A look at inner workings of Joycon and Nintendo Switch(github.com)
github.com
A look at inner workings of Joycon and Nintendo Switch
https://github.com/dekuNukem/Nintendo_Switch_Reverse_Engineering
61 comments
I am really impressed by how the joycons and the switch itself “just work”. I have never had any syncing issues. You can play the console with the controllers attached to the switch or with the controllers wirelessly and you can change at anytime you are playing the game. I’ve never had any issues with syncing and changing playing style. Attaching/detaching the controllers never takes more than a couple seconds even in the middle of a game.
I’ve had frustrating issues syncing joy-cons. For example, if a game only supports 1-2 players, and my controller is using the player 3 slot, the sync screen will simply ignore any attempt to sync that controller. I can fix this by going to “Change grip order” in setting, but there is no UI indication why I can’t use the controller. It’s just a bit frustrating for a few minutes until I figure it out.
Is this a common issue on other platforms? I’ve basically been using my PS4 controller as you describe (although wired more often than not simply because my setup has the console close enough for it to be convenient) and I’ve never had any issues. It always seem leaky switches between wired and wireless.
I think his comment is more that the Joycon syncing is a bit more complicated; as they have ~4 separate modes of operation - Attached, Attached to charging-controller, separate-but-together, and separate.
You can get into scenarios where things are 'off', but nothing 10 y/o me wouldn't have figured out quickly I think. It is pretty well done overall; much better than the bottom charging port.
You can get into scenarios where things are 'off', but nothing 10 y/o me wouldn't have figured out quickly I think. It is pretty well done overall; much better than the bottom charging port.
EDIT: "seem leaky" = bad auto correct for "seamlessly"
There was quite a wide spread issue with syncing the JoyCons on launch, particular with the left one - http://m.uk.ign.com/articles/2017/03/22/nintendo-confirms-sw...
I haven’t had any issues personally though, I think it’s a great console!
I haven’t had any issues personally though, I think it’s a great console!
I wish there were a simpler way to change grip orientation.
Maybe someone smarter than me can explain this:
"Also, in a bizarre move, Nintendo didn't use the traditional "one side pulled-up other side to ground" way of reading buttons, instead they used a keypad configuration where buttons are arranged in rows and columns. They used the keypad scanner built-in inside the BCM20734 with 128KHz clock for reading the buttons. That means it would be extremely hard to spoof button presses for TAS and twitch-plays. Maybe the Pro controller is different, need to buy one though."
Why would a 128KHz polling prevent spoofed button presses? Seems like a NPN transistor could easily spoof that - if TAS players are looking at that. Or am I just missing something how TAS players do it now?
"Also, in a bizarre move, Nintendo didn't use the traditional "one side pulled-up other side to ground" way of reading buttons, instead they used a keypad configuration where buttons are arranged in rows and columns. They used the keypad scanner built-in inside the BCM20734 with 128KHz clock for reading the buttons. That means it would be extremely hard to spoof button presses for TAS and twitch-plays. Maybe the Pro controller is different, need to buy one though."
Why would a 128KHz polling prevent spoofed button presses? Seems like a NPN transistor could easily spoof that - if TAS players are looking at that. Or am I just missing something how TAS players do it now?
I didn't think TAS or Twitch Plays used real hardware anyways.
Console verification is an important part of ensuring TASes are accurate. For that, we need an interface into the console's controls among other things.
Well, given that this controller talks bluetooth, wouldn't a logical way to control the input be to write a BT stack that emulates the controller?
I know its not the legit real controller, but buttons are buttons (analog or digital). As long as you don't mod the console itself, that should be fine for the TAS community, no?
I know its not the legit real controller, but buttons are buttons (analog or digital). As long as you don't mod the console itself, that should be fine for the TAS community, no?
I'm curious about this too. There's also a physical connection between the joycons and tablet which can be used instead of the Bluetooth.
I believe the main issue is that it's very easy to electrically simulate button presses and be confident that the rest of the system is functioning correctly. Once you start simulating the entire controller and its protocol for communicating with the console, you start to enter a realm where it may be possible to trigger inputs that are not physically possible with the actual controller hardware.
I believe the main issue is that it's very easy to electrically simulate button presses and be confident that the rest of the system is functioning correctly. Once you start simulating the entire controller and its protocol for communicating with the console, you start to enter a realm where it may be possible to trigger inputs that are not physically possible with the actual controller hardware.
I'd assume for precise play/speedruns you'd use the controller in the physically connected mode, not over Bluetooth? (with random delays potentially)
They've been doing tool-assisted runs on real hardware for a while now:
http://tasvideos.org/TASBot.html
In addition to TASBot, there's also a growing practice of using physical rigs to "verify" tool-assisted runs:
http://tasvideos.org/ConsoleVerificationGuide.html
This isn't generally regarded as necessary for a TAS to be valid, but it allows it to work on TASBot and similar rigs, which can help it get in front of a wider audience. It also helps figure out which TAS techniques are theoretically possible in non-TAS runs.
http://tasvideos.org/TASBot.html
In addition to TASBot, there's also a growing practice of using physical rigs to "verify" tool-assisted runs:
http://tasvideos.org/ConsoleVerificationGuide.html
This isn't generally regarded as necessary for a TAS to be valid, but it allows it to work on TASBot and similar rigs, which can help it get in front of a wider audience. It also helps figure out which TAS techniques are theoretically possible in non-TAS runs.
the importance of console verification is to make sure a TAS isn't relying on any emulator bugs (such as timing, uninitialized memory, etc) and would theoretically be possible on the actual hardware. it also speaks to the accuracy of the emulations themselves
For a specific example of something like this being important, remember the (very impressive) Watch for Rolling Rocks in 0.5 A presses video from a couple years ago? https://www.youtube.com/watch?v=kpk2tdsPh0A (For the uninitiated, this video presents a TAS of Super Mario 64 which beats the level Watch for Rolling Rocks without pressing the A button outside of keeping it held from entering the level.) It turns out that the route presented in the video actually fails console verification, since the crazy things he does trigger some annoying FPU crashes on console, but not on emulator. There is a happy end, though: a fixed route was published after this was discovered, and it passes console verification just fine (assuming the inputs dont desync over the 13 hour run)
Neat, I've always been into speed running, but my neighbour introduced me to TAS not too long ago.
This is all super cool, thanks everyone.
This is all super cool, thanks everyone.
answer is simple - author doesnt know about transistors, maybe started learning electronics with Arduino or some other end more important than the means tools.
All I'm personally hoping for from the Switch modding community is the ability to back up my saves. It is mind boggling to me that I cannot plug a USB drive into my Switch dock and back up saves. I take this for granted on PC but this should be a day 1 feature on any gaming device.
Doesn't the Switch provide an SD card input? Can you not save data to that?
It does have a microSD, but all game saves go to the internal memory of the console. Even then, the saves are encrypted such that they cannot work on other devices.
Citation: http://en-americas-support.nintendo.com/app/answers/detail/a...
Citation: http://en-americas-support.nintendo.com/app/answers/detail/a...
Nothing in the breakdown explains why the Joycons are so expensive at $80.
The MEMS accelerometer they use costs $1.88 each when buying 500 at a time. Presumably the other chips are about that much as well. There are a lot of connectors on the PCB, but in bulk they, the case, and the battery should be fairly inexpensive.
The entire concept of the Switch as a multiplayer in-couch experience is heavily undermined by the excessive cost of a controller.
The MEMS accelerometer they use costs $1.88 each when buying 500 at a time. Presumably the other chips are about that much as well. There are a lot of connectors on the PCB, but in bulk they, the case, and the battery should be fairly inexpensive.
The entire concept of the Switch as a multiplayer in-couch experience is heavily undermined by the excessive cost of a controller.
I'm not sure if any of this is particularly costly, but the teardown on iFixit[1] says that each controller has a Bluetooth transceiver, LRA, flash memory, NFC reader, and the right side has an IR (depth sensing?) camera.
It's also a lot of small parts that have to be assembled once you add in additional components like the docking rails and physical controls
[1] https://www.ifixit.com/Teardown/Nintendo+Switch+Teardown/782...
It's also a lot of small parts that have to be assembled once you add in additional components like the docking rails and physical controls
[1] https://www.ifixit.com/Teardown/Nintendo+Switch+Teardown/782...
Has there been a videogame console which didn't have overpriced peripherals? As far as I can remember additional controllers were always ridiculously overpriced, even back when they were just a plastic case with a bunch of buttons. Let's not even talk about Sony's price gouging with memory cards (all the way back to the PS1).
I guess it's a way to make a nice profit from what is effectively a captive audience. You can either buy the original peripherals at extortionate prices or you can try your luck with off-brand copies.
I guess it's a way to make a nice profit from what is effectively a captive audience. You can either buy the original peripherals at extortionate prices or you can try your luck with off-brand copies.
> I guess it's a way to make a nice profit from what is effectively a captive audience. You can either buy the original peripherals at extortionate prices or you can try your luck with off-brand copies.
Typically the first generation of controllers are not sold at a large profit. They get cost reduced and a couple iterations in they are major money makers.
> You can either buy the original peripherals at extortionate prices or you can try your luck with off-brand copies.
AFAIK all major consoles have security chips for their controllers, collecting a licensing fee from 3rd party peripheral manufacturers may sound nice, but if the team doing the licensing is the same team who gets credited for profits from first party sales, then there is a huge conflict of interest. This is why the market for third party controllers can be a bit slim for some consoles. It is up to the political whims, and fiscal arrangement, of a company during that console's lifetime.
Typically the first generation of controllers are not sold at a large profit. They get cost reduced and a couple iterations in they are major money makers.
> You can either buy the original peripherals at extortionate prices or you can try your luck with off-brand copies.
AFAIK all major consoles have security chips for their controllers, collecting a licensing fee from 3rd party peripheral manufacturers may sound nice, but if the team doing the licensing is the same team who gets credited for profits from first party sales, then there is a huge conflict of interest. This is why the market for third party controllers can be a bit slim for some consoles. It is up to the political whims, and fiscal arrangement, of a company during that console's lifetime.
Each joycon being 40$ seems pretty reasonable. If one component costs almost 2$ then it's not hard to see how after buying different components, designing the system, assembling the components, packaging, shipping, and everything else adds up pretty fast, especially when you consider that people also want to make a profit on top of all of that
i think the argument made by GP (which i agree with) is that the controllers should be sold basically at the cost of materials in order to encourage the "multiplayer in-couch experience".
i would expect most of the profit comes from licensing fees and royalties, anyway. that's where they'll recouperate the engineering cost.
i would expect most of the profit comes from licensing fees and royalties, anyway. that's where they'll recouperate the engineering cost.
Each joycon is a fully-independent controller, after all!
Wireless Xbox One controllers are at least $40, and about the same complexity.
Or even the licensed minecraft controller, which is identical apart from being green and costing $72: https://www.amazon.com/Xbox-Wireless-Controller-Minecraft-Cr...
Remember, electronics cost has basically nothing to do with BOM cost, other than that the bare minimum for profitability is going to be roughly twice the BOM cost.
Or even the licensed minecraft controller, which is identical apart from being green and costing $72: https://www.amazon.com/Xbox-Wireless-Controller-Minecraft-Cr...
Remember, electronics cost has basically nothing to do with BOM cost, other than that the bare minimum for profitability is going to be roughly twice the BOM cost.
Sure, but that's because Microsoft likely makes virtually no money off the sale of a console. [0] They need to make the cost of advertising and risk taking back somewhere, and accessories is one way they do that.
Also, with regards to the retail markup, big brands generally do not sell directly to department stores, they will rent the space out instead.
[0] https://www.extremetech.com/gaming/251038-theres-no-profit-m...
Also, with regards to the retail markup, big brands generally do not sell directly to department stores, they will rent the space out instead.
[0] https://www.extremetech.com/gaming/251038-theres-no-profit-m...
> Nothing in the breakdown explains why the Joycons are so expensive at $80.
It's because people will pay $80.
This is a common misconception, that the price of a product is calculated using "how much it costs to manufacture, plus some profit margin percentage". That is false. Cost to produce only sets a floor on retail price (and sometimes not even that); apart from that they have nothing to do with each other. You charge what the market will bear.
It's because people will pay $80.
This is a common misconception, that the price of a product is calculated using "how much it costs to manufacture, plus some profit margin percentage". That is false. Cost to produce only sets a floor on retail price (and sometimes not even that); apart from that they have nothing to do with each other. You charge what the market will bear.
The "market" will bear varying levels of cost though. The demand and supply curves are curves and not lines intersecting at 90 degree angles. Often a firm can make more money by lowering price because of the higher quantity.
I believe Nintendo would be better off making little to no money on the Joycons. As a spillover of Switch owners having more Joycons and thus the ability to play with their friends, Nintendo will sell more multiplayer games and more Switches.
I believe Nintendo would be better off making little to no money on the Joycons. As a spillover of Switch owners having more Joycons and thus the ability to play with their friends, Nintendo will sell more multiplayer games and more Switches.
Switches are perpetually sold out. Maybe they could sell more multiplayer games, but I'm betting they can actually charge a premium for the controllers right now because there's a lot of demand to play that can't be met by people getting their own console.
This may be true in some areas but where I live every retailer (gamestop, target, best buy, walmart, fred meyer) is full of available switches, with no bundles attached, selling at MSRP (no additional markup). It's been like that for about a month. I'm sure they will start selling out again as we get closer to the holidays. That said, I'm with the OP in that Nintendo would make more money off me if they sold the controllers for less. So far I bought 3 games (Mario Kart [again!], Street Fighter II [again!], and Super Mario Odyssey), all of which have multiplayer modes, but none of which we have played multiplayer more than once because using the tiny joycons sideways is just terrible, but the cost of an additional controller is too much. It's not that I can't afford it, it's just that the value isn't there. So far, none of the games is good enough that it's worth spending the extra $$ to buy another controller. If the controllers were $40 I would have bought 2 of them by now, probably. I get how silly that sounds.
Many games, especially the ones that let you play with a sideways joycon, are fully-functional with gamecube controllers and an adapter. I know Mario Kart is for sure.
The drawbacks are the missing motion, "minus", and ZL buttons. If the game works without that and you have some gamecube controllers already, getting an adapter could be a good idea.
The drawbacks are the missing motion, "minus", and ZL buttons. If the game works without that and you have some gamecube controllers already, getting an adapter could be a good idea.
Where do you live it took me most of my weekend to find one in a store in Seattle?
Silverdale / Bremerton / Poulsbo area out in Kitsap County.
I believe Nintendo would be better off making little to no money on the Joycons.
What's funny if you abstract this concept out a bit in becomes "I believe the company would be better off making little to no money on their products."
Joking aside, that sounds like quite the gamble to me. And it doesn't appear to me that Nintendo likes that gamble, as they have been pretty averse to taking losses on hardware in the past.
Not to mention I don't know of too many other console makers that are willing to take losses on the accessories.
What's funny if you abstract this concept out a bit in becomes "I believe the company would be better off making little to no money on their products."
Joking aside, that sounds like quite the gamble to me. And it doesn't appear to me that Nintendo likes that gamble, as they have been pretty averse to taking losses on hardware in the past.
Not to mention I don't know of too many other console makers that are willing to take losses on the accessories.
Given that Nintendo wasn't able to keep them in stock for the longest time, lowering the price would have simply left money on the table. And not lowered the cost for consumers--this spring people were buying them and reselling for $120ish.
I bought this for RetroPie use: https://smile.amazon.com/Logitech-940-000110-Gamepad-F310/dp...
It doesn't have wireless, vibration motors, or accelerometers.
This model has wireless and vibration motors, not sure about accelerometers:
https://smile.amazon.com/Logitech-940-000117-Gamepad-F710/dp...
It doesn't have wireless, vibration motors, or accelerometers.
This model has wireless and vibration motors, not sure about accelerometers:
https://smile.amazon.com/Logitech-940-000117-Gamepad-F710/dp...
The wireless F710 does not have accelerometers or gyros, but is otherwise a pretty nice controller if you don't already have a PS or XBox pad lying around.
The default Windows drivers and associated "configuration" program are rather badly out of date though, and to get it working on a Windows 10 system I had to override it to use the XBox gamepad driver instead.
The Linux drivers worked just fine out of the box though.
The default Windows drivers and associated "configuration" program are rather badly out of date though, and to get it working on a Windows 10 system I had to override it to use the XBox gamepad driver instead.
The Linux drivers worked just fine out of the box though.
To get video game consoles on shelves, the manufacturers ask the stores to sell the physical hardware at much lower margins than normal, in return promising sales of peripherals with a huge markup.
Having talked to a few people who were directly involved, for controllers the manufacturer get maybe 10% of the purchase price as profit, the store would take 40%+. The markup was basically 100% from what the distributor charged.
This is also why for the longest time consoles shipped with bare bones everything. E.g. External wifi, sometimes extra video cables were even needed. Anything to help the store turn an actual profit on selling the physical consoles.
Having talked to a few people who were directly involved, for controllers the manufacturer get maybe 10% of the purchase price as profit, the store would take 40%+. The markup was basically 100% from what the distributor charged.
This is also why for the longest time consoles shipped with bare bones everything. E.g. External wifi, sometimes extra video cables were even needed. Anything to help the store turn an actual profit on selling the physical consoles.
Looking at this in the other direction, this is giving them quite a bit of profit now with lots of room to lower the price to remain competitive later. Nintendo is just posturing to undercut the next gen releases from sony and ms.
The HD Rumble on these things is pretty amazing. It makes games like Super Mario Odyssey and Golf Story a real treat to play.
I've yet to differentiate the "HD" rumble from just normal rumble. It just feels like there are more degrees of being "gentle" vs "strong". Is there a hardware difference between what's in the Joycons and the standard counterweight-on-a-motor?
There are several Moons in Odyssey which, in traditional Nintendo fashion, make full use of the new system hardware by vibrating the joycons with increasing intensity as you walk around the area until you feel the strongest vibration, which is where the Moon is located.
Is that so much different than what e.g. even older Xbox 360 / PS3 controllers could do? I remember games like Tomb Raider and several others using similar haptic feedback systems.
If you have 1-2 switch, there's a game in there in which you have to guess the number of balls "in" the joycon. It's a great demo of the HD Rumble feature.
The balls game was the most stunning example of this to me; I really enjoyed the experience of playing it.
In Mario Odyssey, there's a recurring challenge where you use the rumble to find the location of a hidden object. I don't know if it's just a psychological trick based on my perception of movement correlating to stronger vibration, but it genuinely feels like as I move toward the lightly rumbling controller that it not only gets more intense but actually moves under my hand, as if the vibration had moved laterally inside the controller.
This story has a developer explaining:
https://www.polygon.com/2017/5/1/15499328/tumbleseed-hd-rumb...
The gist appears to be that the HD rumble effect allows intensity and speed to be controlled separately. Traditional rumble uses a weighted motor which shakes harder and faster as current increases. I definitely have noticed rumble effects in Mario Odyssey that feel like light, rapid trembles as well as heavy chugging thuds. There's boss fights where the sensation of the rumble actually gives you an idea of how successfully your hits are landing, and it really works. However it works, it's pretty stunning.
This story has a developer explaining:
https://www.polygon.com/2017/5/1/15499328/tumbleseed-hd-rumb...
The gist appears to be that the HD rumble effect allows intensity and speed to be controlled separately. Traditional rumble uses a weighted motor which shakes harder and faster as current increases. I definitely have noticed rumble effects in Mario Odyssey that feel like light, rapid trembles as well as heavy chugging thuds. There's boss fights where the sensation of the rumble actually gives you an idea of how successfully your hits are landing, and it really works. However it works, it's pretty stunning.
The HD rumble is actually precise enough that you can make sounds with it. I'm told that in mario kart 8 deluxe, if you hold up the joycon to your ear as you collect a coin, the joycon itself makes the classic mario coin sound, albeit quietly.
It basically plays audio files. I've played most of my Odyssey playtime with just the joycons by themselves, no strap or grip, which makes it more noticeable. Top 3 rumbles:
1) The "fizz" when you're swimming in a carbonated sea
2) Rumbling motorbikes in the city and elsewhere
3) The scrabbling feet of a "Jaxi"
These aren't the only things with rumble, just the ones that I thought felt most unique. I've also heard that in Mario Kart, the sound of picking up a coin is essentially played by the joycon's rumble, even when the audio is off.
1) The "fizz" when you're swimming in a carbonated sea
2) Rumbling motorbikes in the city and elsewhere
3) The scrabbling feet of a "Jaxi"
These aren't the only things with rumble, just the ones that I thought felt most unique. I've also heard that in Mario Kart, the sound of picking up a coin is essentially played by the joycon's rumble, even when the audio is off.
The difference is comparable to the difference between a traditional vibra Motor in a phone and the Taptic Engine based vibration in iPhone 6s and later, including all the haptic feedback tricks provided by later iOS versions running on these devices
It is hard to explain, but you really do feel it in Golf Story. There is certainly a difference in a "light" rumble when hitting the ball, vs a "light" rumble when the ball is tapping along a green.
Lack of silk screen doesn't have to be an obfuscation measure : silk screen is an extra process step during PCB manufacture and if you don't need it, omitting the step saves money.
Exactly. The test points don't need to be numbered either. The CAD tool they use to design the boards can produce a bed-of-nails pattern that can be fed back into QA testing machines. If something fails, they can pinpoint it through the testing facilities and propagate the failures up the chain. At least in my experience with modern manufacturing, unless you have someone doing manual testing with a spring-nail, labelling the test points is completely optional.
Do you know the term for the QA setup/device to test multiple PCBs at once? Or to e.g. write firmware to more then one more at once? What are some search terms for that?
That's done while the boards are already populated, but not de-panelized yet. All boards in the panel are connected to the JTAG bus, that connection will be broken when you remove them from the panel. JTAG traces run in the "frame" of the panel. So you have a single connector on a panel that you use for flashing firmware and doing boundary testing. Dave Jones @eevblog explains that in one of his videos. I can't remember which one, but I'd start search from this series https://www.youtube.com/watch?v=VXE_dh38HjU
Thanks for the guidance. His design for manufacturing series will be good for me to go through.
e-testing? usually each board/manufacturing house has its own in house custom solution.
btw nice trick for vscored panels to avoid huge number of test points https://www.youtube.com/watch?v=V5BDcEqEaKg
btw nice trick for vscored panels to avoid huge number of test points https://www.youtube.com/watch?v=V5BDcEqEaKg
Yeah I didn't realize that there would be a lot of one-off in-house methods. Besides board/panel level strategies, I was curious physically what kinds of rigs or harnesses are out there, if there were any industry standard setups. But maybe it's doing smart things at the panel level and building custom testing rigs in house.