Dell XPS 9380 No Power - After providing 3VALW and 5VALW once, they persist - still no power-up

[seppde]

Hello everyone,
I've recently gotten my hands on a defective XPS13 9380. I will attach a zip of the schematics and boardview.
My first discovery on the Board was that the reverse current Diode PD103 of the 1st USB-C was blown.
Nothing to worry about, the board got 3 USB-C ports which support Power/Charging. On I go.
Next I inspected the power-rails to see why the diode blew up and found a short on the 3V3 rail which I traced back to the Power-IC for the LCD (UZ2). I desoldered it and the very low-ohmic short is gone. I checked on the LCD-connector that there's no short now on LCDVDD to GND.

I now decided that I can try each one of the charging ports:
On the first USB-C port my USB-meter shows 19V and 0,05A. The ESD-protection diode is getting warm. As I haven't replaced PD103 yet, obviously it won't power the board.
The second USB-C port shows only 5V...
The third one though shows 19V at 0,03-0,09A and I get 3V at the LDO, as well as B+ with the correct 8,4V. At this point I do not get 3VALW and 5VALW though.
I decided to manually inject 3V at the corresponding rail, which causes the Buck-converter to take up it's work and provides 3V3 as long as I keep the USB-C connected. The board doesn't power on either, though...

I need help figuring out what could be the cause for the 3V/5VALW not getting generated. My understanding of the different controllers involved is lacking... I've so far considered the Embedded Controller (I believe that's MEC5105 (UE3)) or the PCH to be defective.
If I inject power to the 3V rail, it draws around 100mA but nothing is getting hot at 0,3W...
The VCCCore and other CPU/GPU voltages should at this point not get created, if my understanding is correct. I measured the resistance of +VCC_CORE (4Ohm) and +VCCGT (6Ohm) to GND, anyway. They seem to be very low, but maybe not yet indicating a dead CPU? What do you think?
The power button does indeed pull the corresponding pin of the EC low, I checked that.

Maybe you can help me figure out the problem of this Laptop. Thanks.
I will provide additional information as quick as possible.
Kindly, Sepp

Link to schematic/boardview https://www.badcaps.net/forum/troubl...dview-required

[reformatt]

Is this something you've bought faulty, and someone has potentially already messed with? Look for any evidence of board rework, and double check anything that has already gone on.

You have a power sequence in the schematic on Pg5 which is a good starting point. PU500/PU501 are your +3V/+5V power supplies. Check that ENLDO_3V5V is high, and that +3VLP and VL (5V) are present. +3VLP powers the EC, which if everything is ok in that asserts the ALWON signal to get your main power rails enabled.

If +3VLP is present, and ALWON is absent then you need to troubleshoot the EC basic requirements etc. If +3VLP is absent, you need to see if there is a short etc, which is usually the EC itself in that case.

[seppde]

Thanks for your ideas.
+3VLP as well as VL (VL is not used as far as I can tell) are present
Same for ENLDO_3V5V.

The ALWON-Signal is LOW, which is the reason for the main power rails not getting enabled.
(Same as with 3VALW, if I pull up the ALWON-Enable-Pin on the 3V3 converter, it turns on and stays)

I'll look into the EC-requirements.

I forgot to mention that the battery is not charging either. That could be a problem of deep-discharge and the Controller-IC locking out the battery. (The fuse on the battery is OK)
Could it also be a problem related to the EC?
I've since carefully recharged it up to 7.8V which didn't change anything. I can't identify the Battery Protection / BMS-IC.
Also, the RTC-battery is at 0,7V.

I've bought this faulty, yes. No signs of previous repair work. Just seems to have been lying around unnecessarily long...

[seppde]

I've made some progress and hope my procedure isn't too questionable:
When I supplied 3v3 at the RTC Connector, the 3V3/5V converter would come up immediately upon plugging in the USB-C cable.
This, too, drove most of my attention to the area around the MEC5105.
As I was measuring 682Ohm on the 3V3-Rail, which I believe is much too low, I was taking a guess on the MEC being defective and therefore decided to cut the trace at RE252, separating +3VALW and +3.3V_EC5105.

Now:
- I get 3V3/5VALW upon plugging in the USB-C without additional steps
- The +3.3V_EC5105 is at 1,65V
- the battery indicator lights give off a very faint light, if the battery is not connected (3,3V at the battery connector), without pressing the button. No reaction on button presses
- when I plug in the battery, without the USB-C connected, I get a full battery indicator with strong lights upon pressing the corresponding button (strange because the battery isn't fully charged)
--> as I was writing this, this behavior changed, I can't recreate it. Must have been some remaining charges somewhere
- if battery and USB-C are connected, there's again the faint light at the battery indicator and no reaction upon button presses
- I measure 888Ohm from +3V3ALW to GND
- I measure 700Ohm from +RTCVCC_EC to GND - voltage is 1,65V, but I've measured different already, too --> not stable

Next on the startup-sequence would be PM_BATLOW# going high. I've read here (https://www.badcaps.net/forum/troubl...er-how-to-work) that RTCRST# and RSMRST# also need to be high for a startup.
- RTCRST# is 1,65V --> if I pull up RTCVCC_EC to 3,3V (drawing 0,03A), RTCRST# also goes HIGH
- RSMRST# is 0,00V (not sure what this signal is?)
- PM_BATLOW# (I suppose that's PCH_BATLOW#) is 3,3V

- VCCDSW_EN is 2,0V
- 3V_PCH is 0,0V
- 3V_PCH_DSW is 3,3V
- PCH_PWR_EN is 0,0V

--> I guess my next problem is that the PCH is not receiving Power? And if so, why...
By me cutting the trace RE252, i should've cut power to the RTCVCC_EC. Why do I still measure 1,65V there?

[seppde]

I left the laptop at peace for a few hours and now just wanted to check back whether the battery indicator would again be "working" if I only plug in the battery - it indeed did.
With just the battery and no USB-C I measured
- 1,21V at +3.3V_EC5105 and RTCVCC_EC
- 7,0V at B+
- No voltage at 3V3/5VALW - UNLESS I press either the power-button OR the battery indicator button. In both cases ALW gets asserted and 3V3/5VALW get generated. And in both cases the battery indicator lights light up, nice and strong.

So somehow the EC still reacts to button presses, although it's not even properly powered...

[reformatt]

Battery charging is irrelevant till you get a working laptop so you leave that till you have a working machine. Laptops are designed for the most part to run without one.

A flat CMOS battery shouldn't be an issue, other than losing BIOS settings, time/date when all power is removed. During normal function, the system should provide that RTC power so as to extend the life of the battery. So on adaptor, (regardless of battery voltage), you should have ~3V at +RTCVCC_EC because this is fed from your LDO +3VLP via RTCD1/RTCR2. RTCVCC is provided by +RTCVCC_EC via a reset circuit. Are these two lower than 3V?

On again reviewing the power sequence timing diagram, EC will only issue ALWON after power button is pressed. EC primarily relies on the LDO voltage to execute it's ROM, other voltages like +3.3V_EC5105 are produced much later.

For you to go further in the sequence, the EC must get power good signals from the +5VALW/+3VALW/+3VALW_DSW converters. If +3VALW is low, the EC won't proceed further in the sequence. Remove PJP500 (will be a solder link I think) to isolate the converter and whether this is a load or supply issue. 800 ohms or so on that rail is low, but not low enough it would pull that supply down. You can however have a short post load switch so resistance checks alone are not conclusive. Thermal camera tends to help in this regard. Check pg 4 of the schematic for the load switch's for the +3VALW rail.

Smells of a bad PCH to be honest.

[seppde]

I've checked the power rail of the PCH, 3V_PCH, somehow I forgot checking it for shorts until now - I'm new to this, still.
There was a dead short, upon injecting voltage it drew like ~2,5A at 1V. I could measure the temperature of the Chipset (it's the smaller chip on the CPU-BGA, right?) to be about 35°C at this point, before I cranked up the current a bit too much and now it's "back to normal" with 200-300mA at 3V...
--> I doubt that things that were broken magically heal upon putting way more power in there then is good for them. The unit is dead? (Please confirm)

[reformatt]

Initial triage of any board is quickly checking each coil for shorts. Saves a lot of time. Dead PCH/CPU unfortunately is pretty common, especially on the modern processors. Well you've learnt something, and troubleshooting in a logical manner is something you hone through experience.

Yes the board is a donor at this point. The CPU's require skill and equipment to remove, and then you need to source a (working) replacement. You could source a replacement board by finding another machine with dead screen etc on eBay, just keep your eye out.