How to wire a Wii's Bluetooth Module to a PC USB Port

[AnyOldName3]

It's going to be harder to make that than the solder-wires-onto-test-pads approach unless you have the relatively specialist tools that tiny surface-mount components need, in which case it would probably be easier.

You also asked about the big number of components, so I'll give a breakdown of what they do:

• Fuse - stops you blowing up the more expensive components if something goes horribly wrong.
• USB connector - Lets you plug a regular USB Micro-B cable in.
• Switch - is a switch. Works as the sync button like the real console had behind the SD card door, but not totally necessary as there's a button in Dolphin that does the same thing.
• Two capacitors - Lots of linear voltage regulators perform better with a capacitor on their input and output to help cover up small wobbles in the supply voltage. Some of the regulators other people have listed have come with the capacitors, so you wouldn't need separate ones.
• Resistor - Pulls down the switch voltage to zero when the switch is disconnected so it's really obvious when it's been pressed and when it's been released. Without a pull-up or pull-down resistor, a disconnected switch works like a capacitor and floats around the same voltage as it was when the switch was closed and maybe gradually goes towards something else, which can leave it in a kinda-pushed state from a chip's perspective, which can confuse digital circuits. I don't actually think this is necessary as I think the Bluetooth board has one of these internally.
• Tiny Molex plug - lets you plug the Bluetooth board in instead of soldering to test pads.
• Linear regulator - supplies the 3.3V the Bluetooth board needs.
  

[themanuel]

(10-11-2020, 12:31 PM)AnyOldName3 Wrote: It's going to be harder to make that than the solder-wires-onto-test-pads approach unless you have the relatively specialist tools that tiny surface-mount components need, in which case it would probably be easier.

You also asked about the big number of components, so I'll give a breakdown of what they do:

• Fuse - stops you blowing up the more expensive components if something goes horribly wrong.
• USB connector - Lets you plug a regular USB Micro-B cable in.
• Switch - is a switch. Works as the sync button like the real console had behind the SD card door, but not totally necessary as there's a button in Dolphin that does the same thing.
• Two capacitors - Lots of linear voltage regulators perform better with a capacitor on their input and output to help cover up small wobbles in the supply voltage. Some of the regulators other people have listed have come with the capacitors, so you wouldn't need separate ones.
• Resistor - Pulls down the switch voltage to zero when the switch is disconnected so it's really obvious when it's been pressed and when it's been released. Without a pull-up or pull-down resistor, a disconnected switch works like a capacitor and floats around the same voltage as it was when the switch was closed and maybe gradually goes towards something else, which can leave it in a kinda-pushed state from a chip's perspective, which can confuse digital circuits. I don't actually think this is necessary as I think the Bluetooth board has one of these internally.
• Tiny Molex plug - lets you plug the Bluetooth board in instead of soldering to test pads.
• Linear regulator - supplies the 3.3V the Bluetooth board needs.
  

Thanks for the detailed explanation.

[DacoTaco]

(10-11-2020, 08:25 AM)LordPimpus Wrote: Hi all,

I created a new version of the Wii Bluetooth to USB adapter, which I have built and tested to confirm it works.  The main differences from DacoTaco's version is my board is smaller, lacks pads to solder directly to test points on the Wii bluetooth module, and has an added pulldown resistor for the sync button.  I also spec'd a much smaller linear regulator, as the one on DacoTaco's version is overkill.

For anyone who wants to make their own, I have posted the PCB project at Oshpark: (https://oshpark.com/shared_projects/vOrhO2HN).  The Oshpark project description includes a link to a Digikey BOM for the required components.  Note that Oshpark requires board orders in sets of 3, but the Digikey BOM only includes parts for 1 set; if you want three complete adapters, you will need to triple the BOM quantities.

For prices current as of this writing, total cost for parts for 3 complete adapter boards (PCBs plus electronics) is ~$11.  This does not include the Wii Bluetooth module, which you have to source yourself (I ordered mine on ebay).  I tested only with the J27H002 module, but either that or the WML-C43 should work fine.

There are some fine pitch components, so I suggest pliers, microscope, and a reflow oven for assembly.

a pulldown on the button isn't a bad idea tbh, but doesn't the module have it internally? i suppose not, but i didn't see any issues with it when i was playing around with it.
ill add it to the design as optional.

as for the regulator , it might be overkill, but i use it for all my 3.3v needs and is very reliable , i went with what i know

(10-11-2020, 11:57 AM)themanuel Wrote: Also, why does the BOM list so many components, when the basic setup works with just the BT adapter and a voltage regulator?

this is why i kept my design big and kept the design open for any way you want to use or make it.
it has options however you like it.
want to go for the original design? just solder on the test pads, attach a usb cable to the usb points, add the voltage regulator and bridge the fuse pads. done.
want to go for a better design? use all the optional stuff (button, regulator, micro usb, capacitors, socket,... )

[Deam]

Hey everyone.
This thread has been very informative - to the point that I recently just soldered my second Wii module. I do have a quick question regarding people's experience. Basically, is the Wii bluetooth module very sensitive to EMI/interference? The reason I ask is this:

- In my living setup, the computer is actually behind a 65" TV. I made the first adapter using the first gen Wii bluetooth module. It worked fairly well, but Wii remote audio was poor and syncing sometimes didn't work so well. Having gotten frustrated, and also thinking that I used too long of a cable (nearly 3ft) in an attempt to have the module/box in front of the TV. I have a lot of stuff on the same TV stand, the TV, Computer, PS4, receiver large centre speaker, Harmony Hub, etc. Lots of radio devices.

- Today, I made a second module with a much shorter cable, though that causes it, at best, to dangle just below the TV.

While syncing, wii motion plus, etc. is all fine, Wii remote speaker audio is still inconsistent. If I am line of site and close, it works very consistently. But even if I move a few feet back, it starts to cut out or get crackly.

Is there some way to add an Antenna wire to to the module? Unfortunately, using a USB extension cable just reintroduces the original issues, in that the audio quality diminishes. d

The fact that it works when close and line of sight suggests the wiring is fine, and I just have a lot of interference in that area. I just don't ever recall the Wii being this sensitive to interference or range issues.

[LordPimpus]

(10-11-2020, 12:31 PM)AnyOldName3 Wrote: It's going to be harder to make that than the solder-wires-onto-test-pads approach unless you have the relatively specialist tools that tiny surface-mount components need, in which case it would probably be easier.

You also asked about the big number of components, so I'll give a breakdown of what they do:

• Fuse - stops you blowing up the more expensive components if something goes horribly wrong.
• USB connector - Lets you plug a regular USB Micro-B cable in.
• Switch - is a switch. Works as the sync button like the real console had behind the SD card door, but not totally necessary as there's a button in Dolphin that does the same thing.
• Two capacitors - Lots of linear voltage regulators perform better with a capacitor on their input and output to help cover up small wobbles in the supply voltage. Some of the regulators other people have listed have come with the capacitors, so you wouldn't need separate ones.
• Resistor - Pulls down the switch voltage to zero when the switch is disconnected so it's really obvious when it's been pressed and when it's been released. Without a pull-up or pull-down resistor, a disconnected switch works like a capacitor and floats around the same voltage as it was when the switch was closed and maybe gradually goes towards something else, which can leave it in a kinda-pushed state from a chip's perspective, which can confuse digital circuits. I don't actually think this is necessary as I think the Bluetooth board has one of these internally.
• Tiny Molex plug - lets you plug the Bluetooth board in instead of soldering to test pads.
• Linear regulator - supplies the 3.3V the Bluetooth board needs.
  

These per-component notes are pretty spot on, though I will clarify the capacitors aren't optional; they are required to keep the regulator from going into uncontrolled oscillations (refer to the datasheet for the part for specifics). Many/most regulators require having external filter caps on the input and output pins for stability. It is uncommon to not require such external components.

As for the other parts, the fuse is optional, but you would have to instead short the pads together. The pull-down resistor probably isn't required, but since I could not locate a schematic for the bluetooth module to confirm whether or not it had internal pull downs, I added one. It's bad design to leave inputs floating, as you don't know what they are doing without that explicit input. If you don't want the pull-down resistor, save $0.10 and don't populate that part. In any event, having an extra pull down won't hurt anything. Likewise, the switch isn't required, since as I explicitly stated on the Oshpark project page, you can configure a hotkey in Dolphin to do the same thing.

As for soldering difficulty, we'll have to agree to disagree; the pads for the SMT parts aren't much smaller than the test pads on the bluetooth module, and the SMT pads will actually be easier to solder (rather than trying to solder fiddly lengths of wire to the Bluetooth module test points). The mechanical connections will also be far more reliable, and you can't accidentally delaminate the traces on the Bluetooth module (since you are won't be soldering anything at all on the module).

Either way, you'll likely get frustrated trying to solder either DacaTaco's or my design with a junk radio-shack soldering iron. Solder paste and a reflow oven is the most painless way of doing assembly, but a decent soldering iron will also do. But hey, assemble however and with whichever design you like. The suffering you inflict on yourself is of no concern to me. :p Worst case scenario, if you try my design you won't to break the bank for $11 worth of parts, and you'll get three tries. You can also use beer/liquor/whatever to bribe a friend with good equipment to do the assembly for you.

Also, in reply to another post from DacoTaco: there is nothing wrong with the linear regulator you used in your design, but that can handle like 800mA, when the Wii module only needs ~30mA. The replacement regulator (including required caps) is like $0.30 total, and can still handle ~300mA. In my design, I was aiming for smallest PCB and cheapest to manufacture (it seemed a bit odd to me that the other gent on the forum was charging $50 each for what is ~$4 worth of parts, especially since you did that design work).

One other thing I didn't mention before is that your design has the RF antenna on the Bluetooth module partially occluded by the adapter PCB, which can cause substantial signal loss (10+ dB loss in Tx/Rx power). My design has the antenna unobstructed by the adapter PCB, which should improve functional range. In all other respects, I regard yours as solid design work.

Lastly for anyone who is interested, the attached image shows the final product. From left to right: US quarter for scale, Wii Bluetooth module, my adapter with Wii module attached (top view), same (bottom view).

[AnyOldName3]

The regulator and especially the mini Molex plug look way too small for a soldering iron to me. Maybe I've never used a good one, but I wouldn't be confident about getting the legs connected to the board without shorting any out. Soldering to the test pads, though, was only hard because I had annoyingly thick wire that didn't want to bend into the right shape, and if not for going over things to try and make them neat and pretty (instead of just declaring victory once I'd got something functional) I wouldn't have delaminated any traces (I'd bought two boards as I'd anticipated something going wrong and didn't want to wait another month for a spare if it did). I'm also fairly sure the mechanical connections of my solder joints are more secure than the Molex plug being a plug. None of this is saying your design is bad, just that I'm surprised by your confidence as someone who usually unsolders surface mount components I'd like to salvage with a dremel after learning that anything else usually kills the part.

Anyhow, even if your board had been available, I wouldn't have been able to use it as I had very limited space as I was trying to fit the whole thing into a sensor bar I'd already converted to USB power.

[DacoTaco]

a lot to go over here, but let me go through it one by one

(10-12-2020, 12:06 PM)LordPimpus Wrote: These per-component notes are pretty spot on, though I will clarify the capacitors aren't optional; they are required to keep the regulator from going into uncontrolled oscillations (refer to the datasheet for the part for specifics).  Many/most regulators require having external filter caps on the input and output pins for stability.  It is uncommon to not require such external components.

i started doing this after finding out power became very unstable when i plugged a gameboy cart into my reader. since then ive been using the same regulator/cap combo to stabilize them and have something i trust

(10-12-2020, 12:06 PM)LordPimpus Wrote: As for soldering difficulty, we'll have to agree to disagree; the pads for the SMT parts aren't much smaller than the test pads on the bluetooth module, and the SMT pads will actually be easier to solder (rather than trying to solder fiddly lengths of wire to the Bluetooth module test points).  The mechanical connections will also be far more reliable, and you can't accidentally delaminate the traces on the Bluetooth module (since you are won't be soldering anything at all on the module).

i too will disagree with you here
in general the usb port and molex connectors are the hardest to do with a soldering iron.
i know i can do it with my 30€ , cheap of the mill iron but only because i have a few tips for the pins (flat head) and because i have experience with them. still, i won't be pretty.
the rest is easy and should be do-able for anyone, specially on my design because i left so much space between all components and (specially) chose 1206 package smd components.
you went smaller and those are harder to do and not get them to stick to the iron.
again, its very do-able, but you need the experience.

(10-12-2020, 12:06 PM)LordPimpus Wrote: Either way, you'll likely get frustrated trying to solder either DacaTaco's or my design with a junk radio-shack soldering iron.  Solder paste and a reflow oven is the most painless way of doing assembly, but a decent soldering iron will also do.  But hey, assemble however and with whichever design you like.  The suffering you inflict on yourself is of no concern to me.  :p   Worst case scenario, if you try my design you won't to break the bank for $11 worth of parts, and you'll get three tries.  You can also use beer/liquor/whatever to bribe a friend with good equipment to do the assembly for you.

oiy, my design (v2) can be done with an iron just fine. align pads with side of pcb and solder. done.
(but i do like your design more cause its smaller and more compact. my design was meant for average diy-er)
i do like the bribe idea..
anyone want to me a few beers? xD
(i only accept good belgian beers. no jupiler bullshit)

(10-12-2020, 12:06 PM)LordPimpus Wrote: Also, in reply to another post from DacoTaco: there is nothing wrong with the linear regulator you used in your design, but that can handle like 800mA, when the Wii module only needs ~30mA.  The replacement regulator (including required caps) is like $0.30 total, and can still handle ~300mA.  In my design, I was aiming for smallest PCB and cheapest to manufacture (it seemed a bit odd to me that the other gent on the forum was charging $50 each for what is ~$4 worth of parts, especially since you did that design work).

thing is, most of that price is the pcb order and the wii modules can be between 5 & 15 bucks a piece. its a fully assembled kit + a little bit of profit (its really not much).
diy is obviously cheaper, but not everyone can do it.

(10-12-2020, 12:06 PM)LordPimpus Wrote: One other thing I didn't mention before is that your design has the RF antenna on the Bluetooth module partially occluded by the adapter PCB, which can cause substantial signal loss (10+ dB loss in Tx/Rx power).  My design has the antenna unobstructed by the adapter PCB, which should improve functional range.  In all other respects, I regard yours as solid design work.

ah, ill take that into mind next time i design a wireless pcb (if i ever will). ill change v2 a bit and move the molex connector to the left, so the antenna hangs over the gap. that should do it right?
and thanks for the compliment. as easy peasy as the pcb is, its still good to hear that since its like... the 2nd pcb i designed xD

(10-12-2020, 12:06 PM)LordPimpus Wrote: Lastly for anyone who is interested, the attached image shows the final product.  From left to right: US quarter for scale, Wii Bluetooth module, my adapter with Wii module attached (top view), same (bottom view).

very sexy.

[LordPimpus]

(10-13-2020, 03:57 AM)AnyOldName3 Wrote: The regulator and especially the mini Molex plug look way too small for a soldering iron to me. Maybe I've never used a good one, but I wouldn't be confident about getting the legs connected to the board without shorting any out. Soldering to the test pads, though, was only hard because I had annoyingly thick wire that didn't want to bend into the right shape, and if not for going over things to try and make them neat and pretty (instead of just declaring victory once I'd got something functional) I wouldn't have delaminated any traces (I'd bought two boards as I'd anticipated something going wrong and didn't want to wait another month for a spare if it did). I'm also fairly sure the mechanical connections of my solder joints are more secure than the Molex plug being a plug. None of this is saying your design is bad, just that I'm surprised by your confidence as someone who usually unsolders surface mount components I'd like to salvage with a dremel after learning that anything else usually kills the part.

Anyhow, even if your board had been available, I wouldn't have been able to use it as I had very limited space as I was trying to fit the whole thing into a sensor bar I'd already converted to USB power.

It is good you got acceptable results.  My point was more that in my design, an assembler won't have to risk any damage to the Bluetooth module (which is the hardest part to source).  Also, while the wire-to-pad solder joints might be mechanically fine, small circular test pads like those on the bluetooth module don't have a very high peel strength; a single accidental tug could delaminate the individual pads fairly easily, and those aren't easy to repair.  A similar tug is far less likely to cause the molex connector to fail, since those stresses are distributed throughout all the pins.

As to removing SMT components, it sounds like you have not experienced the joy of a decent rework station.  With one, you could remove just about any SMT part effortlessly, without damage to either the part or the work piece.  I use a work station that has both IR heater and hot-air pen, and once hot, parts just lift away using a pair of tweezers.  You should get decent results with even just a cheapo $50 hot-air pen.

Regarding my confidence, this stuff is part of what I do professionally.  I have two engineering degrees and have been doing this stuff for ~20 years, so yeah, I'm quite confident I know what I am doing and am pretty familiar with the design tradeoffs. 

[LordPimpus]

(10-13-2020, 04:40 AM)DacoTaco Wrote: thing is, most of that price is the pcb order and the wii modules can be between 5 & 15 bucks a piece. its a fully assembled kit + a little bit of profit (its really not much).
diy is obviously cheaper, but not everyone can do it.

ah, ill take that into mind next time i design a wireless pcb (if i ever will). ill change v2 a bit and move the molex connector to the left, so the antenna hangs over the gap. that should do it right?
and thanks for the compliment. as easy peasy as the pcb is, its still good to hear that since its like... the 2nd pcb i designed xD

very sexy.

I am a surprised you say most of the cost is the PCB; the price from the board house I linked was $0.80 per board for small orders (for my design; they charge by square unit, so your design would be a bit, but not much, more), and only $0.48 per board for bulk orders.  That is not even a particularly good price, but that board house works great for small quantities, particularly since there is no extra charge for domestic shipping (US).  If your fab house is charging more than a $1 per PCB, you are getting gouged.  There are a thousand chinese PCB fab houses that will do runs far cheaper.  The tradeoff is longer turn and ship times.  However, if your fab house is also doing assembly for you, most of what you pay for is labor.

I have a similar price comment for the Wii modules.  I paid about ~$5 each on ebay, but there are tons of listing on e.g. aliexpress for $1-2 each.  The catch is needing to do a bulk order (10-20 units min is typical) and a long lead time (shipping is usually quoted 1-2 months).

Regarding the antenna, the less obstructed the antenna is, the better.  Nearby ground planes are bad, and even the dielectric of nearby bare PCB (i.e. PCB without ground plane) will attenuate the signal.  If you ever modify your design, I suggest moving the molex connector adjacent to the board edge so the wii module protrudes as far as possible over the edge, and removing the nearby ground plane.

The pads for the test points present some tradeoffs, since to maximize signal you would need to rotate those 180 degrees to have the wii module protrude over the edge in that configuration, but then being soldered in that position is just begging to have those traces snapped off the first time it is struck at an odd angle.  But I recommend against soldering to those pads at all, and just using the connector.

I also suggest moving the through-holes for the USB pigtail to the end of the board opposite the bluetooth module, the fast switching of the digital lines of the USB port near the RF antenna can interfere with the Bluetooth RF signalling.

Here is to you getting your beer bribes.

[AnyOldName3]

It was a while ago, but I think I got two modules for $3 including shipping from aliexpress, so risking damaging one wasn't something I considered a major problem. Also, looking at the photos I posted here https://forums.dolphin-emu.org/Thread-how-to-wire-a-wii-s-bluetooth-module-to-a-pc-usb-port?pid=493560#pid493560, apparently I set the solder joints in a mixture of superglue and baking powder to stop the pads lifting off, so they should be rock solid.