820-2850 fan spin 3 seconds and stop, then loop

[protocold]

C9560 replaced.

PPBUS_G3H = 12.5v

ALL_SYS_PWRGD is cycling between 3.3v and 0v

 

[2informaticos]

VCORE cycling too?
Did you try SMC bypass mode?

 

[protocold]

PPVCORE_S0_CPU is cycling between 1.1v and 0v.



I cant even turn it off, when i press the power off button, it doesnt do anything, all i saw was the fan spin up and down every 3 seconds.

Any other way to do SMC bypass mode without the power button?

 

[2informaticos]

You must press (and maintain) power button before connecting charger.

Is this symptom the same as before changing C9560?

Any history fault available?

 

[protocold]

i press and maintain power button, then connect charger, NO GREEN LIGHT, then release charger... same thing, fan spin up and down, the different is this time the fan spin higher and louder.

Nothing shown on the screen.

Symptom same as before changing C9560.

No history available, board looks very clean, untouched, not tampering sign

 

[2informaticos]

GPU voltages cycle too?
You can have CPU, or GPU issue.
Something changes if remove the RAMs?

 

[protocold]

PM_ALL_GPU_PGOOD is cycling...

remove RAMs make no difference

 

[2informaticos]

Work with board alone on the desk.
Check if all voltages from pages 71/72 try to appear.
Also test if PPBUS_G3H and 3V3_S5/SUS are stable.

 

[protocold]

If you means " Misc Power Supplies" and " Power FETs", then I have checked and they all cycling too, except of course PP1V05_S5 because U7740 does not exist.

PPBUS_G3H and 3V3_S5 are stable. There is no 3V3_SUS signal on this board.

When the board is in the machine, how come it doesnt wait for me to press the button before it try to power on? as long as the magsafe is plugged, the fan will start spinning up and down, with the front led going on and off at the same time...

 

[2informaticos]

All iCore machines start directly, if no battery present.

"except of course PP1V05_S5 because U7740 does not exist."
You didn't read the Apple note from schematic .
Do you get a pulse (at least) on 1V05_S5???

 

[protocold]

I dont have any oscilloscope, but my voltmeter is showing PP1V05_S5 is cycling between +0.005v and -0.004v

 

[2informaticos]

If you are able to detect pulsing all the other voltages, should not be a problem to detect 1V05_S5 too.
More than that, this voltage comes from 3V3_S5, which you said is stable.
Post diode mode to ground (at 1V05_S5).

 

[protocold]

i still dont understand, U7740 should be "NO STUFF" and how it comes from 3V3_S5? I checked other 820-2850 board and U7740 is also not presence..

diode mode to ground is 1 (open), other 820-2850 also confirm the same.

 

[2informaticos]

Did you read Apple note from that page???

 

[protocold]

I can only see:

" 1.05V S5 LDO Ibex Peak-M requires JTAG pull-ups to be powered at 1.05V in S5. Pull-ups (3) must be 51 ohms to support XDP (not required in production). 70mA is required to support pull-ups. Alternative is strong voltage dividers (200/100) to 3.3V S5, which burns 100mW in all S-states. "

Are we looking at the same page? Bottom right "PAGE 77 OF 132/SHEET 71 OF 101" ?

 

[2informaticos]

"Alternative is strong voltage dividers (200/100) to 3.3V S5"
So U7740 is not required...

 

[protocold]

ok... where is this voltage dividers then? where should I be measuring?

If i measure at the solder pad of the missing U7740, I get nothing... diode mode to ground is just 1....

 

[2informaticos]

I never found these resistors; also didn't search them.

You said to have another 2850 board.
Do you get 1V05 there?

 

[protocold]

no there is no 1v05 there on another 2850, diode mode also read 1... perhaps these are only meant for the jtag and not required in production(as it said)?

 

[piernov]

PCH has internal pull-ups for JTAG signal, in theory this circuit is only required to actually use JTAG when debugging. Some board models still have it populated.