Hacking Question Will a bricked Switch boot into RCM ?

[WadsRUs]

Should a Switch that shows no signs of life (yet with a fully charged battery) boot into RCM if the Switch is 'only' bricked?

If not, and assuming that the hardware is fully working, is there any way to 'force' it to boot something, not necessarily Horizon?

And no, I've not bricked my own Switch - I bought a really cheap faulty one off Ebay just to experiment with it.

 

[Drejj]

Yes it will boot to RCM. You can never destroy the bootrom. You should be able to boot linux on a bricked switch.

 

[P3T3]

Yes but if it’s been opened and nand swamped for one that’s not paired to the switch you can’t do anything, maybe boot Linux.
Because it’s not the orig nand it won’t have key to read nand, so try to make shore it’s never been opened.

 

[WadsRUs]

Yes but if it’s been opened and nand swamped for one that’s not paired to the switch you can’t do anything, maybe boot Linux.
Because it’s not the orig nand it won’t have key to read nand, so try to make shore it’s never been opened.

It's definitely been opened (some screws missing both inside and out).

If it won't go into RCM then surely I can't even boot it into Linux?

 

[The Real Jdbye]

Should a Switch that shows no signs of life (yet with a fully charged battery) boot into RCM if the Switch is 'only' bricked?

If not, and assuming that the hardware is fully working, is there any way to 'force' it to boot something, not necessarily Horizon?

And no, I've not bricked my own Switch - I bought a really cheap faulty one off Ebay just to experiment with it.

A softbrick will boot into RCM, but when buying faulty consoles on eBay you don't really know what's wrong with them, and they can have all kinds of various hardware faults that would prevent the console from working at all.

 

[P3T3]

Every switch will boot in to rcm as long as it boots, you then can check by checking fuses etc. If it comes up as it should nand is good, if it’s not original then it will display bit of random code..
If random code then only good for Linux!

 

[WadsRUs]

Yes it will boot to RCM. You can never destroy the bootrom. You should be able to boot linux on a bricked switch.

Thanks. I had assumed that if a Switch won't boot into RCM it won't boot into anything at all.

For example, from looking at the instructions here -

https://gbatemp.net/threads/quick-tuto-how-to-boot-linux-on-your-switch.501918/

RCM is required

 

[P3T3]

And if a patched unit go it to rcm but won’t be able to boot payload!

--------------------- MERGED ---------------------------

And if a patched unit go it to rcm but won’t be able to boot payload!

It’s a big gamble, better buying a second hand one!

 

[WadsRUs]

A softbrick will boot into RCM, but when buying faulty consoles on eBay you don't really know what's wrong with them, and they can have all kinds of various hardware faults that would prevent the console from working at all.

So, as this won't even go into RCM (which is apparently needed to run Linux) it's basically non-bootable.

I guess I'll use it for parts then.

 

[Zonark]

I was thinking you may be able to rebuild the NAND If I recall correctly I think there is a tut on here or a forum already made for that. Obviously the files needed are copyrighted so those won’t be posted but you maybe able to read the key (like you used to with the lite on 360 drives) and inject it into the built NAND for repair. Question is if they did try to swap it (NAND) if they damaged it. If your interested in getting rid of it at all I’m interested

 

[WadsRUs]

I was thinking you may be able to rebuild the NAND If I recall correctly I think there is a tut on here or a forum already made for that. Obviously the files needed are copyrighted so those won’t be posted but you maybe able to read the key (like you used to with the lite on 360 drives) and inject it into the built NAND for repair. Question is if they did try to swap it (NAND) if they damaged it. If your interested in getting rid of it at all I’m interested

Thanks. Sounds interesting (as long as I can do it legally).

How though could I even send the relevant code to the Switch seeing as it won't even boot into RCM? It's just not showing any signs of life.

 

[Zonark]

Thanks. Sounds interesting (as long as I can do it legally).

How though could I even send the relevant code to the Switch seeing as it won't even boot into RCM? It's just not showing any signs of life.

This will be all done via a probe device essentially. (USB to serial male with certain pins attached) the device will basically have power kinda like how a tv has power with a broken still is attempting to display a picture even though it can’t.

Quick test though, when you put it in the dock does the green light come on on the dock then turn off?

 

[WadsRUs]

This will be all done via a probe device essentially. (USB to serial male with certain pins attached) the device will basically have power kinda like how a tv has power with a broken still is attempting to display a picture even though it can’t.

Quick test though, when you put it in the dock does the green light come on on the dock then turn off?

I see, thanks. Very curious how this is done. It would certainly be a useful way to see if the Switch is functional in any way. I have good soldering skills and am used to working with very small components and PCBs.

Correct, the dock's green LED briefly comes on then goes off when I place the Switch into the dock.

 

[mattytrog]

Thanks. Sounds interesting (as long as I can do it legally).

How though could I even send the relevant code to the Switch seeing as it won't even boot into RCM? It's just not showing any signs of life.

You have to ask why you aren't getting RCM.

Do a quick check of m92t36 pin 5 (VCC-IN). If between 3v and 6v, then assume it's good for now.

If good, check pmic on reverse of board. If ok, check buck regulators. If ok check usb port hasn't attempted to be swapped and bq24193 is damaged because of that.

Switches are a doddle to repair. As long as you have an intact prodinfo and the stacking connectors aren't fucked. Shame to strip it.

 

[Zonark]

I see, thanks. Very curious how this is done. It would certainly be a useful way to see if the Switch is functional in any way. I have good soldering skills and am used to working with very small components and PCBs.

Correct, the dock's green LED briefly comes on then goes off when I place the Switch into the dock.

The light coming on is a good sign honestly as long as they didn’t swap nand you should be able to fix this. With the power coming on this means the power board should be working properly and that the USB port is working properly. Like the earlier commented it will probably be the chip that controls wether the battery knows it’s charged or dead. I’d replace that first then attempt to boot. Might even be a cable unplugged too

 

[xXDungeon_CrawlerXx]

Well, why not installing Linux to a "bricked NAND" to linux as default?
(I know, this won't work, would be funny, tho)

 

[WadsRUs]

The light coming on is a good sign honestly as long as they didn’t swap nand you should be able to fix this. With the power coming on this means the power board should be working properly and that the USB port is working properly. Like the earlier commented it will probably be the chip that controls wether the battery knows it’s charged or dead. I’d replace that first then attempt to boot. Might even be a cable unplugged too

Are you referring to the M92T36 power charging control chip?

I guess you don't mean the BQ24193 as that's part of the battery and this Switch won't even boot with a known good battery attached.

Cables are all plugged in, no signs of corrosion (I've checked them all). I should say that I've dismantled Switches before, even repaired a few (usually they have damaged charging ports, it seems to be a common problem).

 

[Zonark]

Are you referring to the M92T36 power charging control chip?

I guess you don't mean the BQ24193 as that's part of the battery and this Switch won't even boot with a known good battery attached.

Cables are all plugged in, no signs of corrosion (I've checked them all). I should say that I've dismantled Switches before, even repaired a few (usually they have damaged charging ports, it seems to be a common problem).

Yes the M92T36 I believe will need to be replace basically this unit more than likely had been exploited maybe had Linux used on it or auto rcm running and battery died during it. Have you tried any of the auto rcm repair methods yet? Swapping battery and such?

Also how did you determine that the NAND was bad?

 

[WadsRUs]

Yes the M92T36 I believe will need to be replace basically this unit more than likely had been exploited maybe had Linux used on it or auto rcm running and battery died during it. Have you tried any of the auto rcm repair methods yet? Swapping battery and such?

Okay, I'll replace the charging chip (I have some here).

Have tried replacing the battery with a known good one but still no life in the Switch.

I've not played around with AutoRCM before - can it leave a Switch in the state that this one is in with no signs of life at all and with the Switch refusing to even utilise RCM mode?

 

[Zonark]

Okay, I'll replace the charging chip (I have some here).

Have tried replacing the battery with a known good one but still no life in the Switch.

I've not played around with AutoRCM before - can it leave a Switch in the state that this one is in with no signs of life at all and with the Switch refusing to even utilise RCM mode?

If they used autorcm with Linux and can cause the chip to “forget” if it’s charged enough to power on causing it to forever be in the lower power state where it’s being told that the battery has not enough power to boot. Keep me posted on the chip as well it should technically work as long as everything else is fine with the device ie the nand.

Odds are that the original user of this used an after market charger and messed up that chip

Have you checked online to verify if this unit is exploitable or not? (I’d check this first cuz if not they used and after market charger more than likely and replacing chip should work 100%)

If it is exploitable it could have been pikabricked and a NAND rebuild would be needing done