I fixed a broken Nintendo Switch by watching YouTube videos

George Giannou
8 min readMar 28, 2020

--

It was the beginning of last summer when I started watching YouTube repair videos. For some reason the algorithm though I might be interested in this kind of videos, and it was right, I got hooked. The video platform is full of channels doing repairs on various devices, and it seemed like there was a trend on fixing Nintendo Switch consoles starting. You can easily find videos of video creators that buy lots of broken consoles and try to repair as many as they can. I believed that I could do the same and hopefully I did!

After months of on and off tinkering, waiting for parts and replacements I finally assembled a working Nintendo Switch!

After having watched a good amount of videos, and I guess after some algorithm magic, Facebook suggested that I might like a broken Nintendo Switch that was on sale for 80€ near my location. It only included the console, no peripherals and no controllers. Of course, I was interested, I had watched so many hours of repair videos. And so it began.

The device. Original price was at 80€. Good condition, if you don’t account for the dog bite marks on the front and the back of the device.

At first I contacted the seller, trying to understand the issue of the console. I wouldn’t be interested in a soaked console. You have to be lucky to fix one of them. The seller told me they plugged the console into a public computer and it suddenly turned off, refusing to turn on again. As they were typing to me, I was furiously searching the Internet for clues. A lot of forum threads started popping out mentioning that if the device has AutoRCM and had its battery drained (and/or it suddenly turned off) it might need a couple of days of charging.

But hey, whats AutoRCM? I own a Nintendo Switch since its launch and I had some clues but I wasn’t that sure. Long story short, in order to boot a custom firmware on the device, you need to inject a payload via the device’s USB port. Custom firmware (often aliased as CFW) is used for homebrew development and sometimes piracy. In order to inject a payload and ultimately boot a CFW, the device must be booted into RCM (short for ReCovery Mode) mode. In order to achieve this you need to press the Vol+button and short a pin in the right Joy Con rail. The device then boots into RCM, accepting code sent by USB. Side note, the device only accepts signed code but there is a bug in the SoC, which when triggered runs whatever code is sent over USB.

AutoRCM is just a trick, the internal memory of the device gets corrupted intentionally and the device is forced to boot into RCM each time without requiring any key press or shorting of the pins. In other words, when in RCM, the device screen stays black and you need to plug it into a PC or a smartphone to inject a payload. You have no indication that the device is powered on, other than the sound a PC makes when you connect a USB device!

A Joy-Con jig is used for shorting the appropriate pins in the right Joy-Con header. Using AutoRCM eliminates the need of a jig.

I asked the seller if the device had been hacked and I was really excited when they told me that it was indeed hacked! They revealed to me that it had also AutoRCM enabled. I tried to contain my excitement, but fixing couldn’t be that easy, right?

This was my first guess. My initial thought was that the device had its battery depleted, it couldn’t fast charge in that state it had been stuck to and all it was needed was a good charge. We agreed on meeting the next day and I even got the device for 60€! After getting home, I let it charge for a couple of hours and tried connecting it to a computer. Well, nothing happened.

It had been almost a week of trying. I left the device charging for 4 days straight. No matter what I tried, I couldn’t get it to boot, or at least have it recognized by a computer. The device was certainly charging, it got warm after a while I could verify with a USB current meter that it was indeed drawing power. The device was also responsive to the power button, it drew more current when I pressed it and dropped to zero when I held it for more than 12 seconds (which is supposed to forcefully turn it off). Problem is, I couldn’t get it to connect to any PC, and trust me I tried various different PCs, Macs, smartphones and different ports of them. I also tried RCM on my Switch, which worked perfectly the first time. Well, that wasn’t the issue.

A USB current meter shows the voltage and current drawn of the device.

But I couldn’t just stop there. I mean, how could all those people on YouTube fix them so easily? It surely can be fixed! Albeit a little bit harder due to the AutoRCM hack. I was somewhat disappointed, but I got hooked. I had to fix it.

After watching some more repair videos I thought were relevant to my case, I decided to open the device. I started probing various points with a multi-meter and tried to validate my findings with those on the videos. When the repair guy probed one point on the PCB, I paused the video and took a measurement of the same point to cross-check my findings. There is no point in going into detail here, there are specific measurements that one can make that help quickly identify common issues. Even though I was measuring and checking whatever the guy checked on video, but everything was fine with the console. The same happened with the next video, sadly the same results. My device had everything perfect! All measurements were identical to a working Nintendo Switch. I spend a lot of my summer afternoons trying to troubleshoot this device, when after running out of videos to watch hopped on a forum.

I wrote a quick post on a huge thread mentioning my findings and the state of the device. Someone started replying to my messages and as we were discussing I started checking the continuity of the cable. I thought, the device is working fine but it wasn’t being detected as a USB device. Cable seemed fine and then I started checking the continuity of the PCB traces. I followed the USB data line traces and I was making sure I measured continuity. Until continuity stopped!

Test points 2 and 3 on the PCB match USB ports DATA+ and DATA-. Following those traces (marked with green), continuity seemed to stop on that little component (marked with red on the right of the component).

What was this mysterious device? I had no clue. To be honest, I still don’t. I sent a message on the forum thread hoping someone might know what that mysterious device is and if my findings are correct. I was extremely lucky, someone identified it and provided me a part number! No idea why it broke, or why it might be the cause of failure. All I know is that its a Murata DLM0NSN900HY2 Common Mode Choke. After searching I understood that it was just a filter for the USB data lines, and it was everything I needed to know. My guess is also that that device sacrificed itself to protect the rest of the PCB from that computer the previous owned plugged it into. So, to recap. The device is responsive, I could verify it with the USB current meter, but the data lines didn’t work, so no computer could recognize it and inject an RCM payload!

I quickly sourced some parts online. Shipping ended up costing more than the actual parts. They are only sold in quantities of 10, with 10 being the least you can get at GBP 1.56. Shipping was an additional 5 pounds to my address and adding tax took it to a total of GBP 7.87!

It took less than 48 hours for the parts to arrive, but now I faced an even greater issue. I don’t have the tools or the skills to do the repair myself. I started looking for shops that do PCB repairs. First I went to a local shop but the owner told me it would be really difficult, but he could give it a try outside his regular schedule. I then started looking for other shops. I went to another one and when they looked the size of the component they almost freaked out and told me no one could fix that, it requires special equipment. Yeah right. After leaving that shop I quickly searched for nearby iPhone repair companies and found one almost a km away.

I drove there, got into the shop and went straight to the counter. I tried to explain the guy there what I needed but he stopped me mid-sentence and went to the back of the shop. Another guy showed up and asked me what I needed. I tried to briefly explain to him what I needed, without sounding crazy and showed him the PCB. I gave him the Nintendo Switch motherboard and the ten replacement parts. He gave them a quick look and he told me to wait, took the PCB and the parts and went to the back of the shop.

I heard him shouting at me that the old choke looked like it had exploded! I could then hear the sweet sound of a hot air station (I had only heard it in YouTube videos) and after almost a minute we showed up again with the repaired PCB and a multi-meter. We quickly verified that the traces had continuity now and I told him that he had fixed it! Of course, I didn’t know if that was the only issue, but at least it was something. We chattered for a bit, and then I asked what I had to pay, to which he replied nothing! I was really happy. I thanked him and drove home.

After quickly and carefully assembling the Nintendo Switch, I pressed the power button. Upon inserting a USB cable, I heard the famous Windows USB device inserted chime! That did it, I had fixed it! I now have a fully working Nintendo Switch which cost me about ~70€!

I couldn’t stop there. I changed the case with a clear one, modded my old Joy Cons to match this new Nintendo Switch and finally replaced the touch screen.

As you can see, on the pictures, I couldn’t stop there. I also replaced the touchscreen, which had bite marks from the previous owner's dog (I hope). The back panel was also replaced, and I sacrificed my old Joy Cons to make them match the clear case. These changes weren’t needed, the device was functional since the first repair and added a lot to the budget, but ultimately it ended up costing less than 150€ if you account for a new 256GB SD card!

--

--