820-00928 slow to post and booting to Critical Software Update screen

bjf

Member
Light spill. I attached pictures of the only area that was hit. I cleaned this up. (I had to replace a resistor under the metal shielding on the actual Touch Bar cable. That works fine now with my test board.) Board takes about 15 seconds to post and is booting to the Critical Update screen. I have everything plugged in including the battery. Entire top case is tested and known good using good test board. I've also installed this board in an entirely different known-good case and it does the same thing. Any ideas what might be causing this?
 

Attachments

  • photo2491.jpg
    photo2491.jpg
    1.5 MB · Views: 4
  • photo2492.jpg
    photo2492.jpg
    1.5 MB · Views: 4

2informaticos

Administrator
Staff member
You can try clean ME BIOS for slow booting.
But I recommend you to start with critical update error.

For the critical update check all these recomendations, which I collected from Internet.
Unfortunately didn't get a machine to test yet...

"https://support.apple.com/en-us/HT207567

Have had this the past few days and have some information which may be of use here
- Programs like 'Hands Off' and 'Little Snitch' (essentially ingoing/outgoing network traffic blockers/gatekeepers) were the cause of this not installing for me
- if this happens, then boot into safe mode (Hold down shift when turning on) and disable the program, then restart and install.
As a side, conversations with apple advised that these updates are to do with the OS that runs the touchbar (among other things) therefore, if you can get into safe mode, this should confirm its not a hardware issue.

I solved this by using my mobile hotspot - for some reason my office/home wifi kept going in a loop. Might be worth a try?

Some generic ?fixes? like nvram/pram reset won?t solve or help for this issue.
Full clean install, installation on top of existing only fixes this half way.
If you want to avoid a reinstall - try this workaround:
ctrl+alt+shift+power to turn off the device Connect your macbook to the network WIRED with usbC->ethernet or usbC->thunderbolt,thunderbolt->erhernet, just like me
Turn the device on. The boot takes longer but it boots!
I did this 10 times to reproduce every time with the following results:
disconnected wired network, restart == critical update loop
power down with key combination, connect wired network, power on == booting
I have no persistent remedy for this issue. Everyone please try this, and edit this to be more clear if there?s a workable and repeatable solution here.
p.s. This workaround seems not to work on encrypted machines."

Let us know IF and WHICH helped you...
 

bjf

Member
I always boot into a USB test drive when testing so I'm thinking that would avoid any issues with traffic blockers. Either way, I did boot into the system with safe mode and wiped the internal SSD so there's nothing left there.

Clean install changed nothing.

Ethernet connection didn't change anything. Same message and failure.

I'm guessing there's something happening with the Touch Bar circuit although everything is measuring fine that I've tested. I'm out of time and ideas. I'm going to let this one go. Thanks for the suggestions and tips.
 

2informaticos

Administrator
Staff member
Error message is clear, critical update.
For some reason cannot download necessary update files.
Did you try all recomendations?
Check if router firewall doesn't block the access, as in Apple paper is explained.
 

bjf

Member
I had to let this one go, but I did have internet fine through wifi and ethernet when I was booted into the system.
 

2informaticos

Administrator
Staff member
Did you check what Apple said in that paper?
Can you get acces to mentioned hosts?
Did you check if your network has firewall enabled?
 

bjf

Member
Yes, I did all of it. Our network is very simple, just a basic setup but I did verify that I was able to access everything on that Apple page by wifi and then also by ethernet when I tried connecting with ethernet only. All of those connections were tested on those ports. I still believe this is just something to do with the Touch Bar circuitry on the board not working correctly.
 

bjf

Member
I just realized we didn't get rid of this yet so it's still here. I found out that DFR_TOUCH_INT_L is measuring .18V and it should be 1.8V. Everything else is measuring fine on J4402 including PP1V8_S0SW_DFR at 1.8V. Not sure what's causing DFR_TOUCH_INT_L to be at .18V, which happens to be the exact same voltage as DFR_TOUCH_ROM_WC. Not sure if that's meaningful but I did notice it.
 

2informaticos

Administrator
Staff member
Did you find 1.8V on working board?
Ending "L" means active low.
In such case, an interruption is enabled.
 

bjf

Member
Yes, I have a good board I'm comparing it to. That board has 1.8V on DFR_TOUCH_INT_L. It's sitting on my desk with nothing plugged into it. The bad board has .18V. The diode measurement on DFR_TOUCH_INT_L is fine at .484 on the bad board.

It's hard for me to understand where the 1.8V comes from on DFR_TOUCH_INT_L being that R4485 is no stuff and it's supposed to be an OUT line going IN to U3900. J4402 was one of the few things hit by the spill and I had to dismantle the Touch Bar circuitry on the actual Touch Bar cable and replace a resistor for that Touch Bar to work on a good board. The diode measurements on J4402 are all very close to the good board, but not exact. There may be a difference of up to .02. So maybe .450 vs .430. All very close.

The one difference I'm seeing besides DFR_TOUCH_INT_L is that pins 1, 5, 9 and 11 on J4401 on the good board will go from .570 to .740 every time I test them in diode mode. On that bad board those same pins always stay at .540. Other than that I can't find any differences.
 

2informaticos

Administrator
Staff member
I suppose DFR_TOUCH_INT_L is pulled high inside touchbar electronics, or into U3900.

"1, 5, 9 and 11 on J4401 on the good board will go from .570 to .740"
Maybe 3 instead 1?
Pin1 is ground.

Check the 3 reset signals on J4401/02; compare with good board.
 

bjf

Member
Yes, my mistake, pin 3.

It looks like you might be getting to the issue. DFR_DISP_RESET_L is missing from the bad board.


Bad Board
1 DFR_DISP_RESET_L - 0V
2 DFR_TOUCH_RESET_L - always 1.8V
3 DFR_CLKIN_RESET_L - .110V up to .250V
4 DFR_TOUCH_INT_L - .110V up to .250V It is always at the exact same voltage as DFR_TOUCH_ROM_WC

Good Board
1 DFR_DISP_RESET_L - 1.8V
2 DFR_TOUCH_RESET_L - Sometimes 1.8V and sometimes .03V (I think I might be triggering this on and off by testing voltages)
3 DFR_CLKIN_RESET_L - 0V
4 DFR_TOUCH_INT_L - 1.8V
 

bjf

Member
DFR_DISP_SMC_RST_L is present at 3.4V on the bad board, same as good board.
DFR_DISP_RST_L is missing on bad board. Should be 1.8V
PP1V8_S0 is present on pin 6 at 1.8V on both boards

Diode measurements are the same on all of the lines connecting to U4406.
 
Last edited:

2informaticos

Administrator
Staff member
DFR_TOUCH_LID and SMC_LID same level on both boards?
Did you reflow U3900, with good flux?
Compare DFR_TOUCH_PANEL_DETECT line with o-scope between both boards.
 

bjf

Member
DFR_TOUCH_LID is at 0V and SMC_LID is at 3.4V on both boards.

Did not reflow U3900 yet. Was trying to rule everything else out first.

I don't have an oscilloscope, but DFR_TOUCH_PANEL_DETECT is at .200V on the bad board and 0V on the good board.
 

bjf

Member
No, but I'm almost certain it was from the Touch Bar circuitry on the board being damaged in some way. There are three scenarios where I see this Critical Update happen. One, is an SSD issue. Two is a battery issue, and three is a Touch Bar issue. This board was damaged at the Touch Bar connector. I was able to remove the shield and repair the blown resistor on the actual Touch Bar cable (tested with good board) but I was never able to figure out why the board was having an issue.
 
Top