Acer Aspire LA-7912P Rev.1.0 - brak zasilania SPI, diody LED włączają się na ok 3s

Cześć.

Mam na stole starego laptopa Acer Aspire (nie wiem dokładnie jaki model, bo wszystkie naklejki już dawno zniknęły). Chciałbym go uruchomić, natomiast on nie bardzo tego chce — włączają się diody, cooler delikatnie drgnie i po około 3 sekundach wszystko gaśnie.

Płyta główna to LA-7912P Rev.:1.0

Napięcie na głównej linii zasilania = 19,25V (podawane z oryginalnego zasilacza)
Napięcie na 8 nodze SPI (U36) = 0V (w ciągu tych 3s pozornego uruchomienia także). Rezystancja, względem masy, na tej nodze = 40kΩ

Pomierzyłem także cewki. Wyniki podaję w kolejności - oznaczenie cewki, rezystancja względem masy, napięcie na cewce przy wyłączonej płycie, napięcie na cewce przy pozornie włączonej płycie:
- PL9, 10Ω, 0V, 0V
- PL22, 1,1Ω, 0V, 0V
- PL21, 1,4Ω, 0V, 0V
- PL20, 1,4Ω, 0V, 0V
- PL14, 6,8Ω, 0V, 0V
- PL25, 79Ω, 0V, 0V
- PL7, 35Ω, 5V, 5V
- PL6, 570Ω, 3.3V, 3.3V
- PL15, 5Ω, 0V, 0V
- PL4, R≥20MΩ, 3V, 3V
- PL27, 29Ω, 0V, 0V

Martwi mnie brak zasilania SPI i jednocześnie zastanawia, że rezystancje na prawie wszystkich cewkach są dość niskie.

Co mogę sprawdzić, aby spróbować zdiagnozować tę płytę?

bpjl napisal/-a:rezystancje na prawie wszystkich cewkach są dość niskie.

Hello!
The resistance measurements on the coils must be wrong, or you have a damaged meter.

Did you measure the resistances with the power supply and main battery disconnected from the board? If so, check your measurements with another meter, or at least replace the batteries in your meter.

matic napisal/-a:The resistance measurements on the coils must be wrong, or you have a damaged meter.

It also seemed suspicious to me, that's why I measured several times. And it didn't want to come out differently. Now, after these a few hours, I measured them all again and a few differences came out:
- PL9, 432Ω
- PL22, 1,1Ω
- PL21, 1,4Ω
- PL20, 1,4Ω
- PL14, 6,8Ω
- PL25, 79Ω
- PL7, 12kΩ
- PL6, 1,4kΩ
- PL15, 5Ω
- PL4, R≥20MΩ
- PL27, 103Ω
What could such a situation come from?

matic napisal/-a:Did you measure the resistances with the power supply and main battery disconnected from the board?

Both power sources were and are disconnected.

I must add that with these previous measurements I got advice to look around PL7 because it is far too low resistance. Before I measured the coil again, I checked the PQ17 and PQ19 transistors:

PQ17 - only 1.7Ω in both directions between the gate and the source. In the diode test, it also shows a short circuit.
PQ17 - high resistances everywhere. Diode test - it's hard to say. Once a full short-circuit between drain and source. But only one way. After a while, with the next measuring short circuit there is no longer. As if the transistor "opened" with measuring and "closed" in a moment.

bpjl napisal/-a:It also seemed suspicious to me, that's why I measured several times. And it didn't want to come out differently.

What about another multimeter?
Before processing, we must make sure the measurements are correct, as further diagnosis is based on them.

matic napisal/-a:What about another multimeter?
Before processing, we must make sure the measurements are correct, as further diagnosis is based on them.

It took a while but I managed to give a different multimeter. I compared the readings and coincide with these:

bpjl napisal/-a:- PL9, 432Ω
- PL22, 1,1Ω
- PL21, 1,4Ω
- PL20, 1,4Ω
- PL14, 6,8Ω
- PL25, 79Ω
- PL7, 12kΩ
- PL6, 1,4kΩ
- PL15, 5Ω
- PL4, R≥20MΩ
- PL27, 103Ω

Measure the power consumption (in mA) of the board before and after pressing to the power button without a battery connected.

The switched off board takes 7mA. After turning on the consumption increases to 16ma. When the board goes out, the current remains at 14-15mA.

I borrowed, for this afternoon, the same, working board, with a working CPU. I measured the coils and parameters on the 8th leg to have a comparison.



An working board has no dedicated GPU therefore there is no PL25 and PL27. SPI voltage appears when the board is turned on.

- Measure the voltages to ground on pins 67, 37, 101, 110, 115, 124 of KBC chip.

- Measure the voltages to ground on pins 106, 114 of KBC chip in the following cases:
a) Power button released.
b) Power button pressed.

Now disconnect the power supply, wait about two minutes, reconnect the power supply and perform the following measurements:
- Measure the voltages to ground on pins 6, 14, 16, 95, 100, 112, 116, 127 of KBC chip in the following cases:
c) Before pressing to the power button (machine turned off).
d) After pressing to the power button (when the power LED is turned on).

matic napisal/-a:- Measure the voltages to ground on pins 67, 37, 101, 110, 115, 124 of KBC chip.

matic napisal/-a:- Measure the voltages to ground on pins 106, 114 of KBC chip in the following cases:
a) Power button released.

matic napisal/-a:b) Power button pressed.

matic napisal/-a:Now disconnect the power supply, wait about two minutes, reconnect the power supply and perform the following measurements:
- Measure the voltages to ground on pins 6, 14, 16, 95, 100, 112, 116, 127 of KBC chip in the following cases:
c) Before pressing to the power button (machine turned off).

matic napisal/-a:d) After pressing to the power button (when the power LED is turned on).

On another forum I was recommended to look at the +1.5VP converter:


I measured the resistance of each PIN relative to GND. Here are the results:


After publishing these results, I was ordered to replace the PU6 chip.

Now, when I had the power supply connected for a long time to make the measurements you recommended, I sensed that PU6 probably warmed up. It's not hot but just warm.

Something strange is going on with your measurements.
You measured the resistance to ground of about 400Ω on the coil PL9:

bpjl napisal/-a:- PL9, 432Ω

But you measured the resistance to ground of only 10Ω on pin 16 of chip PU6, which is connected directly to the coil PL9:

bpjl napisal/-a:I measured the resistance of each PIN relative to GND. Here are the results:
[ Obrazek ]

How is this possible, unless the copper trace between pin 16 of PU6 and coil PL9 is broken, which is very unlikely?

matic napisal/-a:Something strange is going on with your measurements.
You measured the resistance to ground of about 400Ω on the coil PL9:

bpjl napisal/-a:- PL9, 432Ω

But you measured the resistance to ground of only 10Ω on pin 16 of chip PU6, which is connected directly to the coil PL9:

bpjl napisal/-a:I measured the resistance of each PIN relative to GND. Here are the results:
[ Obrazek ]

How is this possible, unless the copper trace between pin 16 of PU6 and coil PL9 is broken, which is very unlikely?

You're right, I made a mistake somewhere. Resistance, compared to GND, on the coil PL9 = 297Ω (at this moment). However, re -measurements of PU6 look as follows:

Honestly, I have no idea what is happening here with measurements

Communication between the KBC and PCH chip is missing.
Try to reflash the KBC chip with programmer. If nothing changes, replace the KBC chip.