Nintendo Switch Failed eMMC Fixable?
Вставка
- Опубліковано 3 жов 2024
- See the everyday life of working a retro modding company. Learn hardware design, mods, software, FPGA, LCDs and more.
Discord / discord
Wiki www.retrosix.wiki
Instagram / retrosixuk
Facebook / retrosixuk
Store www.retrosix.c...
Ok now THIS is the exact kind of content I wanted to come across. I’m a beginner learning about repairs and I’ve got pretty good at replacing components/micros soldering, but I’ve never really like the whole “blind swapping” of component. Just like how I used to get in trouble for asking my maths teacher too many questions, I don’t want to be told something is something, I want to know why something is something, so swapping components on a whim never really felt like real fault finding. I have a water damaged switch I was trying to repair for my boss that wasn’t docking, and now suddenly wont boot after being left for a month, this will definitely help with diagnosis. Time to get a bench supply I think.
Thanks, great video!
Awesome glad it's useful to you
@@RetroSix where is the next video about this console?..
Thank you so much for your website and this video. It's so sad that after months of watching youtube videos, your video just now popped up. If I would have seen this earlier, I would have saved so much time and headaches while learning the switch. Again thank you so much, truly a goat!
Do you have the 2nd part to this? I have one with the exact same issue you described here but I do not have an oscilloscope.
I just found this video and want to say THANK YOU!
I do have to say that I'm only about half way into watching and this is the most detailed video I have seen yet. Thank you for your hard work and the wiki that I'm about to check out.. I have fixed about a half a dozen switch boards but this has opened my eyes a lot more.
Thanks. Gkad it's useful
@RetroSix I have quite a few motherboards what didn't come with emmc-s due to data security reasons, is there any way I can recover them to working boards with atmosphere? Thanks
Just out of curiosity, will a switch with bad eMMC boot into RCM and does the eMMC area make a high pitched noise? I don't have a power supply like this to test with but I am intrigued by your video. I'm glad I came across it.
thanks for sharing your knowledge
This just happened to My switch :( (Goes to Nintendo logo and shuts off) Happened when I went to install the latest switch update. Was really hoping to find some kind of fix I'm gonna keep looking but If nothing I'll reflow my NAND and hope.
Same updated Switch and it crashed
Any fix?
@@shaundeonarine3333 update atmosphere
The only way to recover your switch is to install modchip. But u can no longer boot official firmware
@@aduemmanuel1844 nope just needed to update atmosphere:)
Thanks a lot for sharing your knowledge.
Best regards
great video will you be showing same type trouble shooting on a Nintendo Switch Lite?
I realize that this video is now years old, and this probably isn't relevant anymore, but at 24:17 you can see a stray solder ball on the left side of the EMMC. I wonder if that rolled out from underneath the chip?
Hi, I have been repairing devices for 13 years and recently in last few years started on switches and consoles such as PS5 I am very good at the hardware side, but get bit stuck on the more complex board issues, I loved the format and way you explain things, totally awesome. I have a switch I am determined to fix for a good friend, but bit stuck trying to find a device/location to take a 10 ohm resistor from to trick the switch into thinking it has a battery fitted, would really appreciate you help. Thank you so much.
Sure thing. Jump on our discord.gg/RetroSix and ask for assistance
brilliant work ..cant thank you enough ...
Your welcome, glad it's useful
how did it end? Did the switch itself worked? (-> eMMC bad)
I have the same problem. Looks like I need the read how to rebuild the eMMC. It´s an old switch, but I would like it to work "normal". (rebuild eMMC) Do you know a way to do this?
Btw, the wiki is really good !
I have several now with presumed bad eMMC until I find more ways to confirm it (if RCM isn't an option as model is too new) as rebuild eMMC isn't a thing yet either. I've found eMMC boot stages now on a wiki that help at least confirm working eMMC and the more I test the more confident I will be if it's a solid test for bad eMMC or not
@@RetroSix Hello I have a switch in atm where the Emmc has been damaged so I know it’s not working but it’s an old model so can hack. Can I rebuild a different eMMC and get the console working or is this console essentially dead now?
You are a saint for making a guide.
Thank you for the work you’ve put in.
This is great! Thank you for putting up a wiki
Your welcome
Got a system I'm working on now that's missing the 3.3v on V8 around the MAX77620A. What chip supplies that voltage?
Odd, you seem to pass right over V8 and say it's 0, but all the documentation I've seen seems to indicate it's supposed to be 3.3v.
I would love to see an in depth analysis of the oled model. There is way too little info out there on its board
Somewhat unrelated, ok maybe totally unrelated, but I'm looking at your wiki about joycons not connecting wirelessly or charging and it shows a photo of a cap, some resistors and such but I'm looking at my board for the right controller and don't see what I your photo at all. Any chance you could update the wiki to include the other versions of the board? I've looked all over and don't even see anything similar.
Great detailed video there. Do you have Switch OLED version of the EMMC fix?
Not reverse engineered the OLED yet but it's likely similar
thank you for this video. I indeed have a bad emmc chip
I've seen eMMC recovered, but it was a case of relating newer firmware after failing to revert to old firmware.
Hello there! thats really cool, that someone do real diagnostic instead of blind swap m92 or bq chip. Thx for the guide!
where are you located?
Simply amazing, thank you for all your invaluable work! Did I miss the follow up video?
I have 2 switches that take 5v or 15v fine ( reading at BQ cap ) no shorts anywhere.. And are just stuck at .38-.41 Amp? Even with fully charged batteries ( 3 ) system will.not boot or show.anything.. It gets 3.3v at backlight but not even backlight is turning on. Seems like Cpu isnt even booting?
Checkout RetroSix.wiki and follow the guide for checking BQ charger fully, independent of system. Also try the boot from bench guide. Failing that join discord.gg/RetroSix
NO part 2 :(((( ?
woooooooooowwwwww GREAT GREAT GREAT!!! Thanks so much for this guide!! I wish you much success
Thanks
My M92 ic on normal Nintendo switch keeps shorting... I replaced twice. it always shorts out when i try to charge it. Any idea what could be causing it?
Possibly one of the MOSFETs are shorted so once it negotiates 15V it goes back into a non-15V tolerant pin
The best repair video thanks you save my switch
fantastic work! i love this - hate changing things without understanding whats wrong! :)
So I'm confused.. if the EMMC is faulty and requires a correlated CPU; is it just dead? I was looking at performing a reball but it sounds like that won't work.
Correct. If eMMC is bad your only option is RCM booting into unsigned firmwares and running it as a hacked switch if it's old enough to do so
@@RetroSix If you have a backup of keys, you could rebuild an emmc from scratch (including boot0/1) or I am mistaken ?
@@nicolasmop4033 I am not sure if there is a process yet to match the keys but yes it is techincally possible if you can rebuild an eMMC to match
Thank you for the work, the best diagnose video for switch. If I may ask you a question. The vsys line draws around 0.5a on my lite. Using ipa to detect heat showed me only one spot and thst is the APU. It gets hot on the lower left corner of the silver part on the APU the one where nvidia is written on. Is this a dead short of the apu?
Could be normal it does pull that sign a screen. Without a screen though that's too much and likely BGA balls underneath shorting or bad APU
@@RetroSix thank you for being so kind and taking time for an answer. I really appreciate that! If I may ask you an ongoing question? It is just the pcb without battery etc. In your video you said it should pull nearly 0a otherwise look for hot areas. Also when I plug in a Usb meter. The meter makes infinit restarts. Diode readings are good on m92 and bq. Fuse all works. That would then likely mean broken connections under the APU right?
@kurtvongurt758 ah yes if you are injecting as part of dead test then yes. Follow RetroSix.wiki guide on dead test. Injecting 1V into each rail. Confirm that way for sure a short
@@RetroSix Awesome, this will help for sure. I know your site and it is very helpful. I will do the search for dead short today. 🙌
where part 2 of this video havent seen it in yours videos
is amazing all video but is posible make sysnand from 0 to funtional on board erista? thanks desde colombia
This was awesome. Thank you for putting this info out.
Hey man, can you help me? So, after installing hwfly v5 on a v2 Switch and the hekate software, when I try to do the first steps, which include the eMMC backup, I receive the message "Fail to initialize the eMMC, the eMMC is initialized in slower mode." So, what should I do?
1. I tried reflowing the eMMC, but it was unsuccessful.
2. I attempted reballing the eMMC, but it didn't work either.
3. I purchased a new eMMC from AliExpress, but still no success.
4. I removed the hwfly and placed the eMMC back in its original location on the motherboard, but now I see a pink screen.
Therefore, I believe the problem lies not with the eMMC but with the communication between the eMMC and the GPU. Are there any tests I can perform, such as using a multimeter or something else, to confirm this?
The hack interferes with the eMMC data lines so likely during install the CPU ball is slightly shorting to another. Inspect where your mod was and any solder and test points to make sure there are no joins
@@RetroSix OK, I understand, I'll open it and see it again, thanks for answering, any news I'll comment here
although not really practicable for someone to do at home.
I use a working unpatched switch and hakete, to test the suspected dead emmcs in. (even from had boards.
works well.
Yep I do that, but most switches coming in are not hackable so I would like to continue to find ways to prove the eMMC. I found a very good way here www.retrosix.wiki/emmc-booting
great video
Thanks
what if a rail is shorting another rail and you ended up injecting 4 volts into the CPU?
It wouldn't be good. But I've never known that or see a way that would feasibly happen
Hi. I send my switch to a repair shop. I see that they sold a resistor near the battery as you did. I also see a blob of metal on the USB c Port. I was wondering if the resistor can cause some problems because i see that the battery drain a lot in standby and even worse when the switch Is off. It drains like 30% a day on standby and 50% off. Thanks for any help
It shouldn't do no. But depends what resistor they used and it shouldn't be left on
What a fantastic video. I look forward to learning from the way you work.
quick question, is the WiFi chip on the second stage of the boot sequence, I know sometimes it hangs at the switch logo. sometimes it orange screens etc.
I've not checked that. I could remove eMMC and check for signals from chip. I suspect it's first stage
great work congrats
so i thought i would try this on a working switch just to see what happens. So with both wires connected and injecting 4.2v and 0.01A i get 4.02v and 0.00A and 0.00W...Then shorting the 2 pins the voltage drops to round about 3.60v then back upto 4.02V the amps goto 0.001a then back to 0.000a and get 0.003W which then goes to 0.000W...this is on a working switch, is this expected? My bench supply is a ROCKSEED RS305P...
Can you jump on our discord and do video. But you need to limit current to 1A as 0.01A wont allow it to boot
@@RetroSix sorry yes i guess that should be 1.000 milliamps which is 1 volt
@@johncarr3458 do a video on our discord so I can see your full setup and know your testing right
@@RetroSix yeah have posted a video
Muchas gracias esta super completo
De verdad 👍
Hi, awesome channel!! I have one question, my nintendo switch won't boot into system, the only things i see is a black screen, so i decide to install a mod chip to rebuild the emmc, but the custom firmware do the same thing
On your experience its probably an hardware isse after bootup on emmc?
The Switch turn on and i can see hekate, but when i try to launch the SO i cant see nothing
Thanks
Likely MAX77620
@@RetroSix if the ic dosn't work probably there is a cap shorted near the ic? I haven't shorts
Yes short or bad IC
@@RetroSix in the info battery on hekate the max77621 pmic say v4, but on a working switch say v18
The caps near the ic are not shorted so at this point i think i need to replace the chip
Is this correct for you? Or i said a wrong thing?
loving that site yours
love these kind of vids, lots of info
Part 2?
Hi, thanks for the video.
I tested your way, and I got all good voltages on the main regulator, I got all good voltages (1V) en the bottom secondary regulator, and the left regulator I get 0, nothing.
The switch doesn't past second boot (always stays at 0.2A).
Symptoms: Black screen, backlit on.
It feels like bad emmc, but I'd like to have your opinion.
Thankfully this switch is hackable so I'll get to try this way in the future.
Definitely sounds like eMMC or the bottom regulator under the CPU. Easy test use use the fuse.bin file that loads the screen up without needing second stage boot. If it's just bad eMMC then fuse will load and show
Hi. I'm struggeling with no display in handheld mode, but in dock the switch shows up and it boots fully. I've checked the all of the ribbons and cleaned everything. This occured after an hwfly install. The hwfly worked as it should and booted up, but the display does not show no more.
Symptoms: No display, backlit on, fully booted and can hear me navigating the menu.
Do you havce any idea what I could check, I can clearly hear that you have way better control with diagnostics and was hoping for a finger pointer for me to check..
It might of blown the P1USB or you might of damaged a trace or pin. Remove the mod and perhaps replace P1USB
@@RetroSix thanks for the reply. I've checked the voltage and cap/resistors at the LCD driver ic, but it's seems to show the right value and 5v at the Vin.
P1USB? struggling to find info on it(currently on my phone).
@@PRaZTuTz voltages won't check much. Would need oscilloscope. P1USB handles telling CPU it is docked and passing signals. It might be faulting to trigger dock output when not docked. Also when you turn it on can you hear the audio and does touchscreen work when not docked
@@RetroSix Ok, interesting read. Yes, I do hear audio and the switch registers my inputs, so i would assume that the P1USB wouldn't be telling the switch that it's docked when it is in handheld mode? Cause the audio and touch input would've been disabled and gone to external audio/video..
@@PRaZTuTz correct. Is this a switch or lite? I'm guessing just your backlight is dead or LCD bias regulator. Try getting your phone torch and shining it right on the screen and looking from all angles for the slightest visual if you have any screen showing just without backlight
Well done.
when I put voltage on the vsys it goes to 0.2A but then off after a few seconds
Usually it's a bad BQ that does that. But it can by almost any chip. Use thermal camera ideally to see which chip is getting slightly warmer that will be the faulty chip
Could you tell what is the second stage boot chip on a lite? Thanks
The CPU tells the MAX regulator to turn on the second stage regulators after it's read the OS from the eMMC
@@RetroSix Thank you for your responce, but i mean the part name/number
@@oleggrigorev7168 retrosix.co.uk/Nintendo-Switch-Lite-Power-Management-IC-MAX77620H-p478604271
the one where u tested on the 1.3v coil top left of max77620A am not getting 1.3v there does that mean i have a bad chip?
Or that rail is shorted. Do you get high current draw on bench when powering it on?
@@RetroSix well i don't the machine u use to test that so i rely on battery with the charger. Just like everything u did in the video i get voltage just not the 1.3v. On my multimeter in ohms mode it start at 7.6 ohms then continue to climb up. If i check in diode mode it will give me 0.526v.
Measure resistance to ground of 1.3V rail compared to 1.1V rail while power is off
@@RetroSix ok so am getting 2.9 resistance to ground on 1.3v rail and on the 1.1V rail i get 202 resistance
@@zane383 remove the inductor of the 1.3V rail then measure resistance to ground on that top pad where inductor was again
hello can a emmc be replaced
Not without patching the console or cloning a known fully working eMMC
@RetroSix I have quite a few motherboards what didn't come with emmc-s due to data security reasons, is there any way I can recover them to working boards with atmosphere? Thanks
It's not possible unless you already have a copy of the eMMC that came with that switch to clone the internal ID
@RetroSix ah I see so it's not possible even with some cfw magic either at all?
Not that I know of yet @@SzebaHUN
Where can i find your guide, thank you
RetroSix.wiki website
@@RetroSix Thank you, but ive been there and i couldnt find the switch board pics ...
Homepage, Public Files. Then Switch. Think I have them there
Thanks a lot for the videos on repairing the nintendo switch.
I would like to ask a question, i hope you can have time to answer. After watching your videos i came to the conclusion that my Switch has a bad M92T36, but the charging port looks ok. Can it be a just a bad M92T36 and the port still be fine ? What are the causes that IC to get faulty ?
Thank you
@@valmonta3647 The MT92 is responsible for negotiating with a USB-C device for higher voltage charging. The best way to quickly test that is diode mode, red probe to ground, and probe with the black probe around the capacitors of the MT92. A faulty MT92 almost always results in one or more of those capacitors beeping permanently on both sides of the cap not just one.