Grounding wires..

[JeSo]

I'm currently in the planning stages of building a CP and have come across something which confuses me somewhat. I'm planning to do a keyboard hack, and have read numerous articles about this. Some of these articles mention (only briefly) attaching a grounding wire across all of the controls, and others don't mention it at all.. My question is, is this neccessary? and if so (or if it's a good idea), how do I go about this?

Thanks.

[vusteveii]

check the faq for some good info on the keyboard hacks, i think the vast majority of the forum would however reccomend you go with an encoder(ipac, hagstrom, etc) to save a lot of time and frustration.  Doing a keyboard hack right involves breadboards and diodes and a lot of other things far more complicated then running a ground.

[RandyT]

I'm currently in the planning stages of building a CP and have come across something which confuses me somewhat. I'm planning to do a keyboard hack, and have read numerous articles about this. Some of these articles mention (only briefly) attaching a grounding wire across all of the controls, and others don't mention it at all.. My question is, is this neccessary? and if so (or if it's a good idea), how do I go about this?

Thanks.

Most, if not all, keyboards use a matrix for scanning keypresses.  This means that you can't just run a common ground to all of your controls like you would when using an encoder.

With a keyboard hack, you'll probably have about 4-5 commons (not the same as ground)which will have to be connected based on what you find to be optimal when bugging out the keyboard.  BTW, I've actually been through hacking a keyboard myself, and it's a pretty formidable task, even for the experienced.

RandyT

[Apollo]

Oh my god! You better list 3 good reasons for doing a keyboard hack immediately!
Why? why? why?
Why go to all that trouble and end up with a controller that doesn't work properly?
IPAC, KeyWhiz, Hagstrom KE72, MK64 there are so many other better ways to go. A KeyWhiz eco is only $26.95 for pitys sake.

[paigeoliver]

The ONLY reason I could see for doing a keyboard hack would be to do a jukebox, or something else with 100 buttons that you only press one at a time.

[JeSo]

Hmm, everyone seems so happy with the idea of a keyboard hack.
The main consideration with doing it this way is the cost. From memory, when I looked into it, an Ipac would end up costing me roughly AU$100, whereas I just picked up a keyboard for AU$2. Is it really *that* bad a way to go about it though? Why? I've seen quite a few examples and tutorials about the keyboard hacking method. It seems fine to me. I'm only doing a single player CP too , btw..
Anyone who has done it want to jump in with defense/tips?

[Hoagie_one]

For a one player cab, a joypad hack may be easier to do.

Or maybe this alternative

[hooded_paladin]

I recommend hacking a playstation controller.  14 inputs each controller, and it's $15-20 for a two player psx->pc converter.  I'm not sure that there are any sites close enough to you to keep down the shipping costs, though.

[Hoagie_one]

for the price of a PS1 controller and the converter, you could just buy a KeyWhiz eco

[Raleigh]

Well if you are set on doing a keyboard hack, I say do it.

I was having some issues with my keyboard encoder when I first got it and decided to hack a keyboard to get me by while I get my encoder working.

What I found with a keyboard hack is that you learn everything as you are actually doing it.  Sure I read about the details of it before attempting it, but some areas I didn't understand until I was at that step in hacking it.  So I say get a cheap keyboard, or take one you have sitting around and start the process.  By the time you get to the wiring you will probably understand your current question.  A keyboard is a matrix, mine was a 18x8, but they vary.  So each button has to have a wire coming from the 18 and one from the 8.  The more buttons that share a common wire increases the chance of ghosting, which is what you want to avoid. Once you map out your matrix, you can determine a good wire configuration for each button, i.e. up and down can be on the same wire since they cannot be pressed at once, eliminating the chance of ghosting.  But like I said, just go through the process and most likely you will understand it as you go.  I used the following tutorial to get me through the process:

http://dragonsden.emuunlim.com/ddkeytutl.htm

Also the following site is a good resource as well listing pros and cons and helping with the mapping of the matrix:

http://www.mameworld.net/emuadvice/keyhack2.html

Be patient and hack away.  

With that said, I found the keyboard hack as an alright temporary solution, but wouldn't sugest it to to anyone with more that a dozen or so inputs.  It was a painful process, but I am glad I went through the process.  Now that I have my encoder working I can appreciate it even more, and realize it is worth every penny.  

P.S.  I have heard that some keyboards don't work as hacks or can even screw up your motherboard, so you may end up putting in all the time for nothing, but luckly mine worked fine.

[tmasman]

Come on guys... It's not that hard to hack a keyboard and you can still get about 20 - 24 unghosted keys with most keyboards (8x18 matrix).
If you're just planning on 2 player with 6 buttons, then you shouldn't have many (if any) problems. Granted you'll have to remap your entire MAME config, but that's not too big of a deal anyways.

I had never messed with this type of thing before and I hacked a KB my first try. Sure it took alot of time, but it was a fun learning experience and only cost me $1.50 for a used busted up old KB.

If you're try to save $$$, and you aren't worried about taking your time (I mean at least 3-5 hours) then by all means have fun hacking a KB.

To answer your original question, you will not have a common ground wire attached to all your switches.

If you have any questions please feel free to port here, or email me (or PM). I'd be happy to help you save a few bucks.

By the way... For any future Control panels I plan, if it's going to be in a cabinet, I plan on using a KB hack for player 1 & admin buttons, but I'd hack a gamepad for PL2+, just for the convenience of not having to mess with possible ghosting. (But I still recomend a KB hack for your first Control Panel).

[RandyT]

Come on guys... It's not that hard to hack a keyboard and you can still get about 20 - 24 unghosted keys with most keyboards (8x18 matrix).

If you got that many then you are insanely lucky.  Many modern keyboards have firmware to prevent more than 6 keys from being pressed at any given time.  This is probably ok for a single player though.

If you're just planning on 2 player with 6 buttons, then you shouldn't have many (if any) problems.

I think that is being a bit optimistic.  Much of one's success or failure depends greatly on the keyboard being hacked.  You might find that after all the work you put in, it won't do what you thought it would.  Some individuals have hacked 3 or more to find one that worked well and there are even databases that keep track of the "best" keyboards for hacking.

JeSo:

I have a hard time recommending a keyboard hack to anyone who isn't chronically unemployed and to the point where every penny counts, or extremely bored and looking for a personal conquest.  It's a lot of work, it's easy for a novice to ruin parts (like keyboard fuses on motherboards), and the end result is usually pretty sub-standard.  The keyboard I hacked was noticably poorer in performance than a dedicated encoder.

There is a LOT of information about hacking keyboards, both on this site and on the net as a whole.  I suggest you do a lot of reading before attempting this and don't base your decision on the successes or failures of a couple of individuals.  This is going to require a certain amount of introspection on your part, after educating yourself, to decide if you are "up to the job".  If you haven't soldered before, you'll need to practice.  If you don't own a multi-meter, you should get one.  It'll help tremendously.

For some, the task is trivial, but time consuming.  For others, it is impossible and cost them a motherboard to find out.  You should probably try to figure out which group you'll end up in before you get too far into it.

RandyT

[JeSo]

I've done a lot more reading around now, but I still think that this is the way I'll go about it, or at least attempt to.
One question though, I've seen various people mention the dangers to the motherboard of a bad hack, I was wondering, where exactly does this danger come into it? What is the thing that can go wrong and kill/maim/etc my mb?

[paigeoliver]

i THINK that the keyboard port on your computer dies if you blow the fuse on the port, which isn't replaceable as far as I know.

[tmasman]

What is the thing that can go wrong and kill/maim/etc my mb?

The dangers only come in when you've crossed a couple of wires, or have done a sloppy soldering job.  To prevent any problems, follow these steps:
- Remove PCB from KB
- Solder (about) 10" wires to the KB connection points (Where the KB membranes connected to the PCB)
(On many KB there are 26 connection points (18+8))
(There are ways to get around so many soldering points, but this makes mapping the matrix easier IMO)
- Check (with a multimeter) that none of the wires are crossed and for bad soldering.
 (Probe each adjacent wire to make sure there is no current flowing between them)
- Check again
- Check again

* That's the hard part *
* Now for the long part *

- From here you can attach your wires to some kind of screw strip to allow for easy access
- Start mapping the KB matrix (there are plenty of tutorials floating around, or email me for details)

Now you get to figure out the best possible keys to avoid ghosting

This stuff isn't very hard, you just HAVE to pay attention to the details. And if you've decided to give it a shot, just take your time & ask alot of questions before ever pluggin in the hacked KB to your PC.

Above all of that, HAVE FUN!
& Keep us posted here...

[SirPeale]

When I hacked my KB, I used what are called 'wire bits' to drill holes in the board.  Then I put wires thru, and soldered the wires to the board, as opposed to soldering directly to the trace.

Of course, you should only do this on a single sided board, or risk going thru another trace on the other side.

The bad thing is...I can't solder, so I ended up using too much heat for way too long,and most of the traces pulled off the board.  Otherwise, it would have worked like a charm.

[JeSo]

- Check (with a multimeter) that none of the wires are crossed and for bad soldering.
 (Probe each adjacent wire to make sure there is no current flowing between them)
- Check again
- Check again

I *think* I understand. So, for each wire, i test it with the wires on either side of it? Like, say, I grab wire number 7, and test it with wire number 6 and wire number 8? I might add another 'check again' step though for good measure  

One other thing, I've read a few different methods for mapping the keyboard's matrix, and the one that seemed nice and simple to me involved using a multimeter to check each key on the actual mylar(?) sheet itself, is that feasible/recommended?

Thanks for all the help too  

[DarkKobold]

I will soon be getting a cocktail cabinet, that I plan to convert to Mrs. Pacman/Dig-Dug/Donkey Kong. as far as i understand, hacking the keyboard is the best/cheapest/most interesting option. I only need 1 or 2 buttons, and the coin door. Besides, you should only need to hit a max of 2 buttons at once.

[BobA]

I found that the mylar sheet was hard to follow and that the easiest way was to solder wires to the contacts on the encoder and to make up a map of the resulting keypresses as I touched the wires together.   You only have to touch the x wires to the y wires as there are usually 2 distinct areas where the keyboard connects.  you could use letters vrs numbers as easily.


   a b c d e f g h i    (X)
1
2
3      Map Keypresses
4         in resulting grid
5
6
7
8
(Y)


BobA

[Tiger-Heli]

Come on guys... It's not that hard to hack a keyboard and you can still get about 20 - 24 unghosted keys with most keyboards (8x18 matrix).

If you got that many then you are insanely lucky.  Many modern keyboards have firmware to prevent more than 6 keys from being pressed at any given time.  This is probably ok for a single player though.

Alright, this is getting a little misleading.  You get 18 unghosted keys with a 18x8 matrix.  However, the magic number is 16 since the opposite directionals can be allowed to "ghost" since they will never be pressed at the same time.  Thus:

Street Fighter - 12 buttons, P1 Up, P1 Left, P2Up, P2 Left = 16 inputs, and then "Ghosted" inputs for P1 Down, P1 Right, P2 Down, P2 Right.

TMNT - 8 buttons, 1Up, 1Left, 2Up, 2Left, 3Up, 3Left, 4Up, 4Left = 16 inputs and "Ghosted" inputs for Down and Right.

USB Keyboards are limited to 6 inputs by the USB spec.  PS/2 keyboards often employ "blocking" which prevents 3 simultaneous keypresses from registering, but usually do not limit overall number of inputs if they are not in the part of the matrix where they will form a square.

If you're just planning on 2 player with 6 buttons, then you shouldn't have many (if any) problems.

I think that is being a bit optimistic.  Much of one's success or failure depends greatly on the keyboard being hacked.  You might find that after all the work you put in, it won't do what you thought it would.  Some individuals have hacked 3 or more to find one that worked well and there are even databases that keep track of the "best" keyboards for hacking.

I disagree, most keyboards will use a 16x8 or 17 by 7 (or 8 ) or 18 by 6-8 matrix, any of which will produce the results above.  Only a very few will use a 15x9 or 14x10 or 13x11 matrix, which will not produce acceptable results.

I have a hard time recommending a keyboard hack to anyone who isn't chronically unemployed and to the point where every penny counts, or extremely bored and looking for a personal conquest.  It's a lot of work, it's easy for a novice to ruin parts (like keyboard fuses on motherboards), and the end result is usually pretty sub-standard.  The keyboard I hacked was noticably poorer in performance than a dedicated encoder.

Agreed

For some, the task is trivial, but time consuming.  For others, it is impossible and cost them a motherboard to find out.  You should probably try to figure out which group you'll end up in before you get too far into it.

Also agreed.  Some further things to consider:

While you should have enough inputs to play Street Fighter, or TMNT, there is NEXT TO NO chance of having P1 use the arrows keys and B1 = L Ctrl, B2 = L Alt, B3 = Space, etc.

This is not a problem in MAME, b/c the keys can be reassigned.  However, if you ever use any programs without assignable keys, you might be out of luck.

Also, a keyboard hack will be flimsier than a dedicated encoder, and you will have to re-do all your effort (mapping the matrix, soldering the leads, etc) if you break the hack.

[Tiger-Heli]

I found that the mylar sheet was hard to follow and that the easiest way was to solder wires to the contacts on the encoder and to make up a map of the resulting keypresses as I touched the wires together.   You only have to touch the x wires to the y wires as there are usually 2 distinct areas where the keyboard connects.  you could use letters vrs numbers as easily.


   a b c d e f g h i    (X)
1
2
3      Map Keypresses
4         in resulting grid
5
6
7
8
(Y)


BobA

I don't recommend this method (although I am pretty sure it will work).

First off, I don't recommend trying to "follow" the traces on the Mylar sheet, I recommend using a multimeter and checking when you get continuity between the Mylar contacts under the buttons and the contact points on the encoder chip.

Other advantages of this over your method:

1) Can be done (should be done) with the keyboard disconnected, so no chance of shorting anything.
2)  It's not always easy to tell where the connection points are and which ones are rows and columns.  I'm not sure what happens if you connect two row inputs together, but it might not be good.
3)  Many inputs (L Alt, Windows Key, etc.) won't display on a screen like NotePad.  (I realize GhostKey or some of the other utilities will work).
4)  A 16x8 matrix is 128 inputs, but MAME can only use 104, so you have 24 unused and unusable inputs that you still need to map out.  (Add another 36 to that for an 18x8 matrix).
5)  If the keyboard ends up being a poor choice to hack, with my method you can reassemble it and either use it as a regular keyboard or yard sale it.  With your method, you have spent a couple of hours soldering wires to a keyboard encoder that you will never use.

[RandyT]

Alright, this is getting a little misleading.  You get 18 unghosted keys with a 18x8 matrix.  However, the magic number is 16 since the opposite directionals can be allowed to "ghost" since they will never be pressed at the same time.
...
USB Keyboards are limited to 6 inputs by the USB spec.  PS/2 keyboards often employ "blocking" which prevents 3 simultaneous keypresses from registering, but usually do not limit overall number of inputs if they are not in the part of the matrix where they will form a square.

Just because the chip inside the keyboard uses an 18x8 matrix does not, and I repeat, does not automatically mean you will get 18 or even 16 keypresses without blocking or ghosting.  It's all about how the firmware is written, which happens to vary unpredictably between keyboards.

The 6 keypress limitation is not only a USB phenomena.  I have about 40 PS/2 keyboards that only allow 6 simultaneous keypresses.....period.  I did a lot of research on this and tried all the "tricks", like diodes, etc.  None made any difference whatsoever.  There was a 6 keystroke ceiling imposed by the firmware.   Further research revealed that I was not the only one to have come across this situation.

This is why this is so hotly debated.  Each person isn't using the same keyboard for the experiment, so oranges and apples are getting compared on a grand scale.  Inevitably, this leads to some individuals blaming the lack of success on the incompetence of the individual, rather than the keyboard, which is where it usually belongs.

I disagree, most keyboards will use a 16x8 or 17 by 7 (or 8 ) or 18 by 6-8 matrix, any of which will produce the results above.  Only a very few will use a 15x9 or 14x10 or 13x11 matrix, which will not produce acceptable results.

Again, I stand by what I said because it was my experience.  If you fared better than I, then congratulations.  Just because one person returns from Vegas a winner after visiting for the first time, doesn't mean that everyone will .

As I stated, your success or failure will depend greatly on the keyboard you choose to hack.

One of the reasons I decided to offer the Eco version (no, this is not an adverstisement) was to offer a reasonable alternative to the 2 days of soldering, mapping and inevitable troubleshooting that was my experience when hacking a keyboard.  I eventually ended up with a Gravis gamepad hack next to it and using the mouse buttons just to get enough buttons for a 2-player 6-button panel, because I sure as heck wasn't going map out ANOTHER keyboard chip .

But since none of this has any effect on me one way or the other, have fun with it.  It's a real learning experience.  It's just that what I learned from doing it is that it wasn't worth the effort when I was done.  Just trying to save others from the trauma I experienced.

RandyT

*EDIT*

For those following this discussion, I had incorrectly remembered the ceiling set by the keyboard as 6 keypresses.  The actual number was 8, so all references to 6 in the above should be replaced with 8.  The ceiling still exists, but at 8 keypresses.  My apologies to those who those 2 extra keys would make all the difference in the world to.  (off now to punish myself   )

[tmasman]

Once I had checked all of my wires.
(Yes to your example: make sure wire 7 is not "connected" with 6 or 8 with your multimeter)

I plugged the hack into my PC & used a program called keyhook
(Download keyhook here)

You'll have a matrix (for example) of 8 x 18 and (first # = 1-8, second # = 1-18)
connect wire 1 to 1 and write down what keyhook says that keypress is.
Now connect wire 1 to 2 ... etc etc... continue with wire 2... yada yada...

Just remember some cells in your matrix will not be used & keyhook will not register anything for those wire pairs when they are crossed.

[Tiger-Heli]

The 6 keypress limitation is not only a USB phenomena.  I have about 40 PS/2 keyboards that only allow 6 simultaneous keypresses.....period.  I did a lot of research on this and tried all the "tricks", like diodes, etc.  None made any difference whatsoever.  There was a 6 keystroke ceiling imposed by the firmware.   Further research revealed that I was not the only one to have come across this situation.

Ok, I stand corrected.  I tested three of the "Free with Rebate" keyboards (all different brands) and all of them used at least a 16-column matrix and would register 16 simultaneous keypresses.  I thought 3 for 3 was pretty good odds, but I guess that was incorrect.

Having said that, after mapping it out, I never did anything with it due to:

1) Complete inability to solder.
2) Wanted to be able to re-map inputs if needed.

[aricade]

Well this is a long thread, so if my response is the same as someon elses my apologies.  I'm a little impatient, It's one of my weaknesses...

Any hoo....
I tried doing a keyboard jhack fo r fun. and this is what I discovered.. Most modern keyboards do not have a ghosting problem simply because the block the possible 3rd key that would create the ghost (after pressing 1st, and 2nd key), So you can never make an L shape in a matrix; Try and look at your matrix as if it was a CHESS BOARD or graph paper...  if the keys creat a potential L shape you'll have a ghost.

Right but you probably knew that.  So with my keyboard I could barely find a usable 2 player 6 button config, due to key blocking tactics of the keyboard, "barley" I said but I did find it after 4 days stumped...  saying that it might just be easier to hack a gamepad and a keyboard as you would have enough inputs. But since your hacking gamepads why not just hack another...

If your on a mission for hacking a keyboard try finding a really really really really old AT keyboard...  That does not do Key blocking.  If your keyboard does not do key blocking then you can use the diode trick...

If you want to go cheap and know how to solder just build a snespad interface..  You can hook it up to your lpt port in win98 and dos you can use sneskey wich will act just like a keyboard.   And there really is anything quite like burning your fingers with a soldering Iron  .  If your interested in this and cant find info pm me but theres info here at BYOAC and on Steph Hans page about this...

[Tiger-Heli]

Most modern keyboards do not have a ghosting problem simply because the block the possible 3rd key that would create the ghost (after pressing 1st, and 2nd key), So you can never make an L shape in a matrix; Try and look at your matrix as if it was a CHESS BOARD or graph paper...  if the keys creat a potential L shape you'll have a
ghost.

Right, most modern keyboards (last 10 years) use blocking, but you can still get enough inputs with this.  Apparently some newer ones also limit you to six simultaneous keypresses which is not acceptable.