Nintendo Switch sometimes unable to go past boot0/1

Some time ago i attempted to install a trinket chip into a V2 Mariko. I knocked off RST resistor (check photo highlighted)


I was adviced to swap it with a 5k resistor so i did(because i did not have a 4.7k at that time). It worked for a while (couple of hours) but then it freezed in Atmosphere while idle(screen set to never lock) in main menu/games(backlight on, fan on, image frozen on screen. when putting it into the dock and taking out, the image would go black,but backlight still on).After restarting the console. i was no longer able to boot to HOS,but sometimes it would boot to HOS but i would get error 2002-3539.
So i ordered 4.7k ohm resistor (that is what everybody told me that resistor is) and i swapped it back.
Same story,it worked for some time then i would get same behaviour. The job was done with a hot air station.
I tested the line for continuity from EMMC to test points and it looks ok. I also get 4.7k ohms from 1.8v line to RST test pad,without the emmc on board so it tells me that it is properly in place.
I suspected the emmc daughterboard connector to be the problem since it was worn out. I replaced it and i still have same issue. I did a reball on Emmc since i suspected that solder balls cracked while taking it out and i still have same issue. I did a RAM test and i got no errors.
I have decent solder skills now,but i have no knowledge. I dont know what to look for anymore.
In the image above. there is solder paste near the resistor. I did this to avoid soldering to the wrong pads. There is nothing under the solder paste.

Btw i took the chip out, and i still have same problems. So this problem is with nintendo switch

The differece in resistor value you mentioned will make no difference, and will also not be the cause of your current issues.

In regards to your error, see @Calvin post here

Pictures of the board in it’s entirety might be a good idea, front and back.

Checking behaviour in L4T might also be a good call too. As it stands, could be EMMC, Ram, or SoC related issues, bit hard to tell without further diagnosis.

I bent a pin yesterday on a console . I was ruthless with these lcd connectors in the past (until yesterday when i actually messed up one).
Wich got me thinking let me check the connectors. Sure enough when i reached lcd , it had a bent pin. I looked at old photos and i can see that i have this pin bent for quite some time. I do not know if this is the root of my problems. I will replace the connector and i will check again. Is this nightmare going to finnaly end? ;-; No way its this stupid pin that gave me problems right? it has to be more complicated. Can someone tell what the pin is responsible for? (keep in mind the lcd was working even with this bent pin)
Bent pin:


Photo form the past:

Could be, quite common that 1V8PDR will short to ground fully or partially in LCD con bent pin cases and as a result will cause heat, excessive current and which could potentially cause the issues your seeing.

Not sure if the pin you’ve highlighted is 1V8PDR, could be some other rail but :crossed_fingers: this resolves your issue :slight_smile:

Also I do notice something on your GC PCB, is that corrosion? … might explain that error if it is :thinking:

oh no no corrosion. that is thermal paste

1 Like

You also need to use some solder weld on that board it’ll be arcing something cronic without it

i dont understand what you mean?

ok so i replaced the lcd connector. That was not the problem. I had incidents where if i hold my finger on emmc ,it would be “fixed”, but then i had incidents where even if i held my finger on emmc the problems would still be there.
Green and white stuff on the board is solder mask
Also it feels like everytime i apply heat to the board, the problems goes away for a short amount (maybe like a couple of hours) but then next day its guaranteed its going to act up again.

























have you got another EMMC module you can transfer the data too (even if just temporarily) just to take this out the equation?

Possibly due to an open somewhere and the heat is temporarily resolving it as a result of expansion or contraction. Maybe bad joint under an IC. When it’s acting up more consistently, booting into L4T ubuntu might be helpful (for example to check WIFI etc)

When it doesn’t start up, do you see any boot logo/s?

Can you provide pics of the board front and back in it’s entirety so i can see the board as a whole, few areas missing prior :slight_smile:

Can you also give me some photos of the Hekate stats pages if your able to.

I have another EMMC module but no way to transfer the data to it. I have checked for shorts on emmc capacitors and they seemed fine.
I am not familiar with L4T. I would be able to install it but i dont know what to look for.
When it doesnt start up i get these situations:
-Boot heckate, boot HOS, get both nintendo logos, then error 2002-3539
-Boot heckate, boot HOS, get both nintendo logos,then black screen
-Boot heckate, boot HOS, get 1 nintendo logo, then blackscreen
-Boot heckate, boot HOS, get both nintendo logos, and then its stuck on second logo

HOWEVER.Sometimes now instead of booting heckate, i get backlight on ,and no display.
Now the error is changed to 2162-0002(IN HOS! CFW BOOTS). I dont know when it changed bcs i havent payed much attention to it recently. I will provide pictures with the back of the board soon.
This is the front for now (ik its messy but this is not final product. This is in “f** it im done” stage):



Update. There was another theory that the connector that the emmc slides in might be loose or might be cracks under it. I removed it. Everything was ok. I put it back and i checked if its properly on. Everything is ok. I also changed the 4.7k resistors on command and reset lines since they looked shabby. Now they look better but this time the error did not dissapear at all.

I also checked in a rush for shorts ,at the capacitors near the charging chip (since from what i found, for 2162-0002 people just change the charging chip) and they all looked ok. Plus if i had actual power problems i think i would not be able to boot Atmosphere.

well the final hail marry was to rebuild the nand from scratch . After that the error went away. I am able to boot HOS. Ill try to boot atmosphere,load up a software and try to replicate the errors that i was getting before when using it .
However i doubt that its fixed. I either have a faulty emmc that ends up corrupting data, or i have a component that ends up corrupting data on emmc. IDK

Is thi the error your getting now? If no, then probably not M92 issue.

If your still getting error 2002-3539 then as per Calvin’s post, he seems to be indicating it’s related to the GC PCB, so I’m guessing you’ve already tried booting with this disconncted? Either that or more than one thing can cause the error, perhaps the Ram issues he had in his case were responsible :thinking:

Can you show me an image of the (i forget the name) console info / battery info within Hekate?

Can you also show me an image of the otherside of the board in it’s entirety.

You can backup all data on both EMMC’s, then put that in a safe place, then transfer one to the other in Hekate, just to rule out the EMMC module itself. Of course be aware of update fuse mismatches and the like (not an issue if your HOS is on the latest version on your current patient board and of course because it’s modded)

You reference calvin post twice but as i said, i feel like our problems are verry different. I got 2002-3539 non stop. Then it changed to 2162-0002 referencing a power issue. However i have no shorts,charges just fine,boots atmosphere, i dont get errors while reading game cards.
His problems started with 2002-2629 , as he stated in his post, that was bcs GC. And the screen problems were bcs of ram. Totally different from what i was experiencing.

Not entirely… while Ram often causes graphical artifacts, it can (and does) cause boot issues, including no boot black sreen, halt on logo etc which does fit your prior symptoms, hence why I bring up that post. Bad Ram can also cause EMMC issues, which again, semi fits what you mentioned earlier.

Though, I will just ask again as you still haven’t confirmed, did the error remain even with the GC PCB disconnected from the board and digitizer?, I need to know bud in order to rule things out :slight_smile:

If you feel confident enough I’d swap M92 IC out just to take it out the equation, though perhaps after I see images of the back of the board and those from Hekate. I’m wanting to verify your Fuel gauge amongst other things. After, I’d also remove your solder mask from the board, in particulary around M92 and in future refrain from using it :slight_smile: (just incase there is issues below, for example, disconnected passive)

Anyway, verify EMMC module first. Then if/when it plays up again, boot into L4T, if it doesn’t boot into L4T then you can pretty much say for sure it’s Ram/SoC issue or less likely some key rail for them. If it does boot into L4T while it’s in this intermittent no boot state then you can verify WIFI/BT works, possibly even do a Ram test (pretty sure you can do this) etc

I see. I “ruled” ram out since i did a test on them and they did not throw errors. But who knows.
So i havent tested without the gc-digitezer connected.
Last night as mentioned i did a full nand rebuild for 16.1. I tryed to mess with it and it did not do any funny bussines anymore.
Today i wanted to update HOS to 17.0 directly through nintendo servers. That bricked my console XD I get first logo then blackscreen.
Now i am in the process of rebuilding the nand again, but on 17.0.
And about M92 ic , ive got the skills, but i dont want to swap anything for now, since my main suspect for now is the firmware on emmc. Ill finish rebuilding it again, assemble it and see if that finnaly fixed it.
If it acts up again, ill go ahead and do the L4T.

would a test in atmosphere work? It would be the same thing right? I have wifi, bluetooth in atmosphere (in case you are not aware, atmosphere is just a copy of HOS jailbrokened) . I also did a test in the past for RAM adresses in atmosphere. Are these tests valid?

So update. After upgrading from nintendo servers to 17.0, console is bricked again. Now i get first logo,then black screen.
I have tryed to take the whole procces from 0, rebuild the nand with 17.0 (since my fuses are burnt and cant put back 16.1) . I have rebuilt the nand 3 times now, and nothing happenes. I get first logo then blackscreen.I also changed the gamecard reader from another V2 and still same results.
EmuEMMC still works . Boots Atmosphere just fine.