUTING_ISNS = 12.683V
P5VS4_EN = backside of PCB, cannot measure

What logic drives these power states (power button, timeouts, etc?)

**mon2** · 11-01-2021, 08:50 AM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

hi.

see here:

https://www.logi.wiki/index.php/A1398_Common_issues

**BillyJ** · 11-01-2021, 09:00 AM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

Thanks, saw that 😊

None of the issues there have similar symptoms or match my logic board part number unfortunately.

**BillyJ** · 11-01-2021, 04:37 PM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

Further to my question about the computer states, I found the following:

Computer is off - S5
Computer is hibernating - S4
Computer is sleeping - S3.
Computer is on and working, S0.

So given the S5 state is for when the computer is "off", it seems that my measurements are in line with it being off...

but what drives the state change? Which signal is not getting where it needs to, to stop the computer turning on?

For info, I took the PCB out and connected the MagSafe to confirm the following:
P5VS4_EN = 0V
PP5V_S5 = 5.0188

**reformatt** · 11-01-2021, 10:19 PM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

Start with simple stuff first. Closely examine all sections of the board for any corrosion etc. Boards have to come out for that, as often any liquid damage is on the side facing the keyboard. It doesn't need to be big either. I've seen the smallest green tinge on a component, only to find that it's either shorted (if it's a cap), the end cap of the SMD has corroded off, or the pad/track/via it sits on is eaten away. Usually its not obvious till you examine it under a microscope.

The big coils on the board are part of buck converters, so a quick run around with a meter to see if any are shorted to ground.

For in depth troubleshooting, you'll need to study up on MAC power sequencing signals from the SMC and the PCH. SMC_PM_G2_ENABLE and PM_SLP_S4 for example are signals that are enabled after certain pre-requisites are met. Without that latter signal, you won't see any S4 power rails or later. You can also find the rails are coming up momentarily and then shutting down, because of a short etc.

**BillyJ** · 11-02-2021, 02:47 PM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

@reformatt, thank you for your comment! Was very helpful

Cannot see any corrosion and have no reason to believe this is a result of corrosion as no liquid was spilt.

None of the inductors I can reach on the topside are shorted to ground.

I googled the MAC power sequence and found this useful state change diagram:
https://logi.wiki/index.php/G3_to_S3..._and_Broadwell

PP3V3_SUS = 3.3275 V
PM_RSMRST_L = 3.2982 V
PM_PWRBTN_L = cannot measure accurately/without risk of damaging PCB, component too small
PM_SLP_S5_L = cannot measure accurately/without risk of damaging PCB, component too small
S4_PWR_EN = 0 V even when the keyboard is attached and the power button is pressed.

SMC_ONOFF_L triggers correctly when I press the power button.

I'm getting the board under a thermal camera tomorrow but suspect it will show nothing concerning as the current draw from mains is 10mA (@230V)

Thanks again for your help so far.
As far as I'm aware, next steps are check bits between SMC_ONOFF_L and S4_PWR_EN to see where the signal is stopped.

**BillyJ** · 11-02-2021, 04:24 PM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

and I dug deeper...
Found the startup power sequence and took the following measurements:
S5_PWRGD = HI
PM_DSW_PWRGD = HI
ALL_SYS_PWRGD = LO
PM_PCH_SYS_PWROL = LO
PLT_RESET_L = LO

So trying to understand what drives ALL_SYS_PWRGD to HI. Looks like U8160 voltage monitor?
P1V05S0_PGOOD = LO
P1V5S0_PGOOD = LO
Or the S0 Rail PGOOD (BJT Version)
VMON_Q2_BASE = LO
VMON_Q3_BASE = LO
VMON_Q4_BASE = LO

Also looks like PP1V5_S0 is needed to keep the N channel FET Q8151 from pulling down ALL_SYS_PWRGD. There is 3.3V at PM_1V5_PGD_L which I think is the source of ALL_SYS_PWRGD being low.
U7810 creates 1.5V S0
P3V3S0_P1V5_S0_EN = LO
reminder, PP3V3_S5 = 3.3V

P3V3S0_P1V5_S0_EN ios created by the AND gate U8180 and is essentially driven by PM_SLP_S3_L
PM_SLP_S3_L = LO (and can measure the 100k pulldown with POCB powered off)

I can't see where PM_SLP_S3_L is created and what drives it so advice would be great here... looks like it's from the system power management (U1100)

**BillyJ** · 11-02-2021, 06:14 PM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

Also found somewhere online explaining the 11+1 requirements for PM_SLP_S4_L = HI.

I measured all those (below) and odd that PM_BATLOW_L = 0V as there is no battery present...

1. PPVRTC_G3H = 3.38v
2. RTC_RESET_L = 3.3218
3. PCH_SRTCRST_L = 3.3215
4. PCH_INTVRMEN_L = 3.23
5. PCH_DSWVRMEN_L = 3.23
6. SYSCLK_CLK32K_RTCX1 = Pending measurement
7. PP3V3_S5 = 3.327
8. PM_DSW_PWRGD = 3.41
9. PP3V3_SUS = 3.327
10. PM_BATLOW_L = 0
11. PM_RSMRST_L = 3.3
12. PM_PWRBTN_L = 0 (but SMC_ADAPTER_EN = 3.38V)

**reformatt** · 11-03-2021, 06:26 AM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

Forget about ALL_SYS_PWRGD as this is much later in the power sequence. At a minimum, you need all the G3H and S5 rails before you can proceed to S4. If you are not getting PM_SLP_S4, anything post S4 state is irrelevant and narrows down your search somewhat.

Re liquid damage, it's common for inexperienced to identify that with some major spillage incident. It's often the reverse, it can be but a droplet that causes the problem. Pretty sure I saw a bug take a dump on a PCB once on a Paul Daniels video some time back. So corrosion can come from many sources. I had a dead ASUS lappie a while back which had two traces eaten away from the smallest of liquid damage, and this could only be seen under the microscope.

Another thing to consider is that a lot of signals like PM_SLP_S4 go off board to daughterboards like the RIO. Had an iMac a while back which had a shorted cap on a headphone jack, killed the whole system. So remove non-essential flexes to see if your problem disappears.

**BillyJ** · 11-03-2021, 02:47 PM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

really hoping some bug shit induced corrosion isn't my issue here haha but fair point well made about difficult/impossible to see corrosion.

I've had all the other boards and looms removed, only the logic board and MagSafe adaptor attached.

G3H:
PPBUS_G3H = 12.685V
PPDCIN_G3H = 16.721V
PP3V42_G3H = 3.4132V
S5:
PP5V_S5 = 5.0194V
PP3V3_S5 = 3.3281V

All the 11+1 requirements present EXCEPT PM_BATLOW_L are present.
(Haven't managed to measure the clock yet but will be able to tomorrow)

I'm using a genuine 85W MagSafe 2 charger and upon closer inspection, I get 0.316V when battery and charger are connected OR charger only. 0V when only battery connected.

With power off and no battery, I can measure 1K Ohm to ground.
around 10K Ohm to PP3V3_S5 (as expected but also the measurement moves a lot as it's in-circuit measurement)
0.84V diode drop (1.6V drop in reverse polarity with pos cable on GND)
1Kohm to ground

Assuming the R1240 is still 10K ohm, I'm sinking 0.3mA somewhere to pull PM_BATLOW_L down to 0.316V

At what point is it worth throwing the PCB in an ultrasonic cleaner in IPA?

**reformatt** · 11-03-2021, 05:03 PM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

PM_BATLOW_L is generated by the SMC, based on comms to the ISL6259 charger IC. If no battery is connected, then SMC should assert high on PM_BATLOW_L. It'll establish that by comms with the ISL chip via the SCL/SDA data lines.

First off, make sure the SMC supplies are ok, and it is coming out of reset by checking SMC_RESET_L. Try SMC Bypass too, then check the IS6259 area as this is a common failure in MAC's.

Ultrasonic will likely introduce issues to the board if you don't do it right.

**BillyJ** · 11-03-2021, 05:58 PM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

I'll hold off on the ultrasonic bath for now.

SMC_RESET_L = 3.3V
PP1V2_S5_SMC_VDDC = 1.2V
PP3V3_S5_SMC_VDDA = 3.41V
PP3V3_S5_AVREF_SMC = 3.3V
SMC_WAKE_L = 3.125V
GND_SMC_AVSS = <0.5 Ohm to GND

SMC Bypass did not spin up fans.
I connected both fans and the keyboard. Held the on button (which I have confirmed works) and connected magsafe. Did this both with and without the battery connected and nothing from the fans even after 15s of holding the switch after magsafe.

I still haven't been able to check the 32Khz clock yet so keen to see that too but visually the crystal next to the SMC device looks OK.

ISL6259 area looks OK but photo attached

I'm still suspicious of the 0.3V BATLOW signal. makes sense given the 10k pullup and 1k I measure to ground but is 1k a typical measurement?

Attached Files

**BillyJ** · 11-07-2021, 08:26 AM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

SMC 32khz clock is indeed 32khz

Goes high on power up, approx 22mS delay then starts clocking at 32khz
Pic attached. Seems like normal behaviour.

Bat low signal starts to climb for approx 300uS and is quickly grounded. Not sure what signals are used in that logic so unsure what else to put on the other channels for timing etc.

Attached Files

**BillyJ** · 11-07-2021, 09:15 AM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

Ok wtf.

Was playing around measuring Q3000 as it's the only other device apart from the SMC that interacts with PM_BATLOW_L and started seeing a voltage on that line which surprised me.

Long story short, the batlow signal was having me over and I now get S0 voltage. No idea why.
Not sure what to do now…

Ultrasonic clean in deionised water and IPA then rinse in IPA, oven dry and reassemble?

Genuinely unsure how to take this forward.
Will measure all the S0 rails and update.

**BillyJ** · 11-07-2021, 09:29 AM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

PP5V_S0 = 5.049
PP3V3_S0 = 3.326
PP1V5_S0 = 1.515

For info I've attached the rising batlow voltage. You can still see the dip which I assume is where the SMC checks or something but it “passes” and voltage rises.

Attached Files

**BillyJ** · 11-15-2021, 03:21 PM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

Cleaned the PCB and dried in an oven at 60degC for a few hours and it worked perfectly.

Problem solved.

Used it for about an hour and then went to unlock from sleep where it died and I got 3 beep followed by 5 second pause on repeat. A quick Google shows this is a System integrity check failure - typically ram.

Any thoughts?
Feels like back to square one.

**reformatt** · 11-16-2021, 06:52 AM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

Check your rails for shorts. I've been led up the garden path with the POST beep codes, only to find it was a short on a ancillary power rail.

**BillyJ** · 11-16-2021, 03:59 PM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

Was preparing to start exactly that but it was working this AM again.

Left the cover off the base for the day, gave it some hot air gun action in case there was some condensation under an IC or something and it's working fine now.

Expecting it to fail again at some point haha and hopefully getting a new 14” MBP in Jan.

Thanks for your help with this issue. Glad I managed to narrow it down and fix.

**rotceh_dnih** · 03-28-2022, 03:18 AM

Re: Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

Sorry to thread dig this one just thought I'd share some info i found while working on a 820-3332-a with onboard gpu, These boards "mine and maybe yours if your reading this" will not boot without a good battery connected! - i have read the wiki and seen it should with the trackpad disconnected but here we are.

Let me say atleast mine wont, Story time...

Customer bought in one liquid damaged a1938 for repair, soldered new bits anywhere i seen green or had any 2'nd thoughts about "also did the gpu buck while there" then plugged it into my bench supply and it only pulled 0.01a no normal boot but would boot in SMC Bypass mode.. i checked everything many times over.. going crazy ast this point and was about to swap the smc or try a clean bios then it hit me to try a new battery and here we are weeks later and this mac is still running like a champ!

Now i have seen this once before with a A1502 no matter what i did it was dead then i put a battery on and away we go..

So as Im writing this I'm 2'nd thinking myself that something else inst wrong with these boards maybe something todo with battery detect batlow or something regardless if your bashing your head on a macbook that isn't making much sense then try a new battery

Announcement

Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

Troubleshooting dead Apple MacBook Pro A1398 (820-00426-A)

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment