iMac A1311 - Motherboard 820-3126-A, turn on sometimes

[ryogadan78]

Hi guys, I have this 2011 iMac A1311 with a strange problem.

Most of the time, when I plug in the power cord, the first of the 4 LED's turn on, but when I press the power switch the LED turns off and iMac won't boot up.
Sometimes, when I unplug and plug the power cord back in, iMac starts up correctly by turning on the 4 LEDs in sequence.

I have already tried a new power supply, but the result does not change. Do you have any suggestions?

Motherboard 820-3126-A

[mon2]

Remove all power. Meter in resistance mode. Measure the resistance to ground of each inductor. They are usually grey in colour. Checking for a low resistance which means a shorted cap is onboard. Common to see at this age of a machine. Post your measurements.

[ryogadan78]

Hi mon2, glad to see you again.
These are the measurements, the boardview also helped me to identify aluni inductor.
The other measurements are above 1kohm

L7850 290ohm (PP1V8_S0)
L7710 95ohm (PP3V3)
L7750 14Kohm (PP5VS3)
L2406 15ohm (PP1V05_S0_PCH)
L7301 9ohm (PPVCORE_S0)
L7420 15ohm (PP1V05S0)
L7830 490ohm (PP1V5_S3)

[ryogadan78]

Hi Mon2, is there any possibility to inject a small current on some inductor and evaluate if any capacitor heats up? I have at my disposal a thermal camera

[mon2]

L7850 290ohm (PP1V8_S0)
L7710 95ohm (PP3V3)

These 2 look a bit low for their resistance to ground. However, voltage injection may not smoke them out. You can try and see if this heats up the PCH which means the CPU is dead.

The power rail created by L7710 is used to power PP3V3_S5_PCH. At 95 ohms, the PCH is probably ok but the resistance to ground is too low for this power rail.

[ryogadan78]

Hi, I injected a small current on the two inductors. 1.8V and 2.3v ... absorption between 0.05 to 0.28 amps, I don't see any particularly hot components on the motherboard. I tried repeating the test by turning the bench power supply on and off many times, considering that the original problem is not always present, but nothing changed.
How do I find the problem of this motherboard occasionally working and often not working?

[mon2]

Those rails do not look to be shorted with the observed current draw. No heat will be generated @ 280 mA.

Review sheet #3 of the schematics which offers the 'power block'. Start with the root voltage from the external power supply. Be very careful here due to the open frame power supply design. A shock can be lethal. With no power to the unit, apply a cardboard protector onto the power supply connections to insulate this shock hazard.

Then confirm that you have a stable 12v each and everytime you power on.

[ryogadan78]

Hi mon2, I followed your advice and connected the power supply safely (see the picture). When I connected the power plug, the first of the four LEDs lit up. I measured at this time the PP12V_G3H on C8111 and it was stable at 12V. I then turned on the iMac through the switch pins and, as often happened, the first LED went out. I measured PP12V_G3H on C8111 again and 12V was still present. I checked the mainboard and nothing seems to be overheating at this stage
Thank you again for your very kind help

,

[mon2]

The LED that is turning off is LED501 ('It is plugged in') and is driven by PP3V3_S5.

Remove all power. Meter in resistance mode. Measure the resistance to ground (not diode mode) of PP3V3_S5_REG @ L7710 (pin 1 but either side should be ok). Post the measurement.

We are checking if there is a shorted cap on this power rail which is being sensed as an excessive current draw by U7700.

[ryogadan78]

OK, here I am.

When I measure the resistance to ground on L7710 by disconnecting all voltages, the value is 95ohm.

If I repeat the measurement by connecting the voltage, with initially the first LED on, then activating the power switch, after the LED goes off, the value drops to 15ohm

[mon2]

These are low readings for a power rail.

Would you have an adjustable power supply? Use it at 1 volt with 3-5A current to inject onto the same inductor at the output side.

Keep the voltage low but current at max. No power to the board. Hunt for any part that should heat up. That part needs to be removed and/or replaced.

[ryogadan78]

Hi mon2, yes I have an adjustable power supply. I set it to 1.5V, max ampere (5A).

I injected the current on pin 2 of L7710, sometimes the current drawn is 0.05A. Sometimes, switching the power supply off and on, the current drawn is 0.60A. In neither case I can detect any component overheating. I have tried both with the thermal camera and by touching the mainboard with my hand. The thermal camera shows me only slight heat (26 degrees against 23 for the rest of the board) in the area of the U8100

[mon2]

Can you confirm the location of U8100 on the 820-3126-A logic board? Unable to locate this component in my database of schematics and boardview drawings. Do not see U8100 onboard.

[ryogadan78]

I apologise. Sometimes my dyslexia comes back to visit me. I meant U1800

[WTFRules!]

Hi! Have you made a visual inspection looking for missing components?

[ryogadan78]

Hi wtfrules,
thank you for your support. I also visually analysed the board under the microscope and found nothing abnormal. No overheating or oxide or overvoltage damage.
This mainboard also does not always have the problem, sometimes it works correctly.

[amennone]

Are you sure that U1800, INTEL SLJ4F BD82Z68 BGA Chip Chipset, does not overheat? I have the same problem, powering the board with a bench power supply, PP12V_G3H_ACDC, I have a current draw of about 400 milliamps and U1800 heats up about 40 degrees Celsius.
L7710 to gnd I measure 45 ohm

[dqtech]

Hi mon2, glad to see you again.
These are the measurements, the boardview also helped me to identify aluni inductor.
The other measurements are above 1kohm

L7850 290ohm (PP1V8_S0)
L7710 95ohm (PP3V3)
L7750 14Kohm (PP5VS3)
L2406 15ohm (PP1V05_S0_PCH)
L7301 9ohm (PPVCORE_S0)
L7420 15ohm (PP1V05S0)
L7830 490ohm (PP1V5_S3)

hi sir...have you solve the issue?...i got the problem led1 only blink once...the measurements exactly same like this except for L7750, only got 625ohm

[dqtech]

i dont know if this happen to others.....i try to resolder U7700 and now imac work normally...maybe it only cold solder joint?