Keyboard Interface/Encoders

[Infrasonic]

I'm looking to design PC-Keyboard encoders, with between 14-26 non-matrixed/non-ghosting individual inputs.  Maybe a 50+/- input unit too.

What type of features would be usefull for it?  Keyboard pass through?  What characters be would good for it, assuming that each input would have set character?

[crsdawg]

i think randy and andy have this niche pretty much covered. seems like you'd be setting yourself up for some pretty stiff competion.

i do see a market for a basic 10-12 input cheapo interface. something very basic, doesn't even need to be programable. i've got some specific projects on the drawing board that don't need 25-32 inputs(single player desktop control panel for example). i feel a kw or ipac would be a big waste since i would never use even half of their inputs.

i guess i could always hack a sidewinder(which is probably what i'll end up doing), but i'd kick in a few extra bucks for something that was more plug and play, and smaller. a guy on ebay was selling something like this, but the quality was suspect, as was his shipping promptness.

price would be key, since a 32 input programable interface can be had for 28 bucks(keywiz ecco).

2 cents spent, good luck

[Spaced Invader]

I agree with crsdawg. A lot of people who have built their monster cabs are now looking to build something more specialized (a single player vertical is in my future), others have wives that would veto a mulit player space hog but would allow a smaller unit, and then there are always gift machines...so, i think there is a large untapped market for a cheap, 12-16 button controller. Just my opinion!  

[neuromancer]

I agree with the other two. How about a unit that had inputs set up for the arrow keys (4), shift, ctrl, alt, "5",  "1", and esc. That's 10. If I could have two more, they'd be "2" and "P".

It would need to be inexpesive and high quality though, to compete with the Ipac. If it's around the same price, then I go with the proven product.

Bob

[SirPoonga]

I agree, the encoder market has an encoder for anyone's taste.  The mk64 has ikari warriors support built in.  The hagstroms can do both keyboard and mouse in one board.  The IPAC and keywiz are customized to this market.

the one thing that hasn't been made is the uber encoder.  But that would get expensive.  If you could combine the optipac with the ipac that would be helpful.  Also design a break away board like hagstrom, that way you can use a break away board for mudlar/swappable panels or a JAMMA fingerboard to make this an all in one product to compete against everything from ultimarc  Also it would have to support two ikari warrior controllers.
This would have to be a USB product in order to pull that off which would limit you do just people who run windows.

However, I (and 1up) have different idea that could use the help of someone who knows how to build stuff like this

Eitherway you probably wouldn't make much but pocket change with all that competition.

[tmasman]

If you think you could develope/build an encoder with 14-26 inputs & sell it for less than $20 then you might be able to sell quite a few to people looking to do smaller projects.

However, if you can't get away with selling them in that lower price range, then you'll have a hard time breaking into this market.

The KeyWiz is only $26.95 (Hey Look Randy! I got it right this time... Just kidding...)
And it has 32 inputs with 24 more shifted inputs...


As far as features...
A Keyboard pass-through might be a nice optional add-on (compare to KeyWiz options)...
Programability would be a nice feature to help you avoid selling to the "MAME Only" crowd...

Keep us updated with your plans...
(Also... Are you an EE student, or hobbiest, or )

[DarkKobold]

I'm actually writing my own keyboard encoder - I have to say tho, its a pain in the butt. It's finally mostly working... but if you add up the parts, I'm not saving much money....

$11 microcontroller
$3 resistors
$3 vector board
$1  socket
$3 keyboard extender cable
$1 clock
$5 power socket ends

comes to about $27... with all the time I spent soldering, writing the code, debugging, add some for that.

www.Microchip.com saves money, because they send sample microchips for free, and I already had some of the other parts...

It was worth it for me. It helped me learn coding for the PIC so I could do my final project for Senior Design. Also, I have 2 cabs that need keyboard encoders, so it will help there.

Not worth it tho to the normal hobbiest. I applaud Randy for being able to make the keywiz so friggin cheap. It's expensive to build your own.

It was worth it for me

[Infrasonic]

any type of reprogrammability on the encoder board itself would increase the complexity of the unit.  If you mean programmability in the form of a program on the computer itself, that would be easier from a production point of view.  However, I don't know how to make "device drivers" so I would have to make it compatible with someone elses, freeware.  Or make it non-programmable.  14-16 individual inputs, and 1-4 outputs, or LED drivers possible.  Price of the device would be competitive.  When you speak of 'shift' would that be intended to modify the output of certain keys?

How would you connect this to a computer, would a keyboard pass through be appropriate, how would you use such a device if it didn't have pass through?  Do you swap connectors in the back everytime you write an email?  What about a pass through that is contained in an external box, so you wouldn't have to mess with the keyboard to computer connection when plugging or unplugging the stick.

[tmasman]

OK...
I figured "prgramability" would be a big fat NO...
Let's forget that idea for the moment...

The easiest way to describe the "shift" function is you send the shift input along with one of the regular inputs & the encoder sends a completely new key to the PC...

Example: Pressing "Shift" & "Down" together would send "Esc" (or some other key press).
This is a feature present on the Keywiz.

Again, this wouldn't be a neccesary feature, but you were asking about feature ideas, so I threw it out there...

As far as the pass through...
Most of your market for this would probably end up being people who want to put this in an Arcade cabinet... They would either not have a keyboard present, or use a USB keyboard.

If you could make your encoder USB based, you'd definately have more going for you... But I'm going to assume that's not an option (to keep the price down).

[JamIt]

crsdawg and SI are on the money.  I've gotten many requests for simple single player panels and I'd like to give them out as gifts.  But I can't warrant buying the keywiz or ipac (both great products by the way) when I don't need all the inputs.
Also, I think you could reach out to to the jukebox and ddr crowds as well as mame.
good luck and keep us posted,
--JamIt

[tbombaci]

With all the projects I have brewing, I am good for 3 or 4 of them.

-Tom

[eightbit]

The best way I can see would be to get the cost under $20 shipped. I can hack a ms sidewinder for 14 inputs for under $5. Using win98  I can have it send keystrokes with the native driver. Though in that price range I don't see how you can make money at it.

[Infrasonic]

Let's have a vote/poll:  everyone name their 14 or 15 most wanted ascii characters (not in ascii code please) in order of importance so I can build a proto.


up, left, down, right, cntrl, enter, esc,.......

[SirPeale]

One player controller?  

Up, down, left, right, left control, left alt, space, 1, 2, 5, 6

Nice for a bartop.

[eightbit]

One player controller?  

Up, down, left, right, left control, left alt, space, 1, 2, 5, 6

Nice for a bartop.

Secound that execpt maybe one more button, for player 1 button 4. Depends on what the limits of your encoder are.

Absolute min is 8 buttons up, down, left, right, lft ctrl, lft alt, 1, 5. with 10 buttons add 2 and 6, with 11 add player1 button 3 and with 12 add player1 button 4.

[DarkKobold]

Hey... If I programmed just the PIC, would anyone want one sent to them for a certain price? Like, the cost of the PIC, the clock, shipping, and maybe a little extra... You'd have to solder all the resistors and buttons yourselves... But since I've already got the code.... Its an idea to save on cash. Plus, you could pre-order the inputs you wanted. I've got the shift function working. I just need to debug at the moment.

[Infrasonic]

left, right, up, down, space, l ctrl, l shift, l alt, 1, 5, 2, 6 that's twelve so far, anything else?

[hazydave]

Well, I'm designing my own keyboard controller, too. It's not a money saving move, but for fun, I suppose. Also, when you're known as a hardware guru, it's just wrong to use someone else's. And as well, naturally, they're all doing them wrong :-)

I sort of which I had found this forum earlier, maybe I could pick up some hints on what folks found useful, and not-so-useful, for controllers. I'm going the PS/2 route for both keyboard and mouse replacement on the one board, with a keyboard passthrough (or to link two controllers).

My basic controller supports two players. Each player gets a joystick and 8 buttons. Gaming buttons, naturally, are all non-blocking, but also all capable of generating an interrupt-on-change to the microcontroller (most of the large-count controllers do polling only). I'm also supporting two flipper buttons, a coin drop trigger, one trackball and one spinner (with hardware counters, so you don't get the aliasing problem that makes them seem to go backwards), 4 high current LEDs, and eight potentially blocking non-gaming buttons (for player 1/2 buttons, a pause button, maybe some other stuff). A shift is possible, but I haven't given it a great deal of consideration.

This one's small, 2.5" x 1.9" or so, 2-layer PCB.

Anyway, just wanted to add to the proposition that, even with other controllers on the market, there's room for more, or even a hobby or possibly Open Source controller (I haven't quite decided what to do with the design, once it's completed).

[Pipercub]

I would be interested in being able to buy just the progammed encoder chip. I could put together my own PCB and such for allot less than the IPac

[Infrasonic]

I have encoder chips for the PlayStation, requiring only 5 resistors and a recomended fuse.

IBM encoders are at least a week away.  

On that note:  Any good freeware MAME games to test out the IBM encoders?

I don't know much about PC style keyboard controllers except that my equipment was INVENTED to build them so they shouldn't be too much of a challenge.

[Infrasonic]

What would be a good freeware mame setup to evaluate pc-joysticks?  I am also interested in USB.  Are there any freeware Device Drivers available on the internet for people to use?  I'm asking because designing a controller and writing a device driver are too very different things and I would rather make it compatible with existing established formats than to develop my own.

   

[Donger]

As a budding cab builder I would certainly say that there's demand for a universal cross-platform encoder (psx, USB, x-box, etc.).

[Grasshopper]

There's always room for one more design....

I used a keyboard hack for my first panel. I don't mind that it's not programmable but I really miss not having a keyboard pass-through facility.

There is actually already a pretty good freeware keyboard encoder available:

http://surf.to/buttonbox

I have considered building one but the Atmel chips that Leif used in his design are quite expensive where I live. However I've found the following company that sells then quite cheaply:

www.futurlec.com

[Infrasonic]

thanks for the info grasshopper, I haven't even begun to research the keyboard pass through function, so this could help me out a lot.  Since I hope for commercial use, I won't be copying or duplicating their methods in any way.  I have several keyboard methodologies already but I think I will just create one from scratch.  There is something on keyboards that makes it so when you hold down a key it will start repeating itself automatically.  I have decided to do away with that on the encoder itself so that if someone wanted to still have that or any other features on their keyboard pass through, it would still work.  I hope to have some prototype alpha-beta models out soon (tell a friend:).