Hardware Switch doesn't charge at all after RCM-X86 install

[PRAGMA]

After a troublesome RCM-X86 UF2 rodeo, I realized that my Switch no longer charged after installing the RCM-X86 (5-wire setup, GND,3V3,D-,D+,D0).
I do not know what could have caused it at all as I was very careful.

- I've removed the RCM-X86 completely including all wires, bits of solder I made, e.t.c.
- Ive gone over and cleaned with IPA after removing everything.
- Throughout both installation and uninstallation I was being very careful. No shorting ever occurred.
- The switch was always off whenever I was fiddling with it, other than when I was attempting to enter the SAMD21 bootloader.

What's strange is absolutely nothing happens when I try to charge it. I've tried 3x different cables including the official charger.
But, I can still use the Red-OS Gear injector to enter RCM using a Jig.

In Hekate, it never detects the USB as being connected, nor does the wattage draw differ not that I can notice anyway.
It stays around -450 to -500.

I've been trying to install an RCM-X86 for a long time now and my last attempt was 4 years ago.
Strangely both times have had the exact same effect, no charging. Both times I've been careful so I'm very confused why.
Last time I did notice issues with the port itself, specifically balls of solder getting stuck between the pins. But this time I've not got a single thing to point at as a cause.

The issue was noticed while trying to enter the bootloader of the RCM-X86 chip and encountering weird errors.
See https://gbatemp.net/threads/white-rcm-x86-wont-enter-samd21-update-mode.626316

Any advice would be great. I'm done with RCM-X86 and just about done with soldering in general!

 

[linuxares]

Can it maybe be the hdmi chip?
Also double check the fuses to make sure non of those poped

 

[thesjaakspoiler]

It could have been static electricity.
Or the chip just failing at the same time you were working on it.
For most people their Switch just breaks down without them doing anything.
About the power usage, chip like these draw something like 10~15 mA.
You can hardly measure that accurately with Hekate.

 

[PRAGMA]

Can it maybe be the hdmi chip?
Also double check the fuses to make sure non of those poped

Any advice on how to check them? I have a multimeter but barely know how to use it.

It could have been static electricity.
Or the chip just failing at the same time you were working on it.
For most people their Switch just breaks down without them doing anything.
About the power usage, chip like these draw something like 10~15 mA.
You can hardly measure that accurately with Hekate.

I think you may be right. I don't ground myself when working on soldering. My best guess as to what has happened is two things.

I might have been too liberal with the IPA and have not waited for it to dry up. I never really let it dry to be honest. I would solder on a cable, give a spray with my IPA 99% bottle, rub the area with a qtip, then move on. There was often a time when I would connect the battery and continue fairly soon after this.

Or, something had shorted while trying to enter bootloader mode on the RCM-X86. Maybe I moved the chip which would have made the cables move a tad bit and maybe get closer to a short in a way. E.g., a slightly loose connection that moved the solder a bit near a capacitor.

The connection I had to this side of the capacitor was moderately loose. You could jiggle the cable and notice the blob of solder moving. This connection was also the only one that was loose like this. It's possible I made a bridge as shown in the picture at least temporarily/depending on how the cable was positioned.

Considering that capacitor was close to the only loose connection I wouldn't be surprised if this was the cause or at least complications caused by my bad soldering. I would also spray IPA near-ish to this area, it would have definitely hit around that left side of the CPU. It's possible the IPA didn't dry up before I turned it on, causing a short with some of the pins or traces. (Can IPA short traces?)

---

If anyone has any advice for me to check something, let me know. I have a multimeter I could use to test stuff, and a heat gun that I could use to replace a part or capacitor, but I'm not the most experienced. I also would have to wait around a month for the parts to arrive, meaning it could take a very long time to try, regardless if it was the fault or not.

 

[binkinator]

Any advice on how to check them? I have a multimeter but barely know how to use it.



I think you may be right. I don't ground myself when working on soldering. My best guess as to what has happened is two things.

I might have been too liberal with the IPA and have not waited for it to dry up. I never really let it dry to be honest. I would solder on a cable, give a spray with my IPA 99% bottle, rub the area with a qtip, then move on. There was often a time when I would connect the battery and continue fairly soon after this.

Or, something had shorted while trying to enter bootloader mode on the RCM-X86. Maybe I moved the chip which would have made the cables move a tad bit and maybe get closer to a short in a way. E.g., a slightly loose connection that moved the solder a bit near a capacitor.

View attachment 353932

The connection I had to this side of the capacitor was moderately loose. You could jiggle the cable and notice the blob of solder moving. This connection was also the only one that was loose like this. It's possible I made a bridge as shown in the picture at least temporarily/depending on how the cable was positioned.

Considering that capacitor was close to the only loose connection I wouldn't be surprised if this was the cause or at least complications caused by my bad soldering. I would also spray IPA near-ish to this area, it would have definitely hit around that left side of the CPU. It's possible the IPA didn't dry up before I turned it on, causing a short with some of the pins or traces. (Can IPA short traces?)

---

If anyone has any advice for me to check something, let me know. I have a multimeter I could use to test stuff, and a heat gun that I could use to replace a part or capacitor, but I'm not the most experienced. I also would have to wait around a month for the parts to arrive, meaning it could take a very long time to try, regardless if it was the fault or not.

m92t36 is a common failure. Here’s a thread that goes into some detail of some tests you can try.

https://www.tronicsfixforum.com/t/a-funny-switch-blow-up-3x-m92t36/3979

You’re goingto have to learn how to use that multimeter to test for shorts and opens.

There is a parts thread with some m92t36 troubleshooting in it. Not sure it was applicable but worth a search since it’s so common.

https://www.tronicsfixforum.com/search?q=M92T36

https://gbatemp.net/search/3976260/?q=M92T36&t=post&c[child_nodes]=1&c[nodes][0]=283&o=relevance

 

[PRAGMA]

m92t36 is a common failure. Here’s a thread that goes into some detail of some tests you can try.

https://www.tronicsfixforum.com/t/a-funny-switch-blow-up-3x-m92t36/3979

You’re goingto have to learn how to use that multimeter to test for shorts and opens.

There is a parts thread with some m92t36 troubleshooting in it. Not sure it was applicable but worth a search since it’s so common.

https://www.tronicsfixforum.com/search?q=M92T36

https://gbatemp.net/search/3976260/?q=M92T36&t=post&c[child_nodes]=1&c[nodes][0]=283&o=relevance

Thanks for this information. Believe it or not I've had a lengthy chat with none other than... .... ChatGPT about some stuff and boy is it correct, but also fun to talk with! I don't even have to pretty up my sentences like I'm talking to my boss!

I've spoken with it about testing with a multimeter and I think I know enough to begin, other than the resistance value of the capacitors. However, my safe guess is to test in the 2,000 ohms range. 1,000 ohms range is also probably fine (from my research).

Anyway, the chip you stated is also my best guess as to what's gone wrong but before ordering it I will test the chip tomorrow (late for me) and will see first. However, it came to me that it very very likely is that chip that is wrong. Think about it. The USB port does not work AT ALL once the Switch is on. But, remember I said how my Hardware Injector by Red-OS still works. This is likely because when you enter RCM, that chip may not be in control yet. On the GBATemp thread on SAMD21 RCM software, I remember seeing talk about RCM not being active for USB stuff or something along the lines like that. Which sounds plausible.

Also, I found a very very nice place I feel like I should link in this thread for others who end up here as it has a TON of useful information about this exact problem, https://repair.wiki/w/Nintendo_Switch_Not_Charging

Tomorrow will be a fun day of either rejoice, or pain.
If I successfully fix charging on this unit, I'm going to literally piss on that rcmx86 and grab a Trinket M0, it's a cursed device to me at this point.

 

[binkinator]

Thanks for this information. Believe it or not I've had a lengthy chat with none other than... .... ChatGPT about some stuff and boy is it correct, but also fun to talk with! I don't even have to pretty up my sentences like I'm talking to my boss!

I've spoken with it about testing with a multimeter and I think I know enough to begin, other than the resistance value of the capacitors. However, my safe guess is to test in the 2,000 ohms range. 1,000 ohms range is also probably fine (from my research).

Anyway, the chip you stated is also my best guess as to what's gone wrong but before ordering it I will test the chip tomorrow (late for me) and will see first. However, it came to me that it very very likely is that chip that is wrong. Think about it. The USB port does not work AT ALL once the Switch is on. But, remember I said how my Hardware Injector by Red-OS still works. This is likely because when you enter RCM, that chip may not be in control yet. On the GBATemp thread on SAMD21 RCM software, I remember seeing talk about RCM not being active for USB stuff or something along the lines like that. Which sounds plausible.

Also, I found a very very nice place I feel like I should link in this thread for others who end up here as it has a TON of useful information about this exact problem, https://repair.wiki/w/Nintendo_Switch_Not_Charging

Tomorrow will be a fun day of either rejoice, or pain.
If I successfully fix charging on this unit, I'm going to literally piss on that rcmx86 and grab a Trinket M0, it's a cursed device to me at this point.

That’s amazing actually. Both the ”learning how to test for shorts and opens” but AND the awesome link w/ everything all in one place. Great find.

Here’s a place that has the chips you need. They are pretty cheap (meaning you might want to splurge a little and buy more than one…just in case.)

https://m.aliexpress.us/item/3256804793227097.html

While you’re learning to go pro at this, you might want to replace your USB connector while you’re at it. Can be a PITA but if you suspect it could be in the mix…

 

[PRAGMA]

That’s amazing actually. Both the ”learning how to test for shorts and opens” but AND the awesome link w/ everything all in one place. Great find.

Here’s a place that has the chips you need. They are pretty cheap (meaning you might want to splurge a little and buy more than one…just in case.)

https://m.aliexpress.us/item/3256804793227097.html

While you’re learning to go pro at this, you might want to replace your USB connector while you’re at it. Can be a PITA but if you suspect it could be in the mix…

I recently tried to replace the USB-C on a different switch, and oh boy was it problematic to remove the left-over solder. Unlike others who are able to almost swipe their tip over fluxed pads to swiftly and cleanly tin pads, I have not been able to do anything like that either.

Even if I try to manually do each pad, it just jumbles to a huge glob of solder. I ended up removing the bottom rows 2nd from the left pad, when trying to remove excess solder from the pins. Notice how the top row is an absolute mess of bridged blobs of circles.

This is what it looked like just after removing the old port:

How the pad actually got pulled was I was using wick to remove the excess solder, but the wick ended up getting incredibly stuck. When trying to take it off, I ended up using too much force and pulled the pad. But when I say it was incredibly stuck, it was genuinely like concrete.

I generally used wick wrong until very recently and had thought of how wick worked incorrectly (relatively) so I could probably do a better job now. I used to use it while still in the spool, or with a cut-off piece that was way too long. I learned that the bigger the surface area the wick has, the longer and harder it is to heat up a specific area or in general. Therefore cutting off a piece is vital, and a really small almost square cut helps a ton.

This is my 2nd time trying to replace the USB-C port on that exact switch (not the switch of this thread OP). I don't know how to check if the pads are correctly connected, especially the inner/bottom row. Each time I replaced the USB-C port it would only charge one way with the cable. Even still when I placed a new port onto this board with the pulled pad. Of course, the chip we are on about could be related. The issue the original port had was from an RCM-X86 install as well, where a bit of solder got stuck between 2 pins and bridged them. I don't know if that did actually cause anything bad either. Regardless I still don't really have it fixed.

 

[binkinator]

I recently tried to replace the USB-C on a different switch, and oh boy was it problematic to remove the left-over solder. Unlike others who are able to almost swipe their tip over fluxed pads to swiftly and cleanly tin pads, I have not been able to do anything like that either.

Even if I try to manually do each pad, it just jumbles to a huge glob of solder. I ended up removing the bottom rows 2nd from the left pad, when trying to remove excess solder from the pins. Notice how the top row is an absolute mess of bridged blobs of circles.

View attachment 353947

This is what it looked like just after removing the old port:

View attachment 353948

How the pad actually got pulled was I was using wick to remove the excess solder, but the wick ended up getting incredibly stuck. When trying to take it off, I ended up using too much force and pulled the pad. But when I say it was incredibly stuck, it was genuinely like concrete.

I generally used wick wrong until very recently and had thought of how wick worked incorrectly (relatively) so I could probably do a better job now. I used to use it while still in the spool, or with a cut-off piece that was way too long. I learned that the bigger the surface area the wick has, the longer and harder it is to heat up a specific area or in general. Therefore cutting off a piece is vital, and a really small almost square cut helps a ton.

This is my 2nd time trying to replace the USB-C port on that exact switch (not the switch of this thread OP). I don't know how to check if the pads are correctly connected, especially the inner/bottom row. Each time I replaced the USB-C port it would only charge one way with the cable. Even still when I placed a new port onto this board with the pulled pad. Of course, the chip we are on about could be related. The issue the original port had was from an RCM-X86 install as well, where a bit of solder got stuck between 2 pins and bridged them. I don't know if that did actually cause anything bad either. Regardless I still don't really have it fixed.

Yeah, that second row is the PITA. Kinda have to do it blind. I have a cheap USB break out tester that I can test for errors. I still make a lot of errors but I know pretty quickly when I do so I can fix them. Lots of flux and an adjustable power soldering iron is key for a lot of this stuff. Next level would be to pick up a basic hot air rework station. It makes quick work of all these tiny solder situations. As far as the whole situation, it could have been you or it equally could have just been time for one of these components to go south. All that to say this, I hope this doesn’t put you off what can be a really rewarding hobby. I hope you stick with it and get this all sorted.

 

[PRAGMA]

I tested the majority of the caps from this side of the board and this is what I'm getting.
I haven't tested a lot of the smaller black ones. The black color ones (regardless of size) would be resistors not caps right?
Can anyone advise what I should check from here?

Do note this is a HAC-CPU-21 board, so keep that in mind in terms of the layout or what caps/resistors are actually used. There's a fair amount I noticed where caps are not here that are there on the HAC-CPU-01 board.

Post automatically merged: Feb 19, 2023

I've removed the M92T36 chip and the readings of the nearby caps/resistors are somewhat different.
However, I'm having trouble reading them because of the coating of flux residue around the components.
I have 99.98>=% IPA and it doesn't seem to be doing anything.

Does anyone have any guidance on removing the flux residue?

Post automatically merged: Feb 19, 2023

Using another ChatGPT tip, I've managed to give a moderate cleaning of the flux quite easily and safely. For anyone else, the tip is to set to a safe but high temp of about 100-150 celsius (I did 100) and apply IPA to the area (in my case the entire right of the PCB basically) then heat up the area. Of course, don't heat up a specific exact area for too long. Wiggle around a bit. This worked quite well compared to just setting IPA on it and scrubbing with Q-Tips and soft-bristle brushes.

Post automatically merged: Feb 19, 2023

This is the current status after removing the chip. There's still weirdness on one of the caps below the chip where its 5.77~:0, instead of 1:0. There's also still the same weirdness between the USB-C and the chip with 2x caps still being 1:1. At least one of the caps is fine now. Any further advice would be great.

Post automatically merged: Feb 19, 2023

I've taken off the BQ24193 and I still have the same issues with the two caps around below it. Any ideas would be great.