820-3437-B No PP3V3_S5 or PP5V_S4RS3

[Flippers McCoy]

Hey, everyone!

I have a board that won't power on. The charging light comes on and turns to orange, but it shows no signs of life. Occasionally, when I reset the SMC or plug it in, I'll hear a shrill whine for a couple seconds. I see no visible damage on the board, save for some corrosion that cleaned up fine (on and around U3210).

My current issue, however, is that PP3V3 and PP5V_S4RS3 don't exist. This issue seems to stem from PM_SLP_S4_L not existing either, which is one of the rails responsible for enabling U7501, the IC that is responsible for creating both missing power rails.

I looked around on the schematics, and my best guess is that PM_SLP_S4_L is created in the CPU or PCH(not sure which one), suggesting that the CPU or PCH is bad.

Would someone be able to give me advice on how to confirm this, and/or explain how I can troubleshoot this?
Thank you!

P.S: I've attached a schematic and boardview to help out

[Flippers McCoy]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

Update: I found this neat Rossmann forum post that is very similar to my issue.
I get no voltage bursts on L7520 (read with my voltmeter, I don't have an oscilloscope to test with), and U1950 is getting voltage in, but it is not outputting anything through pin 7.
U1900 is outputting voltage through pin 1, so I think it should be working. I'm not entirely sure of how to properly troubleshoot an RTC chip.

Hopefully some of that info is helpful!

[piernov]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

Check SRTCRST#, RTCRST#, RSMRST#, PWRBTN#, and make sure that all the G3H and S5 rails are ok. Also check if you get a spike on PM_SLP_S4_L after plugging in the charger. If you do, check the BIOS, the BIOS resistors and the traces going to the BIOS.

Check if the liquid hit any other part.

[Flippers McCoy]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

I don't get a spike on PM_SLP_S4_L when plugging it in, although I'm only able to measure it with my multimeter, I don't have an oscilloscope to test with.

Here's what I'm reading for G3H and S5 rails, and those things you mentioned going into the CPU (could you briefly explain their significance?)

PPBUSG3_H=8.6V
PPBUS_S5_HS_COMPUTING_ISNS=8.6V
PPBUS_S5_HS_OTHER_ISNS=8.6V
PPDCIN_G3H_ISOL=18.9V
PPDCIN_G3H=18.36V
PP3V42_G3H=3.42V
PPVRTC_G3H=3.42V

PP5V_S5=5V
PP3V3_S5=0V

SRTCRST#=3.42V
RTCRST#=3.42V
RSMRST#=0V
PM_PWRBTN_L#=0V

[piernov]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

Ok so the PCH won't be able to work if PP3V3_S5 is missing, sorry I missed that in your first post. Check SMC_PM_G2_EN and S5_PWR_EN.

RTCRST# is the RTC reset signal, it tells the PCH to enable the RTC (real-time clock) functions. It is the first signal needed by the PCH.
SRTCRST# is the secondary RTC reset signal, not sure how it is different from RTCRST# but it is needed too.
Those signals should always be present if there is power somewhere on the board. Battery, AC adapter, or even coin cell on computers that have one.
RSMRST# is needed for the PCH to initialize and start managing power states. This is needed to go from S5 to S4. On newer platform there is an intermediate power state, deep suspend so this is a bit more confusing.
PWRBTN# is the power button signal, it goes low then high to tell the PCH to go to a higher power state (powering on or resuming from sleep for example). It needs to be low for a short time after RSMRST# gets high, then go high.

[Flippers McCoy]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

Ah, thank you for the good description, that makes sense!

SMC_PM_G2_EN=3.42V
S5_PWR_EN=3.42V

[piernov]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

Ok so U7501 should turn on and generate PP3V3_S5 but it doesn't. What is the resistance to ground on L7560?

[Flippers McCoy]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

~1 ohm on both sides of L7560, and when I test for continuity to ground it gets continuity on both sides as well.
I can't believe I didn't realize that it was shorted to ground!

[Flippers McCoy]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

I tried injecting 3.2V at 2 amps to PP3V3_S5 and nothing got hot or smoked. Are there any other ways to find the short to ground?
Also, would PP3V3_S5 being shorted to ground prevent PM_SLP_S4_L from being created? To me, it seems that there are two problems, PP3V3_S5 is shorted to ground and PM_SLP_S4_L isn't being created

[SMDFlea]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

Could it have anything to do with U5590

[Flippers McCoy]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

It looks like U5590 isn't included in this model, there are just bare solder pads where it should be.
(if anyone has an explanation for why a component would show up in the schematics but not on a board, I'd love one!)

[piernov]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

If PP3V3_S5 is not present, the PCH will not work. So no PM_SLP_S4_L.
You need to figure out why PP3V3_S5 is shorted.
If you don't have more than 1ohm to ground on PP3V3_S5, something has to get hot. Or your PSU doesn't deliver enough amperage.

[BlueMidnight]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

It looks like U5590 isn't included in this model, there are just bare solder pads where it should be.
(if anyone has an explanation for why a component would show up in the schematics but not on a board, I'd love one!)

Some components are intended only for development and testing. Production boards do not have them.

On other schematics, you might see something like "@" or "DY" next to such components. On boards such as this MacBook Air, you will also see things like the "P3V3S5_ISNS:YES" that is written next to U5590. This indicates that those components are populated only on boards manufactured with that option.

There is a table on page 2 listing two different versions of the current sensing circuitry on this board. "ISNS:ENG" and "ISNS:PROD". The "PROD" (production) version lacks some current sensing components that the "ENG" (engineering?) board has. You'll find that the table shows "P3V3S5_ISNS:YES" for one and "P3V3S5_ISNS:NO" for the other.

There is also a table on page 42 that shows, when the current sensing components are not populated, a 100k ohm resistor is placed where C5595 would otherwise have been.

Those little notes next to the components on schematics can be super important sometimes.

[SMDFlea]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

Thanks for the detailed explanation.

[chris-swe]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

Just something to check.
Look closely under the U1950 on pin 1 (you might have to remove it to be sure), i have had a couple of them where everything looks good but under the chip on pin 1 its burned away and you will never get PM_S0_PGOOD.
I guess you checked for corrosion on all the probe points on those rails as well.

[Flippers McCoy]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

You need to figure out why PP3V3_S5 is shorted.
If you don't have more than 1ohm to ground on PP3V3_S5, something has to get hot. Or your PSU doesn't deliver enough amperage.

When I inject 3.2V at 5.2 amps, it drags down to .66V-.88V and stays at 5.2 amps, would this be enough. It gets warm near where I'm injecting the voltage, but I think that is to be expected. I can't feel it get hot anywhere else, though. Is 5.2 amps enough?

Some components are intended only for development and testing.

Thank you so much for the great explanation! That makes a lot of sense

Look closely under the U1950 on pin 1

PM_S0_PGOOD is 0V, but ALL_SYS_PWRGD and CPUVR_PGOOD_R are also 0V, so I don't think that U1950 is necessarily the culprit there, it just not being told to let PP3V42_G3H through. Is this a good assumption?

[chris-swe]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

When I inject 3.2V at 5.2 amps, it drags down to .66V-.88V and stays at 5.2 amps, would this be enough. It gets warm near where I'm injecting the voltage, but I think that is to be expected. I can't feel it get hot anywhere else, though. Is 5.2 amps enough?

Yes, 5.2 Amps is good enough, and voltage drop would be normal as well since your psu wont deliver any mor voltage and amps since you limit it.
But its a nice good short
So follow that rail systematic and you will find the short.




PM_S0_PGOOD is 0V, but ALL_SYS_PWRGD and CPUVR_PGOOD_R are also 0V, so I don't think that U1950 is necessarily the culprit there, it just not being told to let PP3V42_G3H through. Is this a good assumption?

Ok, Then you have to check where ALL_SYS_PWRGD is coming from (Q8150). Is that turning on. If not you follow that back and so on..
You also have to do the same for CPUVR_PGOOD_R. Both of them needs to be present for it to close and let PP3V42_G3H pass to the next step witch is out of the chip and then back in at pin 7.

[Flippers McCoy]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

So follow that rail systematic and you will find the short.

How am I supposed to do this? Are you suggesting I test every component on the rail in order to find which one is shorting to ground? I was considering dousing the board in alcohol, then injecting voltage into PP3V3_S5, then seeing if any particular area dried out first due to warming up. Is this a legitimate technique?
I also checked along both ALL_SYS_PWRGD and CPUVR_PGOOD_R, and it looks like they both need S0 power rails to be active in order to be generated? I'm also unsure of why I'm checking these things to begin with, I don't see how they relate to PP3V3_S5 or U7501

[chris-swe]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

How am I supposed to do this? Are you suggesting I test every component on the rail in order to find which one is shorting to ground? I was considering dousing the board in alcohol, then injecting voltage into PP3V3_S5, then seeing if any particular area dried out first due to warming up. Is this a legitimate technique?
I also checked along both ALL_SYS_PWRGD and CPUVR_PGOOD_R, and it looks like they both need S0 power rails to be active in order to be generated? I'm also unsure of why I'm checking these things to begin with, I don't see how they relate to PP3V3_S5 or U7501

Yes, Pure Alcohol would do. I use Pure Methanol myself, but just coz i have alot of it.

Yes, investigation is a bitch. You go around and look on the board, and if you see something that look odd you remove it and check if the short is gone. And the area around U5000? looks good?

[SMDFlea]

Re: 820-3437-B No PP3V3_S5 or PP5V_S4RS3

If you remove L7560 and the short is on pin 1 ,check C7562. If the short is still there on pin 2 remove
C7590,C7592,C7572 and C7593 one at a time,check if the short has gone.