I just got a switch which had a blown fuse close to the USB as well as no 4V line around bq24119. I replaced both and now the switch seems to turn on but I get no image on the screen (I am sure that the LCD is working because I have tested it with another motherboard). I also had to reflow P13USB because it did not make contact on a pin. In any case, after all the replacements, I can feel the CPU getting hot after I turn the switch on. I have checked and compared all the voltages around BQ24119, M92T36 and P13USB while using a charged battery without connecting the USB port (I got no port at the moment) against a working motherboard and everything checks out apart from the GPIO1 on M92T36, marked as “Alert#” on the datasheet. I get a low GPIO1 whereas I am supposed to get 1.8V when fully working. I am wondering, could be that the switch starts in recovery mode and needs autorcm to turn on? In this case, why would the cpu warm up like this if it is in recovery mode? I do not know if the switch was modded before, I am trying to figuring it out without connecting the USB port. Is there any way to check if autorcm is active without using the USB? What else could impact the Alert# pin? The GPIO1 pin is marked as IO, so, could it be a faulty M92?
Here are some updates. I have soldered back the usb and apparently autorcm is active, if I inject the payload I get the following message:
Failed to mount SD!
Missing LP0 (sleep mode) lib!
Missing or old Minerva lib!
Update bootloader folder!
I have copied the “bootloader” folder of hekate 5.6.5 to my SD card but if I turn on the switch with the SD inserted I get no image on the screen.
In the hetake options I noticed that the emmc was missing and, by physically checking the connector I could see that the port is not soldered at all. So I soldered it back and now I am getting a blue screen of death every time I turn on the switch. If I remove the emmc I can still access hetake though. Does it mean that the cpu is dead? Or that the nand is corrupted? but if the cpu was dead or badly connected I would not be able to access hetake when the emmc is not connected right?
I am no bluescreen expert, but they can be caused by a number of things. Given that without the emmc you are able to get into hekate I would look closer at the emmc side of things first.
Check that the connector is solid and has no bridged pins to start. If that all looks fine look over the emmc for corrosion and start checking the diode readings against a good board if you have one. There is a picture around here somewhere of one I drew up you can use if not.
I swapped eMMC with one from another working board and read it while running hekate but I get the same error as when I try to read its own eMMC (exception LR 400144E7). Before showing this exception it shows this for an instant.
Here are few updates, I was able to run biskeydump and copy the keys, but how can I generate the nand backup and check whether the nand is corrupted or invalid? Unfortunately hekate does not show anything, unless I remove the micro SD. I noticed that the caps close to max77620h on the bottom of the Apu (which I guess it is related to the GPU) were OL, so I reflowed MAX77620H and now I get normal readings in diode mode and the 4V power line on those caps. Unfortunately this did not get me anywhere when using the latest hekate. so, I checked the APU and noticed that someone attempted to reflow it and that a cap on top of the APU is missing. Here is the photo of the missing cap. Do you know its value by any chance?
To sum up I think hekate does not show anything because the GPU is faulty or is missing power. I am trying to find a new cap and if that does not work, I will try reballing the APU just in cause the guy who reflowed it caused some bridges.