OLED/LCD Arcade Buttons

[stigzler]

I'm sure this has been much discussed - but it feels to be getting closer as time goes by. I'd love to make up a prototype and wondering if anyone's managed this? I guess the ideal would be to have one fit into a standard slightly modded transparent arcade button so that the image on the button could represent the move/button type etc.

It's looking like micro-oled's are the way to go - some examples:

http://www.oled-info.com/emagin-announces-new-brighter-svga-oled-microdisplay-digital-interface
http://www.aliexpress.com/item/0-5-inch-amoled-display-google-glasses-display/844978829.html
http://www.emagin.com/oled-microdisplays/

It's looking like the full colour ones are still in early dev + are incredibly expensive. Not sure you can get LCDs small enough to fit inside a standard button - most appear to be around 1.5". Also - there are a few round lcds available, but again, seem too big.

However - there do appear to be some small monocrome OLED which may do the job - increadibly cheap too! For e.g.:
http://www.aliexpress.com/item/0-5-inch-OLED-module-small-lcd-shield/1990269450.html
http://proto-pic.co.uk/micro-oled-breakout/?gclid=CjwKEAjwh6SsBRCYrKHF7J3NjicSJACUxAh7VM3XdYXhEQpY1V-aeVN_EF92cEmtZgPFVvvUtDzeVhoCjfTw_wcB
https://www.sparkfun.com/products/13003

The middle one appears translucent which would also be good for preserving RGB LED lit buttons? Bottom one shows it in action - measures .6" across so wondering whether this would fit in a button?

Think one of the big problems is going to be extending those thin ribbon cables that run from the screen to the driver board - I'm wondering whether there are any standards for these cables/strips and if so whether it's possible to buy extenders....

Lastly, from watching the vid on the last link - would you need 1 Arduino board per screen - or could one board run several. Looks like there are different ways to run the display (thinking it may be 'serial' vs 'paralell?) and if serial wonder whether one board could run several displays?

Just spitballing, talking about stuff I have very little knowledge about! Wondering if anyone else interested in this or indeed has any ideas/successes in this area?

[SavannahLion]

I had the same thought years ago. There was an OLED keyboard called the Optimus that was released. As soon as it was announced, this was what kept bumbling around in my brain. Sadly, it's pretty expensive, a quick Google shows still retailing for around $1500.

Anyways, the point is, the technology is certainly possible and I imagine the Optimus would be used as a reference Unfortunately, if there is a patent that might preclude someone like Andy or Randy from coming out with an arcade product.

Interesting thing to note and something to keep in mind. The screens on the Optimus are actually fixed and the keyboard buttons move around the screen. Not unlike how an Arcade button might function. However the caveat is that there is a small gap between screen and key to clear the movement. That might be a sticking point if the same technique is applied. Just a thought.

[stigzler]

Yeah. Remember seeing that. Sad to hear price still so high.

You did make me wonder about a different idea though. Dunno if sci fi of actually possible. What if you had an lcd l a yed flat at the bottom interior of the control panel. Then youd transfer portions of the screen via fibre optic light pipes to the button faces? Lol. Feels like total fantasy cos dont even know if this possible.

EDit: oh. I wasnt imagining it! Like this:
http://video.mit.edu/watch/fiber-optic-bundle-11499/

[RandyT]

There are already buttons (although not in the traditional arcade sense) available which have active video displays inside.  HAPP sells them for $75 each, and that doesn't include the hardware to actually drive the displays.

The value to cost ratio is abysmal for something like this.  Unless it is a very specialized application, where considerable income can be generated by the machine (i.e. gambling), it's probably impractical.

[stigzler]

Hmmmm... yer right $100 per button is gonna be silly.

http://na.suzohapp.com/all_catalogs/pushbuttons/130-08019-01

Besides that, it's square and wouldn't fit into a standard button...

So, the Sparkfun micro OLEDs are $15 a piece. Run on a arduino - if you could get one arduino driving a number of these - that'd probably work out quite cheap.

If only I knew the first thing about arduinos, oled driving methods ("OLED over an SPI or I2C interface") or where to source tiny ribbon cable extensions, this might be a go-er...!

[JimmyU]

Definitely not cost effective right now, but give it 5 or so years and it might be. Imagine a control panel with buttons that can say "High Punch" and then "Fire" when you switch games in MAME. I think it would pretty kick ass.

[yotsuya]

When I see how much money dudes drop on 4 Player Panels with 8 buttons each that all light up, I'm sure they'll drop the same amount on OLED buttons without batting an eyelash.

[lilshawn]

just a quick example....


http://ca.mouser.com/ProductDetail/NKK-Switches/IS15ABFP4RGB/?qs=sGAEpiMZZMtqO%252bWUGLBzeL%252bfTmcilAg8

http://ca.mouser.com/ProductDetail/NKK-Switches/IS15DSBFP4RGB/?qs=sGAEpiMZZMsmUjwTDAf1t5IDP72lP7Lh



until such time as it becomes ridiculously cheap to make these it will continue to be prohibitivly costly. I think even that company that had made a keyboard all with LCD modules in the keys retailed for 10 grand. and they probably bought quantities in the 1000's to get a "deal".

[lamprey]

Not quite the same as the button itself changing, but..

When I was trying to decide how I wanted to do my "virtual" cab, I played with the idea of using some cheap projectors to project up onto the underside of the control panel (think rear projection tv) so that, instead of the button changing, the area surrounding the buttons could/would change to indicate the button's functionality on a per gam basis.

In the end it was more effort than I wanted to put into a control panel. However, I think it's feasible to have a dynamic CP. I haven't been looking, but if there is (or will be) a flat panel technology that allows you to drill holes into it without breaking the entire thing, then things will get interesting for CPs.

[SavannahLion]

The AVR Arduino's are abysmal at driving any sort of display and the Arduino stock code suck ass so there is a shitload room for improvement. It's doable for one small screen. A sheer nightmare for two or three smll screens. Just don't expect anything beyond NTSC, ane barely just that.When you need a supporting ARM to get that ---steaming pile of meadow muffin--- working, look elsewhere.

The ARM Arduinos are a little bit more sensible and the PJRC guy has done a very good job stripping the cruft out of the Arduino code for his ARM boards and demonstrating capabilities.  But cost is still a factor.

@Lampry

There is a service that does allow one to drill holes in flat screens. It's crazy expensive though. And I got the impression that it was exponentially riskier as you drill more and more holes. It was posted here sometime last year? Similar tech that chops wide screens to 4x3 sizes for legacy hardware.

[lilshawn]

There is a service that does allow one to drill holes in flat screens. It's crazy expensive though. And I got the impression that it was exponentially riskier as you drill more and more holes. It was posted here sometime last year? Similar tech that chops wide screens to 4x3 sizes for legacy hardware.

yeah somebody was looking for screens that they could use as replacement DMD panels for virtual pinball. someone had contacted one of those companies about chopping screens down. it would have been cheaper to have a custom panel made.

[stigzler]

OFF TOPIC: Cut down LCDS!? Also perilously close to buying ultra-wide litemax - but wondering whether this may be cheaper...Was it this thread:
http://forum.arcadecontrols.com/index.php/topic,110882.0.html
and this service?
http://www.symbolicdisplays.com/LCD_glass.html
I'm guessing it aint going to work out much cheaper in the long run.

BACK ON TOPIC:
Bad news about the arduino boards - so sounding like they can't run more than one micro-oled at once? Back to the drawing board....
Also can't find any dimension specs for the micro-oled screen+breakout board so not sure they'd fit into a button

Dammit - there must be a way!??!

What about LCD matrices? There must be someone manufacturing the basic lcd type thingies you find in calculators/watches etc. You'd just need a transparent one with a 16x16 or 32x32 Matrix. I suppose the problem would remain around how you'd drive each of these... hmmmm

[stigzler]

Hmmm... fused fibre optics? Not sure though how far the light can travel or whether the conduit is flexable:
http://www.incomusa.com/solutions-for-industry/display/
http://www.display-central.com/free-news/display-daily/incoms-polymer-fused-fiber-optics-may-be-game-changer-for-displays/

[SavannahLion]

Didn't say it was impossible. Just hard. The AVR arduino guys tend to make it look easy by selecting screens with i2c or SPI. That means there's a controller on the screen itself doing the timing. IMHO that makes their libraries generally crappy since they're depending on better equipped tools to do the heavy work.

Is that bad? For a beginner I think so. Makes things easier early but hellish later.

The samples are a bunch of stupid Hello World nonsense. Get some screenies up there, ever wonder why almost no one talks about displaying 32x32 24.bit color video? Take a look at the SRAM to understand.

I've done HUGE I/O on AVR already. There are always compromises. If you agree to the compromises, have at it. If not, find a different solution.

My point is, if you want to drive multiple LED screens using the Arduino platform, save youself the headache and select an ARM core.

[Jakobud]

An alternative to mounting the LCD in the button is having it somehow below the button and projecting the image up into the button using a Fresnel lens or something like that. But of course buttons have all kinds of hardware at the bottom of them so I dunno how that would ever work.

[Xiaou2]

Why would one want or need a button display?    98% of the time... your fingers are on top of the button.
The only reason would be for pre-game... and problem is still the same... unless you have a photographic memory.

Better would be a small display that looks like an instruction card... that you mount on the monitor bezel. .or non traditionally on the control panel.

 Furthermore,  remember that many people change button configurations.

[Xiaou2]

I will add... that its quite possible to use a pico projector, mounted above the control panel... to display anything you wanted.
Just glue some projection screen material to your cp surface.  The biggest problem may be keeping it clean.

[stigzler]

There has to be a way... there has to be a way... to quote Ben Stiller:



(sorry - a deleted troll thread started me on this and now I just can't drop it - wish I had that link).

[ChanceKJ]

I almost feel that should be a picture of me with that caption 

When I read the title of this thread I instantly thought of the Optimus keyboard line too. Love Art Lebedev stuff.

And yeah, Yot's $100 a button isn't TOO crazy when it comes to making something cool

[yotsuya]

The day that happens is the day I give up MAME for good.

[stigzler]

Didn't say it was impossible. Just hard. The AVR arduino guys tend to make it look easy by selecting screens with i2c or SPI. That means there's a controller on the screen itself doing the timing. IMHO that makes their libraries generally crappy since they're depending on better equipped tools to do the heavy work.

It sounds, Savannah, that you know quite a lot about this stuff. Sadly, I don't - extent of my electronics knowledge is solder hacks and coding knowledge is visual basic. So feel like I'm a footballer asking Stephen Hawkins about the difference between quantum theory and a 4-4-2 defense (low shared common points of reference).

However, these $15 sparkfun oleds come with the controller board, I think? So - is it possible to wire them in serial or something and have one controller board sorting each display? Does it have to be Arduino? Could it be a (increasingly maligned) Raspberry Pi?

As I say, feel like the Pope discussing different types of condom here, but is it possible??

I almost feel that should be a picture of me with that caption 

Funnily enough, Chance, I posted it as a p**s-take outta some guy, but, as ever, it could actually refer to me too! I think I know that with my tongue in my cheek though - this guy was saying it like he meant it and like he was quoting a great guru. For the record: I have searched and can find no official record of Mr Stiller imparting this comment.

[dkersten]

I have a full RGB keyboard for my gaming PC.  I can set colors for each key for specific games and even have them do macros under certain conditions or blink in patterns based on conditions.  I thought it was cool until I played one game with an intricate profile loaded up.  Turns out I haven't looked at a keyboard while typing or gaming in about 25 years... go figure.  I would think the same goes for arcades as well.  You memorize which buttons do what and it is all positional.  You are far better off (from a practicality standpoint) to have a dedicated button that will flip your display to a control panel view for that game.

Slot machine games in Vegas already have buttons with full active displays in them, and as stated they are expensive and large. 

As impractical as it would be, it would be cool as hell.  This is coming from someone who has full RGB buttons on his CP already, which is also impractical for the most part.  It is good for a quick reference, but once the game is going it isn't worth anything.

[stigzler]

Damn. I didn't think about it being impractical. Yeah, yer right, bad idea. Let's forget it....

Oh no. Wait. Hang on.... Weirdly, I've got a six foot black cabinet standing in my dining room that does nothing but suck up my time, energy and money. Never play any games on the bugger. It just keeps asking me for more bells and whistles, but also insists that I respect its class and dignity as it doesn't want to be like all those trailer-trash cabs.

For reference - that British self-irony and depreciation. You make a good point, dker, but I'd still like to see it happen. Just..because...we....can.

[dkersten]

Hey, I get it, I can't even tell you how much went into my RGB buttons, lol.  I have a 48" lcd tv on the wall next to the arcade cab.. it hasn't been turned on since I installed it a year ago.  The pool table next to these items gets played 3 times per year, and the dart board on the other side of the arcade cab has been unplugged for close to a year as well.  I am an apprentice of the school of "because we can". 

On another note, I did fire up the RGB lighting around my fire pit the other day.. actually came in handy too.  Once a year justifies it, right?

[ChanceKJ]

Funnily enough, Chance, I posted it as a p**s-take outta some guy, but, as ever, it could actually refer to me too! I think I know that with my tongue in my cheek though - this guy was saying it like he meant it and like he was quoting a great guru. For the record: I have searched and can find no official record of Mr Stiller imparting this comment.

  Yeah i saw that in the other thread.  I was just crafting my response when they deleted the whole thing.

I think a few of us who go to "the end of the earth" in a build do it because it shows what type of personality we are, just the same as people like pbj who craft a router table out of a clamp and a board form an old wood pallet. There's nothing wrong with either way, its when you criticize or attack that person for using their methods. There is however a difference between that and notifying people of best practices for safety concerns, or general playability (like angling joysticks). 

UGH! now i really want an Optimus Popularis keyboard to go with my Art Lebedev Battery mug.  Damn you!

[stigzler]

On another note, I did fire up the RGB lighting around my fire pit the other day.. actually came in handy too.  Once a year justifies it, right?

Do you really need to ask that of anyone here? 

...people like pbj who craft a router table out of a clamp and a board form an old wood pallet.

OK. Stop it now... as if we haven't all got enough on our todo list already!

[SavannahLion]

Didn't say it was impossible. Just hard. The AVR arduino guys tend to make it look easy by selecting screens with i2c or SPI. That means there's a controller on the screen itself doing the timing. IMHO that makes their libraries generally crappy since they're depending on better equipped tools to do the heavy work.

It sounds, Savannah, that you know quite a lot about this stuff. Sadly, I don't - extent of my electronics knowledge is solder hacks and coding knowledge is visual basic. So feel like I'm a footballer asking Stephen Hawkins about the difference between quantum theory and a 4-4-2 defense (low shared common points of reference).

However, these $15 sparkfun oleds come with the controller board, I think? So - is it possible to wire them in serial or something and have one controller board sorting each display? Does it have to be Arduino? Could it be a (increasingly maligned) Raspberry Pi?

As I say, feel like the Pope discussing different types of condom here, but is it possible??

Good gravy... Visual Basic?? Yeah, I was like you when I started using Visual Basic during the v4/5 transition. Yeesh... Ever run across those Xtreme VB guys? Don't waste any more time with VB, ditch that ---steaming pile of meadow muffin--- and go to C or C++, Python or ::shudders:: Java. I was so into VB that I even got sucked into that godawful OOPic fad (like the Arduino but Visual Basic/PIC based instead of C/AVR based).

Sure... why not. I was going to have this whole technical write up until I realized the RasPi is actually a pretty good choice for this. Not like you're going to be updating your screens at 15FPS right? For a bare metal approach, I would do ARM unless I can slot a bunch of AVRs in there for cheaper.

I think the new RasPi has 40 I/O now? I may be wrong about that. Using SPI, I think you can certainly pull it off using the R2 without any additional hardware. I believe there is a Python script (It'll take me a bit to look for it) that can directly access the 40 I/O. That gives you an interface that you can modify to set up your various SPI channels and I think Python would be a very good direction to go from Visual Basic (C is better) due to its very friendly beginner language structure, strong syntax sugar, and having now passed my personal fav goto language on the market, Perl. If you use it as a MAME console, I would use a dedicated I/O card like the iPac or the KeyWiz to avoid flirting with that nonsense trying to share the IO with the SPI and the buttons.

Actually... wow... more I think about it, the R2 sounds like a good choice.... I just checked the specs and it does appear that the GPIO are directly compatible with the 3.3v requirements of the screen. You should check the rest of the specs (I think I spotted a 7v requirement somewhere).

The only drawback to that would be if the RasPi2 is replaced, you might have to redo the entire screen interface from scratch if the new board doesn't have the GPIO. But then again, by the time you want the R2 to be replaced, I'm sure micro screen buttons will be a bit more feasible (read: cheaper).

I seriously don't think you're going to get a .66" screen inside of an arcade button anytime soon but I was thinking similarly to another poster, nail the screens to the cardinal points on the CP with the buttons and sticks. I would do that kind of project myself but at $15 a screen? For a joystick, six buttons and a start that's $165 per player not counting credits or admin buttons just for the screens. 

[stigzler]

Wow. Thanks for the reply, Savannah, but in the words of Alan Sugar - "I'm out"!!! Whoa - just reading all your technical language made me realize that this is maybe beyond me as all the hardware + I/O references were way beyond my comprehension.

Maybe I just concentrate on my already stoopid to-do list rather than keep starting 'another thing...'

Also, maybe one day some bright-spark will knock a prototype together. Who knows - the funspark fellas also have some arcade related stuff - maybe their interest could be piqued?

[Projekt Produkt]

I just got done reading through this thread and your idea is very cool. 
I can't think of any reason not to do it, after all this forum is "Build Your OWN Arcade Controllers".

I2C protocol allows up to 128 devices to be addressed on the bus (http://forum.arduino.cc/index.php?topic=159860.0) 
thats plenty for enough buttons and still will be able to run all your other I2C devices such as haptic feedback
and Real Time Clock, etc.  Since I2C breakout boards exist (the SparkFun comes with one) it makes it quite a bit easier
to accomplish.  Each I2C breakout board should have it's own buffer to store and display a picture and I can see from the
picture of the PCboard that the address can be changed through jumpers.  However, the data sheet states this one is designed for
only 2 different addresses (https://cdn.sparkfun.com/datasheets/LCD/Monochrome/Micro-OLED-Breakout.pdf)  so a different breakout
board might be prefered.  However, a shift register might get around this problem with the sparkfun unit.

Getting it into an existing arcade pushbutton housing might be another matter but I don't think we are interrested in
how hard that could be.  Technology changes fast and so do prices.  The more we build this idea the more manufacturers
will cater to us and deliver cost effective and smaller tiny screens.

So I propose a change to the definition and scope of the project to start the ball rolling. 

1. Make a One and Two player start button screen display the player's character for each game selected. (For example if
Galaga is selected it will display one ship above the one player start and two ships above the two player start button.)

2. Use tiny screens either in the top of bushbuttons or mounted on the control panel above the one and two player start buttons.
   a.  Maybe this http://www.adafruit.com/products/684 screen from Adafruit is a better solution.  It's in color and has an SD
       holder built in.

3. I like the Teensy++ 2.0 microcontroller, so I will use that.
   a. SparkFun MicroView - OLED Arduino Module (https://www.sparkfun.com/products/12923) has a microcontroller built in but not color

4. Use USB serial protocol, from a frontend, to tell microcontroller what graphics to send to the screens.

So there is the definition.  Shall we proceed?

Kevin
www.projektprodukt.com

[SavannahLion]

I think it's interesting when people use dynamic addressing methods using things like shift registers or decoders when it comes to i2c. Completely defeats one of the bullet points. Might as well straight up use SPI or some other serial protocol more suited to the task.

It's your money though. Might I suggest a couple of things.

There is a i2c shift register that will help save the pin count off of the controller. There is also a sort of i2c "hub" that was specifically designed for this scenario. I don't have part numbers on hand but any reputable supplier will have them. Remember that either system adds latency which can be problem in some designs so work around them.


I would really test using an i2c system over such a broad area... let's see.... 22... 24 screens maybe? What's the capacitance on those lines going to be like? What's that going to do for the bandwidth there? Those i2c hubs might help here.

[Projekt Produkt]

Points taken.

However what this forked definition of the exercise is about is not the what if's and how many ways can it be done.  It is about taking the idea seeing what is readily available and easy enough for anyone to accomplish and going ahead with the development phase.

Let's get a couple screens going.   Who knows, at the end of the exercise we might just have what amounts to an open source multi-monitor controller that anyone is free to build and improve upon.

But like any project, it has to start at a beginning.

[SavannahLion]

I wasn't exactly offering what ifs. Just options.

Open source? I missed that one. In that case avoid PJRC products like the Black Plague, his bootloader is proprietary.  Not saying they're bad products, they're quite good. But if OS is the name of the game, look elsewhere. Really, there is an enormous number of ARM boards out there to choose from. I use the T.I. boards because the dual ARM boards are super nice for debugging and status but I doubt it's Arduino compatible. And yes, I know T.I. uses proprietary code for the firmware but there is OS firmware that is maturing quite nicely. To that end, make sure you see what the Arduino ARM boards are doing when configuring Serial over USB on windows boxes. Windows 8 and later are pure ass to configure, even for the Vanilla Serial ports. I don't use Arduino myself so I don't know if they're depending on FTDI for the driver signing.

I would personally go with i2c hubs as wiring for a new guy for them is going to be simpler. It just loads more work onto the software side. I would avoid an addressing scheme via shift registers, that's just one more thing to deal with when wiring.

[fischb]

Reviving an old thread.  have you guys seen the Elgato Stream Deck Mini?  Maybe this could be programmed to create a dynamic arcade control panel.  Im sure their buttons are not that costly.  But they would need to be round and good enough quality for bashing on an arcade panel

https://www.elgato.com/en/gaming/stream-deck-mini

[meyer980]

I watched a teardown of the Elgato. The individual buttons aren't screens - underneath is one large lcd panel that just pragmatically shows the proper image underneath the clear button.

Makes sense if you think about it. Much cheaper and more efficient to buy a premade lcd panel.

[lilshawn]

interesting so it's an LCD with a resistive touch overlay with clear buttons over top? so when you push a "button" you actually push the touch overlay with the clear button?

sounds like an interesting DIY project.

[Howard_Casto]

Yeah which is why it amazes me that people will pay $100+ for those stupid decks.  It's $20 worth of parts guys.