I think I killed it. Short on LCD connector, no longer boots

Hey all,

So, having managed to get the emmc working on the switch I was talking about in the “Is this CPU failure thread”, I then set about fixing the less pressing issues on the device. All of which went well. However, as is often the case, I bent a pin in the LCD connector when putting it together.
So, I took a connector off my spares board and swapped it over. What I didn’t notice before attempting to boot it from battery however was a tiny short between two of the pins on the left hand side of the connector.
On attempting to boot it from battery only, I got nothing, and I could feel heat from the LCD connector area. It was only on closer inspection and testing that I found the short. I have now reflowed the connector and no short is present.

With the newly working emmc installed it WAS booting this morning (pre-short). Now I see no response. With the emmc removed, I am able to get it to connect to the PC with the USB in one direction, but not the other. The bad way round the PC knows it has been detected, but shows an error. With it the good way round, I am able to connect to RCM and inject a payload, and it returns 0x7000 as it used to, but I no longer see anything on the screen and the backlight does not come on.

I have checked for shorts in obvious places, including under the CPU, but haven’t found any. All the rails around the PMIC look to match what I would expect. I will continue to do some reading etc, but does anyone know if there is any hope on this one, or where to start?

I would start by inspecting the M92 IC, if it’s only getting charge/detection in one orientation then that could be at fault. failing that then check the diodes near USB.

As for display, sounds like you had a run of bad luck here, I would probably recommend switching LCD connector once again, then confirm if you get backlight or not. Do you get fast charge?

Can only guess that the short on your LCD conn was on your 1V8PDR and it killed the M92 IC… but just complete guess.

Yeah, my luck wasn’t so good… and it was going so well before that! I really want expecting to get the emmc reading.

It actually charges both ways round (at ~0.42a) but as its not booting, it doesn’t get up to a fast change, it is the detecting on the PC that seems to be only one way, and I dont remember if this was the case before the short. I will start by testing the pins on the connector and the links to the PI3 as you suggested. When you mention diodes, do you mean the filters below the Pi3?

I didn’t find any of the usual shorts around the M92, and all the pins in that area of the connector run directly to the CPU which has me worried :worried:

Post from Calvin Here detailing the location of the diode on the rear

There is another one on the otherside of the board near the USB, same package IC. readings there will be approx 0.8

Depending on your board rev you may not have these diodes present, but take those measurments all the same.

M92 does not always fail short to ground. It could also be down to bad joints too, after verifying you have continuity from the USB to the various respective destinations I’d start by swapping the M92 IC out and seeing if you then get PC USB detection with the USB in both orientations and if so, that rules out that problem, though I don’t know that this is your primary issue.

Blackscreen and no boot logos might point to more serious issues, did you transfer the data from that EMMC to another IC/module or are you using the same reballed one? did you backup all partitions originally including boot0/1 etc ?

Ahh, OK, I will take a look for those reading at lunch.
I am still using the reballed one, as the other issue the console had was a damaged SD connector, and it was putting it back together after that in order to start the backup that it all started going wrong!

The console should boot / connect to PC without an M92 fitted at all right?

It will typically boot (with an error) but I don’t think you’ll get PC detection with it removed as it’s effectively managing USB power. In fact I wouldn’t connect a USB to it at all with the M92 removed.

Does your rev board have the filters around the LCD connector? if so might be worth checking them and making sure none are open, as it is, you should be getting backlight/display after sending Hekate payload, but I would resolve the potential USB or M92 issues first then we’ll move on to testing the various things which are needed for image and backlight.

Yes, this one has the filters coming from the LCD, they all tested OK yesterday, but I will double check them.

Ok, I have replaced the USB port, and now it charges and connects to the PC both ways round.
I have re-checked the filters under the LCD connector and they seem good. Checked the diodes, and get a reading of 0.5v (rear) on one side of the board, and 0.8v on the other (front).

Which leaves the lack of booting, and the lack of display / backlight.

Oh, I also measured the pins on the LCD connector in the area around where the short happened (so the first 6 or so on each side) and compared to another board I have, and nothing stood out. I also just checked the emmc on the PC again, and it still shows partitions so hasn’t given up on me yet.

So next up I’d check the 8316 IC is being told to turn on the outputs and if yes if the outputs are at the correct voltage, that being +5V and -5V

So, check that your getting a high (1.8V) at the highighted enable points then check the outputs at the highlighted caps.

Careful with your probes here, you really don’t wan’t to cross your probes here.

Not that I think it will make a difference, but just incase, take these measurments following sending the Hekate payload.

If that’s all good, then next up would be backlight driver IC.

Ok, I think this looks promising: I have 4v on the input, +5 and -5 on the outputs, and 1.8v on both inputs. However I have almost nothing (mv) on the inductor.

That’s all good :+1:

Next up is backlight driver, with the board connected to LCD and BL ribbon, send the Hekate payload and measure backlight voltage directly at the connector though careful here again (it’s easier than directly at the chip with LCD connected) if your getting BL voltage (which from memory is greater than 15V) then we go back to suspecting the LCD connector or the LCD itself, maybe the bent pin killed the panel, though, it might be a good idea to try another LCD anyway as I don’t think backlight output is enabled without the LCD connected, so if there is a fault on the current panel it could prevent it boosting.

I have tried it with a couple of screens, (and broke a bl cable in the process…. Sigh ) so I dont think its that.
However I only get 5v on the bl connector…

Oh no :frowning: you can fix that though, just scrape the polymer coating back and tin and jump the traces with wire.

Hmm, tommorow I’ll go over the datasheet and create a simple pinout photo of a few things you can check.

Also, I don’t know what revision LCD’s your trying but I’ve found the ones with the black shield coating on the ribbon that they seem to “stretch” the LCD connector out which will sometimes prevent the ones with the non shield coated ribbons from making proper contact in some cases, maybe that’s what’s going on here?

Yeah, I have repaired one of those before so I know it can be done, but will worry about that another time!
It was originally using one of the screens with black on it, and the other 2 didnt, so its possible? But it didn’t feel loose…

Ahha! I was just taking a closer look in the BL IC area, and I think the chip may have swum around a little while replacing the SD connector slot, as it look fine top down, but side on its wonky! I will get this swapped out or reballed at lunch and report back.

Ok, new BL IC arrived today, so I fit it and the screen is now back up and running! The LCD that was connected when it was shorted also still works, so far so good.

It is not booting in to HOS any more though, which is annoying, can get into RCM now though, so will get a back up of the emmc next. But now just need to work out why it has decided not to boot again…

I would first check the “Console info” section of Hekate and verify the everything looks in order, check the fuel guage etc.

If all good, I’d do the whole rebuild FW nonsense using EMMCHacGen as it’s possible the EMMC was slowly corrupting due to the prior damage.

If still nothing, I’d measure voltage at CPU buck and seeing if your getting your CPU-01 voltage output at the inductor, which from memory is anywhere from 0.8V to 1.2V, if this is being generated then the issue is likely after this.

It is same symptoms as before right? power button and black screen, no boot logos or anything?

Previously, once I had repaired the emmc, it was able to get into HOS. But yes, this is a blackscreen, no BL, no logos.

I am also having issues getting into RCM with the emmc installed. I have soldered a wire between the two end pins of a joycons black rail part, and put it in the right joycon slot, but its not connecting…
I worked around that by inserting the emmc after it was in RCM, but then it wasnt detecting my sdcard…

No bootlogos point to an early boot error, so it’s possible boot0/1 partitions are corrupt, though it’s very hard to say as several other things can cause the same symptoms.

For sure though, keep your original backup safe on your PC before attempting FW regen

It’s also worth doing the benchmark test for the EMMC in Hekate and comparing your reults to a known good, as Hekate will read/write an EMMC in single bit mode, HOS itself (as far as I know) will not boot without all data lines.

I guess when your trying to get into RCM, your holding vol +/- and then plugging in the cable in order to prompt the console to boot?

If yes, then it goes back to M922/USB related issues again. With the console still plugged in, hold the power button for 10/15 secs (to ensure it’s for sure powered off) , while still plugged in, reconnect the EMMC module, hold the vol +/- and then push the power button and see if it starts in RCM state