Wireless adapter for Amiga joysticks

[Solo761]

I've been experimenting a bit and came up with wireless adapter for joysticks.

Motivator behind the idea was having Amiga connected to TV and sitting on couch that's 3-4m from TV. Obviously, joystick cable is not long enough so why wouldn't this communication go wireless?

And the result was this, two boards, receiver and transmitter. One plugs to Amiga and joystick plugs into other one. Only problem is power for joystick part. At the moment it comes from mobile phone charger (in video it's connected to monitor's USB port, same thing). First thought was to use mobile cell phone power bank , the ones you can use to charge cell phone on the go. But the one I got turns out to be "smart", if it decides nothing's using it it turns off after a minute. And transmitter uses too little power so battery decides nothing is using it and turns off . I have to get some stupider one . But the concept works . (this was my first time using SMD components, it shows, and it's kinda obvius which one I soldered first )





Here's video of Lotus 2 in action. Range should be up to few hundred meters, more than enough .

[ Show youtube player ]


I used ATMega8 microcontroller to crunch data, NRF24L01+ wireless receiver/transmitter and Arduino as IDE since I have Arduino Uno already. But since it fits in ATMega8 and they're cheaper final boards use them instead of ATMega328 that's the core of Uno board. They (8 and 328) are pin compatible either one can be used with this schematic. Only firmware needs to be recompiled for ATMega328 (UNO) as ones in my archive are for ATMega8.


Unfortunately mice don't work as they should. I guess it's because they send more data than joysticks so while it's transmitting one packet it looses few others that are generated in the mean time so movement is jerky. If I moved it slowly it worked fine, but normal usage is no-no. Maybe some kind of buffer or continuous transmission would help but first I need to figure out how to implement something like that .

I haven't noticed any lag, according to Arduino tests with NRF24L01 it takes 2-3 ms for test data to get sent to reciever and then returned to transmitter. And that's with bigger packets than this (I send only one byte at a time) and don't return it to sender. So I'd guess it's under 2 ms.

*Update*

I finally got back to it and managed to speed up communication. Mice now work fine with v1.3. Tested with old tank mouse and with PS/2 to Amiga adapter, both worked fine over wireless.


Here are all arduino files and schematics

v1.3
https://www.dropbox.com/s/9kiyzbckgn...0v1.3.rar?dl=0

Old v1.0
https://www.dropbox.com/s/6l89sulbs6...0v1.0.rar?dl=0

[lesta_smsc]

How much power does it need? I'm assuming AA/PP3 batteries should be enough if it is running off 5V?

By the way, your Lotus video looks like its running on Turbo or maybe it's NTSC lol!

Great project and concept... It would be nice if there was a way to configure PC wireless controllers to use on Amiga. I suppose it would only require one board which connects to the Amiga and draws the necessary power from it whilst the board configures the wireless connection with the controller accordingly.

[Solo761]

It needs 5V so AA/PP3 batteries won't help, AA are 1.5 so you can get either 4.5V or 6V by combining them, and PP3 are 9V which is too much. This could be fixed by adding voltage regulator to lower the voltage from 9V to 5V or step up converter to bring voltage to 5V, but for prototype this seemed too much.

Another problem with PP3 batteries is that they have small capacity, they're in range of 300-400 mAh and you'd loose some more because 4V out of those 9V would be converted to heat by voltage regulator. Rechargable AA batteries go up to 2000+ mAh.
For this prototype I decided to go with USB connectors so you can power it via cell phone charger.

I planned to use power bank type battery, but it's "smart" so it turns itself off to save power if nothing is using it. And as I mentioned transmitter uses so little power it decided nothing was there . I ordered few more cheaper, and hopefully stupider, ones .

Step up converter looks to be the best solution if one wants to have it all in one box, these power banks use 18605 li-ion batteries which are 3.2V, and they output 5V, so they already use step up converters. Maybe even cannibalizing them for parts is an option since you can get them for less than $2 without batteries...

Strange, I haven't noticed it goes faster. Both amiga and game should be PAL O_o.

That's how they say, another bag of worms. I doubt there's standardized protocol for wireless controllers. Every manufacturer uses their own, as well as components, so I doubt this WiFi module could be used for that. I.e. XBox 360 controllers use proprietary protocol, that's why there aren't that many knockoffs (AFAIK they started to emerge only few months ago), similar with standard PC USB wireless gamepads, it depends on manufacturer, they all work at 2.4 GHz range, but protocol and encoding are proprietary.
Only potential solution would be using PS3 controller, they use bluetooth for communication which is known and open, and you can get them for as low as $10 from China (knockoffs of course). Of course, different WiFi module would be needed, I'll have to check how much is bluetooth module for arduino.

[Jope]

http://www.adafruit.com/products/1944

[Solo761]

Or something like that , although it costs almost as all the parts in these two boards, plus the price of the battery.

But first thing first, I need to be happy by the way it works before I'll start thinking about battery powered work . With something like that I wouldn't need AMS1117-3.3 regulator as NRF24L01 could get it's ~3.3V directly from the battery...

[Solo761]

I finally got to update this, now it works fine with mice, tested with old tank mouse and with optical one via ps/2 to amiga adapter.

Next step is to simplify PCB and to decide if to remain at current USB 5V power supply for transmitter, or to switch to some other power source. Since I got it to work reliably at 8MHz 5V is not required. ATmega8A needs at least 2.7V to work at 8MHz so if I use some battery at, let's say, 3.7V it could supply both MCU and WiFi part (which needs 1.9 - 3.6V). Easiest would be to use 3x AAA or 3x AA rechargeable batteries, that would make 3.6V since one cell is 1.2V, but I'm afraid what would happen if standard batteries were to be used, they're 1.5V so that would make 4.5V and it would fry WiFi part...

[GhstWlf]

Here is the "old" mintyboost project: https://learn.adafruit.com/minty-boost/download
It is a diy 3v (2*AA) to 5v thing.

[Bastich]

Very cool. I was thinking of doing something similar with some of my Arduino bits (so cheap from China)

[Solo761]

Originally Posted by GhstWlf

Here is the "old" mintyboost project: https://learn.adafruit.com/minty-boost/download
It is a diy 3v (2*AA) to 5v thing.

Thanks, it looks interesting, that could solve power supply issue

Originally Posted by Bastich

Very cool. I was thinking of doing something similar with some of my Arduino bits (so cheap from China)

Feel free to improve or tinker with it , I think I've included everything in rar archives, from schematic to source files

[hooverphonique]

How about running the transmitter on 3.3V only? The AVR also runs on 3.3V..
Then you could add a small boost converter and have it run off 1 or 2 AAA batteries..

[Solo761]

I'll have to see what's the cheapest solution, those converters from GhstWlf's link can be set to output 3.3V or 5V.

After that, it's only a matter if one battery can power it long enough or it's better to use two. As far as regular joystick goes they're not power hungry. I have USB voltage/amperage/wattage meter that goes to three decimals and it rarely shows 0.001A drain, mostly it's at 0.000, too low to register. Only if I use PS/2 mouse via PS/2 to Amiga adapter, it goes to 0.009A.

[hooverphonique]

Originally Posted by Solo761

Only if I use PS/2 mouse via PS/2 to Amiga adapter, it goes to 0.009A.

well, there's a point, because most of those are designed for 5V supply, which means you need 5V at the transmitter, anyway...

[MrFluffy]

Liking this, thinking of building one but wondering if I can shrink the pcb a bit so it can be fitted underneath a joystick in the hand somehow, maybe 3d print a housing which matches the joystick. Or I'll be brave and see if theres any space inside my kempston to squeeze it inside while I replace the microswitches :-)

For power 5v buck regulator, have you seen these :-
http://www.ebay.co.uk/itm/1-x-PCB-Li...-/390845137551

Also wondering if I should get one of these to hand off charging a lipo inside the joystick and give it a "android" port to charge with between uses, save any nasty accidents...
http://www.ebay.co.uk/itm/Micro-USB-...-/291355449287

[Solo761]

It definitely can be shrunken down. I made it like this just to make it easier to etch at home. It can be made doublesided and that would definitely shrunk it down considerably.

Also, crystal and it's two capacitors can be thrown out. I left it in this first design because I didn't know if it would need to work at 16 MHz or it could work at 8 MHz. It works fine at 8 MHz and for that we don't need external crystal, we can use ATMega's internal one.
If you want to make it internal to joystick you can also skip DB9 connector which is chunkiest part of it all, same with power connectors on transmitter board, I made three different ones to make my life easier with this prototype, but in the end I used only micro USB one, android port, as you said. Actually, I use mobile phone charger to power it .
That all depends on how you want to use it, my idea was to have it externally so I can plug whatever in it.

There are many options for powering it, unfortunately . It would be easier if there weren't so many. GhstWlf option is OK, it can use standard AA batteries, and it's only few components. Only problem is that it doesn't charge them.

At first look these boards looked fine, but on second look I'm not so sure :/. First one is step up charger, it doesn't have any "smarts" so battery has to have it's own overcharge protection. But that's not a problem. Charger part is, from what I can tell it from schematics (as they "forgot" to mention it in description) it needs 3.7V to charge battery so you couldn't use mobile phone charger for it. Output is fine, it steps up batteries 3.7V to 5V so no problem there, only charging would be a problem.

Second board is vice versa . It takes 5V and steps it down to charge 3.7V battery. Problem is it also outputs 3.7V and not 5V. At first I thought it has separate output, but closer look at images shows that B+ and OUT+ are connected. So that's just to make it easier to solder. Here's a similar board, only difference is that it has only one set of 3.7V contacts, but it's cheaper. They both have charge detection.
Now, depending on usage this might be OK. If you're only going to use it for joystick, and this joystick doesn't have any electronics on it's own (like autofire function) it would work. These joysticks are kinda stupid , commands are detected only by if there's contact or not, like famed Competition Pro, it only has switches that make contact, no IC's or whatever. In this case 3.7V or 5V doesn't matter, it would work. And both ATMega and WiFi module can work at 3.7V (what's more, WiFi module requires it, it would fry at 5V, that's why I have voltage regulator to lower 5V to 3.3V). In this case voltage regulator too can be thrown out and board made even smaller. Downside is what I mentioned, it (transmitter) could only be used for joysticks without autofire function. Autofire needs some IC and it works at 5V, since it gets 5V from joystick port. Unless the chip used in that particular joystick has wider range of input voltage, in that case it would work fine, but I don't know how much, if any, ICs back then used lower than 5V. Also, mice definitely need 5V

Another potential problems are batteries. They tend to be expensive and if you want to install it inside your joystick you'd need those flat ones, like for cell phones or tablets. I don't know if these boards (ones with micro USB connector) can be used with more available (and cheaper) 16850 li-ion batteries, or with RC model batteries like these

http://www.ebay.com/itm/231581370655
http://www.ebay.com/itm/371324908015
http://www.ebay.com/itm/251990085772

If they could that would simplify all of this and make it cheaper .


*edit*

Blah... I should have searched more before writing a sermon like this . Here's the board that does all this. It charges 3.7V battery with 5V and it outputs 5V, plus it has charging protection. This one would do nicely. I'll probably order few to test it.

http://www.ebay.com/itm/231276963178

[MrFluffy]

I'm looking at that last board, and a nintendo ds battery... Of which I can steal the old ones out the kid's ds's now they have 3ds xl's instead :-)
For the pcb, I can pcb mill double sided pcb's in theory although I've never done one yet so might be a interesting project to test it out on.
Also this to charge the stick might be interesting if it works with the above board.. Then you wouldn't need any sockets on the outside of the joystick, no charging connectors to get broken etc...
http://www.ebay.co.uk/itm/UK-STOCK-Q...item20f461aef1


A plan is forming...

[Solo761]

I guess NDS battery could be used. You'd only need to solder wires to it's contacts. At least on NDSL battery contacts are soldered to the motherboard itself so you don't have shell with battery contacts you can repurpose .
Soldering directly to battery could be dangerous because of heat. Maybe solution could be to glue wires to contacts with conductive glue.

Well, idea looks fine, if you have wireless charger, if not...

[Solo761]

Finally, voltage regulator slash battery charger slash voltage up steppers have arrived

In the end I ordered these http://www.ebay.com/itm/231276963178 they're really small. Not just to decide on battery and case. I just ordered this battery

http://www.banggood.com/Wholesale-Re...s-p-11598.html

I think this will do nicely .

[Daedalus]

Nice work! I've been working on a similar project for a while now, but it's been neglected for recent months. I was working on (Amiga) keyboard support as well as joystick and mouse. As for the battery problem, I was using a nice little chip that gave a guaranteed 5V output and reasonable efficiency from a wide range of input voltages but doesn't support charging. Was thinking of running it off 2 AA batteries instead if I can get the power consumption low enough, otherwise 4 AA batteries should give a reasonable life...

[Solo761]

I finally have all the parts so now just to redesign PCB and find some case to install it into (not necessarily in that order ).

I don't know what components you use, but if it's like in my case (ATMega and NRF24L01+) consumption is low. I first tried to power it with powerbank via USB. But current draw was so low that it went to standby instead of powering it. I tried USB power meter on it and it's usage is 5-9 mA, that's around 45 mW.

[mfilos]

Nice work man! Can't wait to see the new design