2012 MB Air 820-3208-A voltage fluctuations

[da_muffinman]

This one is stuck in the perpetual boot loop (fan spins a few seconds, shuts off, rinse & repeat). Charger light comes on / no chime / no damage I have discovered yet.

I'm sorry I am a noobsicle

pls see attached

 

[2informaticos]

You make dificil to follow the usefull information.
Please don't post sheets like that, I will not be the only one doesn't read it...

CPU core pulses?
Which voltages are steady there?
Any SUS, S5, maybe S4 steady?

 

[da_muffinman]

My apologies.

Rails with stable voltages include but are not limited to:
PP5V_S5
PPBUS_G3H
PP3V42_G3H
CPUIMVP_AXG_PGOOD

Rails with fluctuating voltage include but are not limited to:
PP3V3_S0
CPUIMVP_VR_ON
PP5V_S0
PP5V_S3
(ALL_SYS_PWRGD)

Rails reading 0 voltage include but are not limited to:
PPV3_SUS
PP3V3_S3 through S5
PP5V_SUS
PPVCORE_S0_CPU
P3V3S3_EN
PP1V8_S0
PP1V05_S0
P5V3V3_REG_EN
CPUVCCIOS0_PGOOD
(ALL_SYS_PWRGD)

ALL_SYS_PWRGD shows 0 to 3V fluctuation on pin2 of R7962, however that voltage does not continue to U7960 or anywhere else on the board where I have measured ALL_SYS_PWRGD; instead, other ALL_SYS_PWRGD points read 0V.

 

[2informaticos]

"ALL_SYS_PWRGD shows 0 to 3V fluctuation
other ALL_SYS_PWRGD points read 0V"
You MUST get same value on any point of the same line!!!
Corroded trace?
BTW, fluctuating ALL_SYS_PWRGD supposes all the secondary power supplies try to appear; at least the ones involved in this signal.


"CPUIMVP_AXG_PGOOD"
How, if AXG VCORE is not even present?

You also mess many signals and voltages there.
Board will not try to start without 3V3_SUS.
Also 3V3_S5 and 5V_S4 cannot appear without P5V3V3_REG_EN.
And nothing on S3, or S0 states pulse on the board without these voltages...

Please pay more atention when you measure.
If ALL_SYS_PWRGD really pulses, check if all the big coils from the board pulse too; maybe not the two from CPU.
O-scope can capture any short pulse, which a DC voltmeter can't.

 

[da_muffinman]

"ALL_SYS_PWRGD shows 0 to 3V fluctuation
other ALL_SYS_PWRGD points read 0V"
You MUST get same value on any point of the same line!!!

Thank you - due to the timing of the voltage pulses, it seems I did indeed mis-measure some points as 0V when they are in fact fluctuating. The entire ALL_SYS_PWRGD rail that I have yet measured has the same fluctuation 0V-~3V every few seconds.

"CPUIMVP_AXG_PGOOD"
How, if AXG VCORE is not even present?

CPUIMVP_AXG_PGOOD is actually fluctuating. However, as opposed to every other yet-measured fluctuating rail, CPUIMVP_AXG_PGOOD "sits" at 1.22V for 2 seconds, then fluctuates down in voltage. The other rails sit at 0V and fluctuate up in voltage.

PPVCORE_S0_AXG is in fact missing.

Board will not try to start without 3V3_SUS.
Also 3V3_S5 and 5V_S4 cannot appear without P5V3V3_REG_EN.
And nothing on S3, or S0 states pulse on the board without these voltages...

PP3V3_SUS is a steady 3.25V at Q5040 and U6100.
P5V3V3_REG_EN is steady 3.35V. I mistakenly had it at 0, sorry! Idk how I mucked that one.

If ALL_SYS_PWRGD really pulses, check if all the big coils from the board pulse too; maybe not the two from CPU.
O-scope can capture any short pulse, which a DC voltmeter can't.

While I do not currently have access to an O-scope, I can get one if necessary. I will measure the coils with my voltmeter in the meantime and report back,

Thank you thank you!

 

[2informaticos]

Pin4 of U2750/60 pulse too?

 

[da_muffinman]

L7510
Pin1: 0V
Pin2: 0V
L7550
Pin1: flux +/- 150mV
Pin2: 0V
L7330
Pin1: flux 150mV - 1250mV
Pin2: flux 150mV - 1250mV
L7630
Pin1: flux 0V - 1000mV
Pin2: flux 0V - 1000mV
L7100
Pin1: flux 2mV - 1000mV
Pin2: flux 2mV - 1000mV
L7030
Pin1: steady 8.44V
Pin2: steady 8.44V
L7220
Pin1: flux 0V - 5V
Pin2: flux 0V - 5V
L7260
Pin1: steady 3.32V
Pin2: steady 3.32V
L9701
Pin1: steady 0V
Pin2: steady 0V
L3895
Pin1: flux .9V - 2.5V
Pin2: flux .9V - 2.5V

 

[da_muffinman]

Pin4 of U2750/60 pulse too?

PM_S0_PGOOD on U2750 (Pin4) & U2760 (Pin1) is steady at 1 mV

 

[2informaticos]

Please, stop to check voltages in both pins of a coil.
DC voltmeter will practically show the same value; guess why?
An o-scope will show you that one pin has oscillation in fact; there is no DC tension!!!

Don't forget about U2750/60.
Also check if R7402 is OK.
Can you catch a short drop 8V-0.5V-8V at CPUIMVP_TON?

 

[2informaticos]

If pins 1 & 2 of U2750 both pulse, then pin 4 must pulse too.
Check pins 1 & 2 now...

 

[da_muffinman]

Also check if R7402 is OK.
Can you catch a short drop 8V-0.5V-8V at CPUIMVP_TON?

Yes! R7402 reads:
Pin1 (CPUIMVP_TON) reads 8.36V but abruptly drops every few seconds then rises and holds at 8.36V
Pin2 (PPBUS_S5_HS_COMPUTING_ISNS) is a steady 8.44V
R7402 measures .0911 MΩ

Pin2 on R7402 has continuity to ground and Pin1 does not

 

[da_muffinman]

If pins 1 & 2 of U2750 both pulse, then pin 4 must pulse too.
Check pins 1 & 2 now...

U2750
Pin5: 3.3V
Pin1: 0-3V flux
Pin2: 0.02-0.5V flux
Pin4: steady 1mV

Does this mean U2750 is malfunctioning?

 

[2informaticos]

"Pin2 (PPBUS_S5_HS_COMPUTING_ISNS) is a steady 8.44V
Pin2 on R7402 has continuity to ground"
What do you mean???

Are you sure no pulse at L7510?
Keep voltmeter connected there more time.
See if reaches 1V sometime. I recommend to use o-scope...

 

[da_muffinman]

Pin2 on R7402 has continuity to ground"
What do you mean???

As in there is a path of low resistance (short?) from that pin to ground, ie the multimeter beeps in diode mode...I wasn't sure if that was relevant when you mentioned a short relating to this circuit.

Are you sure no pulse at L7510?
Keep voltmeter connected there more time.
See if reaches 1V sometime. I recommend to use o-scope...

My meter shows some voltage activity on this coil right when the connections are made and right when the connections are removed. When the connections are held, it reads 0V, but it keeps switching the +/- sign as if it is trying to read something. Probably my meter sucks.

 

[da_muffinman]

Could I be affecting the current by measuring it, like with quantum mechanics?

 

[2informaticos]

"As in there is a path of low resistance (short?) from that pin to ground, ie the multimeter beeps in diode mode..."
Don't tell me you are checking conitnuity on a powered board...

 

[da_muffinman]

"As in there is a path of low resistance (short?) from that pin to ground, ie the multimeter beeps in diode mode..."
Don't tell me you are checking conitnuity on a powered board...

oops...it won't happen again! I'll chalk that L up to the cost of learning this trade...and me being an idiot...and this being my first board...let me just replace that meter...needed a better one anyway...nothing to see here...

 

[da_muffinman]

Ok I got my hands on a proper multimeter...no O-scope yet but I'm hoping we can make some progress using the MIN MAX recording function on this Fluke 87 III...the response time is 250 µs or100 ms...is that passable for now?

Are you sure no pulse at L7510?
See if reaches 1V sometime.

I am reading minimal pulse at L7510: it fluctuates +/- 0.2V on 100ms detection and +/- upwards of 8V on 1ms detection

With these better readings, it is now evident that pin 4 of U2750 and U2760 is indeed pulsing

U2750 & U2760 Pin 4 (100ms mode) +/- 0.4V
1ms mode +/- 10V

They're not pulsing as much as most the other pulsing components.

U9410 and U9420 are hot like temp-wise

Could it be related to the clock chip U2700?

Sorry my skills are bad but I am learning lol I'm in it to win it - what should we do next?

 

[da_muffinman]

I should reiterate this is a quarter fan spin issue

 

[da_muffinman]

Should I go ahead and replace the clock chip just to see? Maybe I'll just set this one to the side for a bit and stays board? I know this problem in particular can be a challenge to diagnose I could just return to it when I'm more well versed but any suggestions welcome