iMac (2012 model) A1419, EMCNo: 2639 not starting up

[Braith]

Hi, EMC 2639 .iMac not starting up. Single dim diagnostic LED (LE0301) flashes off briefly when power button pressed. Working through supply rails at present (PP5V_S5 appears shorted to ground but still early days)

Using Boardview 820-4652 but not sure if that is correct one to use. Have searched Forum but only found bios found for 2639 so far.
Have also tried 820-00134-08 and 820-3478 but have noticed discrepancies compared to my board.

If anyone has correct BOARDVIEW and SCHEMATIC for a EMC 2639 I would really appreciate it before I get too far into this.
Thanks.

[Braith]

Update: PP5V_S5 was not shorted to ground, probe was probably touching conductive screen ribbon when I was probing deeply under edge of case.

The Logic board has a "CE3E 31PITMB00H0" sticker if that is of any help.

I have worked thru the boardview netlst, testing all obvious PPxxVx_xx lines (PP12VACDC, PP12V_G3H, PP12V_LCD, PP12V_S0, PP12V_S5, PP1V05_50, PP1V2_S4_camera, PP1V5_S0, PP1V8_S4_camera, PP3V3_G3, PP3V3_S0, PP3V3_S3_VRefmrgn, PP3V3_S5, PP3V3_TBTLC, PP3V42_G3H, PP3V42_G3H_SMC_SPVSR, PP4V5_Audio_Analog, PP5V_S0, PP5VS0_GPU_VCore_VCC, PP5V_S0_HDD, PP5V_S4, PP5V_S5 and PP8V_BLC).

As this is the first time I've worked thru an iMac like this, I'm hoping I'm headed in right direction.
If anyone see's a power rail I've glaringly missed, I'd appreciate a heads up.

Background: This is my wife's computer. It had an episode identical to this 6 months ago, refusing to start after a period of no usage.
Before I could open it to investigate, it started up itself in the middle of the night and ran flawlessly until a week ago.
Once again it has failed to restart after a period of no usage (possible powered down during that period).
Sadly the current backup drive has been temporarily? mislaid so I'm rather interested in getting it running again.
Have tried lightly using freezer spray on logic board in order to tempt it into firing up - to no avail.

Struggling to get my head around the schematic. I find the 800-3478 Boardview and Schematic is mostly correct, when it is glaringly wrong I find the 800-4652 is correct (and vice-versa)
I find that some power rails are obviously marked wrongly (in boardview & schematic) in different parts of the board as I have got vastly different resistances to ground at different points on the board (i.e. PP3V42_G31 when tested at D1900 compared to R4605 or L6961)

I'm struggling to get my head around which power supply / rail to monitor on an oscilloscope when I press the POWER button. Hoping there will be brief signs of activity that will point me in the right direction

Research continuing.. Suggestions welcome.

[mon2]

Hunt for the 820-**** part number etched in the copper layer so will be covered by the black soldermask. That is the identifier we require. Or post some good quality pics of the board which will help to locate the same details for your board.

Be very careful of the open frame high voltage power supply which can be lethal if touched. Best to remove all power, press the ON/OFF button and wait a bit for the caps to discharge.

Then use your meter in resistance mode (not diode mode) and measure the resistance to ground of each and every inductor (2 leads and often grey in body color). Checking for a low resistance which may denote a shorted cap on the power rail. Each inductor is linked to a buck regulator (takes in the high 12 volt or so as input -> creates a lower voltage rail for downstream parts).

Be sure the ON/OFF button is not defective. Start with probing across this button which lands on the power supply board. Then press the button. The resistance should be ~0 ohms when pressed and stable. Is it? If not, the button could be defective.

Next, hunt the power rail you suspect which is also linked to the USB Vbus rail and a common point of failure from USB devices / powering crap devices that are not supposed to be powered by this port.

What is the resistance to ground of PP5V_S5? (in ohms) If you read 'OL' on your meter then the scale selected is too low to measure the resistance. Skip to a higher scale till you have a measurement.

[Braith]

Thanks 'mon2', info much appreciated.

-Found - Logic Board Part No 820-3478-A See pic below. Located on bottom edge of board, hidden by case when installed. Easy to find when board removed & you know it's copper etched under black soldermask.

-High voltage warning acknowledged. (Have been working with same for 50ish years, had a few zaps but still here)

-Have metered out (to ground) every inductor big and smallish. No shorts obvious. Inductors near CPU around 20 ohms (as expected?). (I have also checked vast majority of caps for shorts)

-Power button confirmed operational. Bottom diagnostic LED, which is normally continuously lit, flashes off briefly about 1 sec after power button is pressed. That LED was dim / failing, I have replaced LED, it is now steady and bright. PP3V3_S5 driving the LED via 1k resistor seems OK.

-USB ports haven't been abused to my knowledge. Regardless I'm backtracking through the schematic to locate the voltage source. They appear to originate from PP5V_S4 and pass through individual inductors and mosfet switches. Will meter lines next time I power up board, however I suspect PP5V_S4 is not running. Will check earlier up the power supply chain once I get my head around how it is sourced.

-Resistance to ground of PP5V_S5 is very high. I measure somewhere around 19Mohms once it settles. Quite fascinating watching an auto-ranging ohm-meter trying to cope with charging capacitors on a high impedance line - starts low-ish resistance and jumps around wildly as meter auto ranges. works better if auto-ranging disabled.


I'm currently working through schematic trying to follow the power supply chain of regulators and current / voltage sensing. Not easy.

Before I removed the Logicboard, the only voltages I confirmed as being present were PP12V_ACDC and PP3V3_S5.
PP3V3_S4 and PP3V3_S0 were both missing. The rest didn't get checked.

Will reassemble, power-up, and, hopefully get a feel for the different regulators and what they are trying to do and what's inhibiting them.

It is certainly tempting to try and force-enable some of the earliest power regulators (or inject suitable power), however I suspect that is a last resort prior to binning the device.

If there any articles (etc) that discuss benefits / hazards of over-riding iMac regulator inhibits, I'd be interested in checking them out.

Thanks for your help, will report on findings.

[maniche]

Thanks 'mon2', info much appreciated.

-Found - Logic Board Part No 820-3478-A See pic below. Located on bottom edge of board, hidden by case when installed. Easy to find when board removed & you know it's copper etched under black soldermask.

-High voltage warning acknowledged. (Have been working with same for 50ish years, had a few zaps but still here)

-Have metered out (to ground) every inductor big and smallish. No shorts obvious. Inductors near CPU around 20 ohms (as expected?). (I have also checked vast majority of caps for shorts)

-Power button confirmed operational. Bottom diagnostic LED, which is normally continuously lit, flashes off briefly about 1 sec after power button is pressed. That LED was dim / failing, I have replaced LED, it is now steady and bright. PP3V3_S5 driving the LED via 1k resistor seems OK.

-USB ports haven't been abused to my knowledge. Regardless I'm backtracking through the schematic to locate the voltage source. They appear to originate from PP5V_S4 and pass through individual inductors and mosfet switches. Will meter lines next time I power up board, however I suspect PP5V_S4 is not running. Will check earlier up the power supply chain once I get my head around how it is sourced.

-Resistance to ground of PP5V_S5 is very high. I measure somewhere around 19Mohms once it settles. Quite fascinating watching an auto-ranging ohm-meter trying to cope with charging capacitors on a high impedance line - starts low-ish resistance and jumps around wildly as meter auto ranges. works better if auto-ranging disabled.


I'm currently working through schematic trying to follow the power supply chain of regulators and current / voltage sensing. Not easy.

Before I removed the Logicboard, the only voltages I confirmed as being present were PP12V_ACDC and PP3V3_S5.
PP3V3_S4 and PP3V3_S0 were both missing. The rest didn't get checked.

Will reassemble, power-up, and, hopefully get a feel for the different regulators and what they are trying to do and what's inhibiting them.

It is certainly tempting to try and force-enable some of the earliest power regulators (or inject suitable power), however I suspect that is a last resort prior to binning the device.

If there any articles (etc) that discuss benefits / hazards of over-riding iMac regulator inhibits, I'd be interested in checking them out.

Thanks for your help, will report on findings.

boardview and schematic , hope it's will help you https://www.badcaps.net/forum/docume...7-mlb-820-3478

[mon2]

Good. So the first LED = LE0301 is lit and solid on? This is the 'It is plugged in LED' to confirm the power supply is working. No other led is lit?

Next one to light up should be LE0302 = ALL_SYS_PWRGD. This led will only light up if all power rails are deemed to be OK (based on the power good signal of each rail; ALL_SYS_PWRGD must be true).

[Braith]

Thanks again 'MON2'

In answer to your questions -
-Confirmed - Only LED0301 is lit (It dips for 1/3 sec, slightly after the power button is pressed). All other LEDS are off.
-ALL_SYS_PWRGD is at 0v, Oscilloscope confirms no bursts of signal.
----------------
----------------
These are the Boardview and Schematic files I'm using
" Apple iMac 27 A1419 J17 PVT 820-3478-A – Boardview (.brd).brd "
and " ***********_J17_MLB_051-9884_820-3478.pdf "
----------------------------------------------------------------
-----------------------------------------------------------------
I have been working thru Schematic & Boardview files comparing them to my logic board to see what signals control ALL_SYS_PWRGD.
The Logic board is still mounted in the imac case using original power supply and power switch. I have disconnected peripheries (Screen, Hard drive, Camera, Speakers, WiFi board and Audio connector)

Long story short - I have traced backwards from ALL_SYS_PWRGD over many hours. I have found every stage disabled in one way or the other.
The principal originator of the disabling is PM_SLP_S5_L, the output of POWER SYSTEM MANAGEMENT pin 43 U1100. This ensures that PP5V_S4 is NOT generated!

My next stage will probably be to remove the logic board from the case, power it from a 12v bench supply and jury rig a 'Start Button'
Hopefully this doesn't produce side effects that will lead me on a wild goose chase.
I will then attempt to analyze why power management (U1100 PM_SLP_S5_L) is disabling enable pin24 of U7600 thus disabling PP5V_S4 (even after PwrBtn is pressed)

Somewhat Tempting to pull PM_SLP_S5_L to 3.3V to force generation of PP5V_S4. (Page 76 under heading "S4 Enables").....THOUGHTS????????
--------------------------------------------------------------
-------------------------------------------------------------
NO NEED TO READ FURTHER IN THIS POST UNLESS YOU WANT TO SEE HOW I TRACED BACKWARDS FROM "ALL_SYS_PWRGD" back to the "POWER MANAGEMENT" enable pin.

I have included the info below in case it is of use to any one trying to get their head around the power management chain.
Beware ...Getting a bit weary, there are possibly errors in my logic (sic)
I'll have another go in a few days.
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------


------------------------------------------------------------
Working thru 820-3478 schematic on page77 to see what signals control ALL_SYS_PWRGD
PM_PGOOD_REG_P1V05_S0 = steady 0V (pin1 of U8601 AND gate i/p) (probably due to PP3V3_S0 being down?)
PM_SLP_S3_L = steady 0V (pin2 of U8601 AND gate i/p)
M_PGOOD_SLP_S3_P1V05_S0 is the o/p of above AND gate = assumed to be 0v with dip to 0V for 1/3 sec following pressing power button (due to aforementioned dip in PP3V3_S5)
PM_PGOOD_REG_CPUVCC_S0 = 0V (probably due to PP3V3_S0 being down?)
------------------------------------------
Investigating why PP3V3_S0 is down - following "S0 ENABLES" schematic on right top of page 76
PP3V3_S0 is turned "ON" by PM_EN_FET_P3V3_S0 which is sourced from PM_PGOOD_FET_P5V_S0 (via R8531 0ohm).
(PM_EN_FET_P3V3_S0 = 0V)
(PM_PGOOD_FET_P5V_S0 = 0V)
------------------------------------------
Investigating why PM_PGOOD_FET_P5V_S0 is down
PM_EN_FET_P5V_S0 turns "ON" U8420 (pin2) which produces PM_PGOOD_FET_P5V_S0 if happy.
(PM_EN_FET_P5V_S0 = 0V)
(PM_PGOOD_FET_P5V_S0 =0V)
-----------------------------------------
Investigating why PM_EN_FET_P5V_S0 is down
U8500 pin 11 (AND gate being used as buffer) drives PM_EN_FET_P5V_S0 via R8530 (100ohm)
The input of buffer (pin 12 U8500) is shown as PM_PGOOD_FET_P12V_S0 = 0V (page 76 top right - poss schematic error?)
(PM_PGOOD_FET_P12V_S0 = 0V )
-----------------------------------------------
Investigating why PM_PGOOD_FET_P12V_S0 is down
U8450 pin 8 drives PM_PGOOD_FET_P12V_S0 via R8454 (0 ohm)
U8450 is turned "ON" by pin 2, PM_EN_FET_P12V_S0
PM_EN_FET_P12V_S0 = 0V
note that U8480 pin2 (which produces PP12V_S0) is also turned "ON" by PM_EN_FET_P12V_S0 which probably explains the name of the signal)

-----------------------------------------------------------------------
Investigating why PM_EN_FET_P12V_S0 is down
PM_EN_FET_P12V_S0 comes from pin 8 of U8500 (AND gate output)
U8500 pin 10 input PM_SLP_S3_L is driven by U1100 (SYSTEM POWER MANAGEMENT) pin AK40
Pin 9 input PM_PGOOD_REG_VDDQ_S3 is driven by REG_VDDQS3_PGOOD which comes from U7300 output pin 20 (PGOOD) . This is an 'interesting' component (Memory Power Solution Synchronous Buck Controller - have datasheet))
PM_SLP_S3_L = 0V and no activity is seen if mains power is connected or when pwrbtn is pressed
PM_PGOOD_REG_VDDQ_S3 = 0V no activity when mains power is connected or when pwrbtn is pressed
---------------------------------------------
U7300 is labeled as "VDDQ (1.5V / 1.35V) S3 Regulator", appears to be a major component and appears to only care about producing PPVDDQ_S3.
PPVDDQ_S3 =0V no activity when mains power is connected or when pwrbtn is pressed.
Further investigation of U7300 needed.
Datasheet shows it needs 5V on pin 12 - this would be PP5V_S4_REG_VDDQ_S3, which is at 0V, SO NO WONDER U7300 is not functioning.
--------------------------------------------------------------------------------
Investigating why PP5V_S4_REG_VDDQ_S3 is down (page 66)
note - some sort of error here... seems like USB power has it's own regulator running off PP5V_S4 and I've got mixed up in the next few lines.
PP5V_S4i is provided by U4700 (TPS2561DR 2-ch, 0.25-2.8A adjustable ILIMIT, 2.5-6.5V, 44mΩ USB power switch, active-high) datasheet available
Pin2&3 are power in for U4700 (TPS2561DR). This is also labeled PP5V_S4_USB mmmmmm
on page 87 it shows PP5V_S4 is source split to PP5V_S4_USB and PP5V_S4_REG_VDDQ_S3
OK, struggling to find actual source of PP5V_S4.
On page 67 it appears that Q7650 & Q7655 (driven by U7600) together with L7650 produce PP5V_S4
U7600 is a "High-Efficiency, Quad or Triple-Output System Power Supply Controller" ISL62383CRTZ - datasheet available
These require PP12V_S5 for the switching Mosfets & U7600
PP12V_S5 = 12.1V that dips to 8.5ish volts for 1/3 sec after pwrbtn pressed
So...Why no PP5V_S4 ... if U7600 and switching mosfets Q7650/Q7655 have power ??
--------------------------------------------------------
So, lets see why U7600/Q7650/Q7655/L7650 are not creating PP5V_S4_REG.
Checkin
It appears that U1100 (System Power Management) pin 43 PM_SLP_S5_L (via U8500 pin1 & 3) is not enabling pin 24 of U7600.
(pin 2 of U8000 is at 3.3 V apart from 1/3 sec dip to 0v after PwrBtn press)
PM_SLP_S5_L =0V
Hence no PP5V_S4.
--------------------------------------------------------------
So, now I need to find out why 'POWER MANAGEMENT' won't enable the PP5V_S4 regulator.
More to come.

[mon2]

Based on a quick run through of the above, would you have a flash chip programmer? Perhaps worthwhile to consider a known good bios.

Would you have a thermal camera? Have seen cases where one of the USB side load caps were shorted and took out the logic board till the shorted cap was removed. Do inspect each of the USB load switches for shorts / damage.

[reformatt]

Most common issues I see on old iMac boards are shorted caps. These are usually at the bottom end of the board, either in line or below the bottom of the screen (as the air is drawn up from the bottom by the fan, and dust/moisture accumulates mainly in this area). The headphone jack at the very bottom is also a common fault as PP3V3_S0 routes here and there's a decoupling cap on it that goes short.

Initially I eyeball each cap under the microscope in the bottom of the board (e.g. C3387, C3287, C7661 etc) and check if any sub-rails are shorted. You can run the board on the bench from a 12V bench supply. J5120 is SMC ON/OFF and you can short that with tweezers. You can also use the Ivy/Broadwell power sequence from LogiWiki for troubleshooting if needed.