Macbook Pro a1502, S4to S0 power cycling every 5sec

[kamthor]

Hello
I have a a1502 820-4924A motherboard with no apparent failed components and no signs of corrosion.
I get green then orange light on a genuine 85w magsafe2 connector supplied from a bench PSU with an 80mA spike drawn every 5 seconds.
here are the readings (more or less .1v I m at work right now writing from memory)
1. PPVRTC_G3H (VCCRTC) 3.3V
2. RTC_RESET_L (RTCRST#) real-time clock circuit reset. 3.3V
3. PCH_SRTCRST_L (SRTCRST#) secondary real-time clock circuit reset. 2.8v
4. PCH_INTVRMEN_L (INTVRMEN) internal voltage regulators 1.05V and 1.5V enable. 2.8v
5. PCH_DSWVRMEN_L (DSWVRMEN) deep sleep well(power) voltage regulator enable. 2.8
6. SYSCLK_CLK32K_RTCX1 (RTCX1) 32khz clock for RTC circuit. waiting for a usb oscilloscope
7. PP3V3_S5 3.3V
8. PM_DSW_PWRGD (DPWROK) 3.3V
9. PP3V3_SUS 3.3V
10. PM_BATLOW_L 3.3V
11. PM_RSMRST_L 3.3V

All S4-S0 power rails are pulsing every 5 seconds.
the board draws 90mA for a fraction of a second every 5 seconds
checked with alcohol and thermal camera and there's no overheating elements.
there's no short to ground on any rails (CPU coils read 11ohms)

I replaced U1900 and Y1905, tried 2 other U6100 chips, reflashed with different bios bins to no avail.
removed J6100 and corresponding resistors.

all resistors around U6100 are fine, changed C7826 and C7821.

what's really weird (to me at least) is that if I remove the bios chip and power up the board all power rails are up and the board draws steadily a current ( between 190 and 500mA if I'm not mistaken)
Are the PCH voltages at 2.8v abnormal? (they are supposed to be 3.3v if i'm not mistaken)
CPU coils have 11ohm resistance.
I ordered an oscilloscope so I'll get the Y1905 readings tomorrow. I have two other spare boards to harvest parts from.
I read almost all similar threads here and over at Louis' forums and the logi.wiki but none of the cases apply here. I"m learning as I go and this is my first apple product so be kind if I made some gross mistake
Thanks in advance for any input and pointers in the right direction.
Cheers!

[mon2]

Re: Macbook Pro a1502, S4to S0 power cycling every 5sec

Measure the voltage to ground of PPBUS_G3H.

Also, remove all power. Measure the resistance to ground of each SMBUS (aka I2C) SDA / SCL line -> checking to see if any such line is shorted.

[kamthor]

Re: Macbook Pro a1502, S4to S0 power cycling every 5sec

Thanks for your reply,
PPBUS_G3H reads 12.57v
all i2C SDA/SCL could check were not shorted to ground

[mon2]

Re: Macbook Pro a1502, S4to S0 power cycling every 5sec

Try a fresh known good battery (Amazon?).

See here under battery not detected case:

https://repair.wiki/w/A1502_2013-201...ook_Pro_Retina

[reformatt]

Re: Macbook Pro a1502, S4to S0 power cycling every 5sec

Are you working on this board only? Cycling is usually a short (or missing) rail; this can also be a switched rail such as those on pg 80 of the schematic. So it's worthwhile running your meter over the outputs of each quickly to rule them out. Also check the caps above the CPU (like C1280 etc) for shorts or presence of volts during the cycling.

BIOS loading can also cause cycling due to bad or open circuit connections for the termination resistors from U6100. Meter these out point to point (e.g. from R6110 to R6120).

Also check your PCH resistance to ground; should be over 100 ohms usually when cold on L7630. Remember resistance will be less when the CPU is hot.....Should be 1.05V on this coil, which should be fine as you have CPU Vcore.

As far as corrosion goes, this can be very subtle on these boards and easy to miss. Usually from insects I find, and the corrosion is often across a bunch of pins of a QFN IC. Look at each IC under a microscope to rule this out.

[kamthor]

Re: Macbook Pro a1502, S4to S0 power cycling every 5sec

Try a fresh known good battery (Amazon?).

See here under battery not detected case:

https://repair.wiki/w/A1502_2013-201...ook_Pro_Retina

battery is charging fine, I'm working on the board outside the case

Are you working on this board only? Cycling is usually a short (or missing) rail; this can also be a switched rail such as those on pg 80 of the schematic. So it's worthwhile running your meter over the outputs of each quickly to rule them out. Also check the caps above the CPU (like C1280 etc) for shorts or presence of volts during the cycling.

BIOS loading can also cause cycling due to bad or open circuit connections for the termination resistors from U6100. Meter these out point to point (e.g. from R6110 to R6120).

Also check your PCH resistance to ground; should be over 100 ohms usually when cold on L7630. Remember resistance will be less when the CPU is hot.....Should be 1.05V on this coil, which should be fine as you have CPU Vcore.

As far as corrosion goes, this can be very subtle on these boards and easy to miss. Usually from insects I find, and the corrosion is often across a bunch of pins of a QFN IC. Look at each IC under a microscope to rule this out.

L7630 to ground resistance is 232.9Ohm, the coil is also pulsating when the board is powered.
R6110 to R6120 I have 2 Ohms, same between R6111 and R6121.
should I inject 1V on the pulsing rails?
Thanks!