The Intellivision controller matrix - Guide!

[bacteria]

Worked out the matrix, works fine. A lot of games don't use the 16 D-pad directions on the Intellivision anyway, and those that do are perfectly playable with four directions. I played Tron Deadly Discs for some time yesterday absolutely fine - in fact, seemed to have if anything more control on the character - the Intellivision controller was a bit "hit and miss" on accuracy at the best of times anyway. Further evidence, if needed, that 4 direction for the D-pad is fine is the "Intellivision Lives" compilations on the PS1 and PS2 as well as the rather poorly ported plug'n'plays for the Intellivision.

 

Here's the product of several hours work and testing - all works fine. Using tact switches is so much better than the membrane ones, and, the tact switch matrix I built (10mm tacts) is the same size as the original mebrane set, so it would be quite simple to make a new controller with a better keypad and d-pad than the original Intellivision one, and still use the original game overlays! (I'm not making a replicate controller for my project, however this diagram would be useful if you wanted to make one).

 

[bacteria]

BTW - if anyone would be interested in commissioning me to make a replacement controller for the Intellivision, based on the above, using the original d-pad and replacing the stock keypad with tacts, etc; PM me. I can use the original controller although of course it will be thicker in the middle section than the original (the keypad section). I think £50 per controller including labour and parts is a very fair price, you provide a controller (I only have one spare one), i'll do the rest. Please PM me if you are interested, and we can discuss. Postage - UK £3, rest of world £10 incuding tracking. Payment in UK British Pounds, PayPal. First commission I would do i'd make the item first before charging, proof of concept, follow on commissions payment up front. PM me if you are interested.

[cparsley]

your missing something, each bottom button maps differently. While the top button is same on both sides, the bottom buttons are different from left/right. As well, games like Vectron will epically fail with only a 4 direction controller.

 

PS - The "plug and play" abomination is actually a NOAC (Nes On A Chip), and not anything remotely emulating the INTV. Those should burn in hell....

 

PPSS - GREAT GREAT WORK on all the rest, bravo!

Edited June 28, 2010 by cparsley

[bacteria]

Indeed, I never said it was perfect! If it plays the considerable majority, that is fine.

 

Yes, the NOAC you refer to is awful, the games do not play the same as the original, the characters move differently, look differently - bad.

 

You are right about the shoulder buttons - reworked it, tested, fine now. This forum doesn't allow me to edit my own posts (!) so re-upped pic below, corrected:

 

[cparsley]

No problem, happy to help with the side button mapping

[IntellivisionDude]

This looks cool. But not sure i could afford one myself.

 

Hope this isn't going off topic but what i'd love to have is like a NES controller modded just so i could use it for Burgertime. Ever since i found out my NES controller (that was modded to use on the 7800) had some response in my Intellivision II, i thought it would be cool to have one modded for my Intellivision. Sure it would probably be pointless to mod a controller just for 1 game, but it is my favorite game

I was using my NES controller to play Burgertime on my Colecovision and it so much easier to control.

[bacteria]

You can, that's the thing. The NES only has a D-pad with 4 buttons (A,B,Start,Select) so as long as your Intellivision game doesn't need more than 4 buttons to work, you can do this mod, quite easily actually. There is no PCB after all to worry about, just lots of wiring and some diodes.

 

Tell you what, if the above would map to the buttons you want for your Intellision and you want a modded NES controller for that, PM me, i'm sure we can make a deal...

 

Only thing is though of course, you'd have to open up your Intellivision to change the controllers; or make a plug for 9 connections so you could interchange the NES controller with the Intellivision one, or better still, split the cable so the official controller would still work but you can also use the NES one. Ideas, ideas!

[Rev]

if a new controller wont work only on a few games like vectron...i say oh well! i think most of us would be happy if it would work on most of the arcade style/popular games.

 

 

bacteria.what kind of controller would be best for this king of application? pc? dreamcast? arcade /fightingstick ? what about the overlays?

Edited June 29, 2010 by revolutionika

[cparsley]

with this, for the most part, you'd be going without most of the keypad... Unless your going to use, say a Jaguar controller to modify.

[bacteria]

Acually, no, you can use the same Intellivision controller!

 

If you notice, the tact switches on the circuit board I used is a perfect size - the buttons lay EXACTLY in the same positions as the game overlays, and of course are far easier to press than the membrane ones. The tacts can be covered easily with a printed piece of paper with the same overlay picture as the official Intellivision controller keypad, so will look the same, and of course, still use the game overlays too.

 

The d-pad only needs the tacts under the disc and that is it (apart from the wiring, thicker area under the keypad area, etc). The end result will look like an official Intv controller, only isn't

 

Illustration (from a broken up controller, but you get the idea):

 

 

Not played Vectron and can't see any YouTube videos on it, so can't comment on if this controller will play it or not.

 

The new keypad works great in games like Tron Deadly Discs, which after all need most of the keypad in gameplay to shoot a disc or block.

 

To all ends and purposes, with this mod, you get a controller that looks the same but works better. Only downside as mentioned is its primarily 4 way not 16 way on the controller, but be honest, I certainly have memories even when using the console in 1980/1981 that the controller often didn't go in the direction you intended... (so the 16 way d-pad wasn't THAT great - 8 way would have been far better IMHO). Games like Night Stalker, Lock'n'Chase Burgertime, Astrosmash, Space Armada, etc were only 4 direction anyway.

[jeremysart]

What I would like to see done, or do myself, is find a non working INTV or INTV controler, gut it, and replace the innards with a modern USB game pad that has 10 or more buttons.

 

You would be able to play most of the games, so long as at least the top and bottom side buttons, 1, 2, 5, and Enter are mapped.

 

How hard of a project do you think this would be?

[jaybird3rd]

What I would like to see done, or do myself, is find a non working INTV or INTV controler, gut it, and replace the innards with a modern USB game pad that has 10 or more buttons.

 

You would be able to play most of the games, so long as at least the top and bottom side buttons, 1, 2, 5, and Enter are mapped.

 

How hard of a project do you think this would be?

It would probably be easier to use an external USB adapter for the old controllers. Such an adapter has already been designed (I posted a link in another thread).

[Akito01]

I'm really curious to see if anything came out of this project, because ever since I installed an extra action button on my Atari trak ball controller, I've been thinking how nice it would be to do something similar with the Intellivision (keeping in mind I have an Intellivision II, you can see why I'd be motivated to do something about those action buttons).

 

Adapting a Jaguar controller looks like a cluster-copulation, and the inherent problem with using a Genesis pad is that you need a 5v line. Creating a custom 'action controller', or at least a supplemental right hand sub-controller for the three action buttons, seems like a really interesting solution. I have some questions, however, and I hope you'll forgive my open ignorance as I pose them.

 

http://www.buy.com/prod/1-foot-9-pin-serial-splitter-cable-db9-1m-2f-rs232/q/sellerid/24351715/loc/111/212435389.html

 

I'm assuming I can use this kind of splitter to connect a regular Intellivision II controller along with the custom controller, in order to handle keypad inputs or the control disc in conjunction with the custom controller. Is there likely to be any obvious issue with such a cable in this scenario that I'm not taking into account (other than the physical insertion into the Intellivision II, which I've already covered via a Genesis controller extension cable).

 

Even though it worked alright for the purposes of Missile Command on my trak ball controller, I wasn't overly impressed by the selection of buttons at my local Radio Shack. Any recommendations on what kind of buttons would be good to use, and where to order them?

 

I've never had to wire anything with diodes before now. What sort of diodes should I be picking up at Radio Shack, and how are they wired (they have an orientation, that much I know). In terms of wiring the buttons, I'm presuming that one line goes to the ground (pin 9), and the other end to the pins that activate the response you want the button to activate.

 

In terms of wiring to a connector, what is the recommendation? Do I get a regular 9-pin extension cable and simply chop off one of the connectors to access the bare wires? Are the wires easy to identify according to the pins they connect to?

 

I have an excess of Genesis 3 button pads. There are posts in this thread that imply you can adapt any already existing controller by simply remapping pins through this controller-adapter. Is this entirely true? It would be easier than creating a new controller from scratch, but I've been led to believe that without that 5v line, the Genesis controller isn't going to work.

 

Sorry to bring up an old post and annoy everyone with questions, but I'm quite earnest about this project and I admit to not knowing enough to be able to pull it off without appropriate feedback.

[bacteria]

Sure, something came of this project. I had to remove the matrix and make the buttons standard fare in order to work on my generic master controller; one controller that works on around 19 or so original console systems, all normally incompatible with each other. This part of the project wiring (with diodes) would allow you to remap the buttons onto a normal controller if you so desire, as you will have the unique button press, and of course grounding for the other connection.

[Akito01]

http://www.buy.com/prod/1-foot-9-pin-serial-splitter-cable-db9-2m-1f-rs232/q/loc/111/212435390.html

 

I made a mistake, as this is the correct splitter I assume I would need. Naturally, me being me, already placed an order for the wrong splitter, but put in a request to make the switch, so fingers crossed that all works out.

 

The Genesis controller is odd. I've found the pin out online. With just the ground, it looks like, at least in theory, that the four controller directions, the A button and Start button should be usable; the others depend on the 5v line. Using a multimeter, I wanted to see if it would respond as expected. I admit that my tests may not have been perfect (no diodes in use, for example), but they were consistent. With the Up and Down directions, it worked exactly as expected; pressing the control pad in that direct created a closed, measurable circuit. With Right and Left, there seemed to be a consistent, closed circuit regardless of whether the control pad is pressed or not. The A and Start buttons didn't seem to generate a response at all. The thing is, I don't know enough to know if any of this means anything at all; after all, the Genesis pad works fine in a 2600, despite the fact that the 5v line and the ground pins don't match (I think they're swapped?).

 

Well, I won't be able to say anything more until I get the cords and connectors I ordered, and then make a run to Radio Shack for some diodes and other odds and ends. It's good to know that this is all on the right track, and is not an unsolvable problem.

[Akito01]

I'm getting way ahead of myself, but I think I solved a big problem I was having with the Genesis controller. Having come across a web page that described how to convert a Genesis controller for use in a Vectrex, I realized that the only solution was to rip out the resistors and the IC from the inside of the Genesis controller. At that point, I followed the traces and soldered in some jumper wires to match the directions and buttons each to a specific pin (up and down on the pad, and the ground, didn't require any extra treatment since they were already directly wired). Perhaps those more clever than myself could have figured how to properly wire it from the inside to exactly match the Intellivision pin outputs, but as it is, I will still need to make an adapter to convert the pin out of this modified Genesis controller to the Intellivision.

 

Hopefully I can report back on greater success next week, but I think the bigger more interesting project (at least for myself) is coming up with a custom controller from scratch that incorporates the full keypad and diagonals. I realize I'm following in other people's footsteps, but other than this thread, I've actually seen very little in the way of custom Intellvision controller projects, and so it's hard to know exactly how to proceed in many cases.

[bacteria]

Problem you'll have with diagonals is what I found - you can use the pinout as per my diagram to get the keypad working no problem (using diodes or it doesn't work), the problem is that if you're using a normal d-pad, which is 4 directions, when you press diagonals it registers two of them, eg up and right to give diagonal up/right, which works for everything else, however, with the Intellivision, you don't get proper diagonals at all. The reason for that is the controller is set out as 16 way, as we all know (even though it didn't work too well in practice, even with a new console in 1980 as I recall), so when you press diagonal on the d-pad it uses the normal connection (wire 8 as I recall) which gives more like NNE or NEE as direction and not NE. What happens on the Intellivision is if you press NE it uses pin 9 instead of 8 (as I recall off memory). This means the only way to make true diagonals on a d-pad is to use a programmable chip/logic chip to say "If up and right pressed then use pin 9 else use pin 8". I don't know how to do this, if anyone does and can provide an easy circuit that works, please post it.

[Akito01]

Without experimenting, I can't say yet what will or won't work, and I am in the same boat in terms of being uncertain as the implementation. I should also say that I've also been using the pin out information on the Beeslife FAQ, which is pretty much the same, but they use a different numbering system for their pins. They have a matrix for the full 16 directions, and assuming that their numbers are correct, this is the kind of logic one would need:

 

 

In fact, you would only need one of these as opposed to four, because you could put the Down line on the same line going into the gate as Up, and the Left line with Right, and get exactly what you need -whether to activate pin 9 or not. There are ICs and diode solutions that provide an AND gate, but they also require a 5v line. The open question is if there is a way to create an AND gate circuit that doesn't require that kind of additional voltage line. I guess a really easy but somewhat stupid solution is to create a supplemental diagonal button that is wired to pin 9.

 

Because of the potential size of this still theoretical custom controller, I'm still undecided if it will feature a D-pad, joystick or a ring of press buttons. I can easily imagine something long the lines of the arcade stick controllers they currently sell for modern systems, which would offer enough surface area for the keypad, and all the necessary wiring without need for an additional adapter unit such as my Genesis pad is going to need.

 

 

Problem you'll have with diagonals is what I found - you can use the pinout as per my diagram to get the keypad working no problem (using diodes or it doesn't work), the problem is that if you're using a normal d-pad, which is 4 directions, when you press diagonals it registers two of them, eg up and right to give diagonal up/right, which works for everything else, however, with the Intellivision, you don't get proper diagonals at all. The reason for that is the controller is set out as 16 way, as we all know (even though it didn't work too well in practice, even with a new console in 1980 as I recall), so when you press diagonal on the d-pad it uses the normal connection (wire 8 as I recall) which gives more like NNE or NEE as direction and not NE. What happens on the Intellivision is if you press NE it uses pin 9 instead of 8 (as I recall off memory). This means the only way to make true diagonals on a d-pad is to use a programmable chip/logic chip to say "If up and right pressed then use pin 9 else use pin 8". I don't know how to do this, if anyone does and can provide an easy circuit that works, please post it.

[Akito01]

Now I'm a little excited. Step one has proven to be a complete success. I now have my 'action controller', which gives me four way movement and the three shoulder buttons mapped to my modified Genesis pad.

 

 

Me being me, in my original set up, I had the diodes oriented the wrong way. I still have to mount the adapter unit into something, but in all other ways, this thing is good to go. Once I get my splitter cable, this pad will be usable in conjunction with the normal Intellivision keypad for number entry. As it is, I can already play Mission X with comfort.

 

The next step is to see if I'm right or wrong about my assumptions about diagonals and the Beeslife pin out matrix.

[Akito01]

This will likely be my last post until I have something more physical to show off. As far as the Genesis pad goes, now that the adapter bit has been properly enclosed in a kind of box, that is done.

 

So, the matrix from Beeslife has proven correct, and activating the 9 pin goes give a 45 degree diagonal, as well as giving a slight diagonal to the four main directions. For the most part, diagonals have proven rarely to be necessary, so as with the Genesis pad, there is a strong argument for not bothering with diagonals at all.

 

Still, given the effort and time involved with creating some kind of controller more or less from scratch, one would like to include as much functionality as possible. There is no AND gate solution that I can find that doesn't involve additional voltage, so this is my compromise solution. Basically, both the Up and Down directions also output to the 9 pin. The end result are clean diagonals, completely straight directions Left and Right, and a slight diagonal on Up and Down. My thinking is that this will be perfectly functional for games like Deadly Discs or AD&D that offer a greater range of player movement. I also intend for this 9 pin line go through a two way switch, so that this diagonal functionality can be turned off if it somehow proves to be problematic.

 

Arcade parts are not terribly expensive and easy to acquire online, but they all tend to be of one size, which is a minor bummer. The real challenge is going to be just the physical mounting, because unless I can find something to retrofit, I guess I'm going to have to resort to some woodwork.

[bacteria]

Given the choice of proper 4 way directions and not good diagonals or good diagonals but not so good 4 way directions, i'd have thought proper 4 way would be preferred? Problem is of course that games will be unplayable in either combination that need more sensitive movements but then those types of games were hit and miss on accuracy anyway.

 

What are the AND logic gates you saw? Other option is a proper 8 way digital joystick, however the only small one I know of is the one used on the GP2x which is far from good anyway as the directions are not proportioned properly.

[Akito01]

I'm glad I went ahead and experimented a bit more before I answered, because as it turns out, the whole diagonals thing is more complicated than I first expected. I went ahead and built a test circuit, using the modified Genesis pad, and an adapter that mapped the Up and Down directions to a switch that was matched to Pin 9 on the Intellvision input. This part of it all worked as expected. I was curious to see which games used all 16 directions and which didn't. Utopia and Deadly Discs do, while AD&D only uses 8.

 

But, this wasn't the puzzling thing. When I had Pin 9 switched on, for some reason I lost functionality of the 7,8 and 9 buttons on the keypad. I tried this on both Deadly Discs and AD&D with the same result. This was the case regardless of any other input that was going on. I have no idea why this should be the case, since those keypad buttons have nothing to do with the 9 pin, and with no directions being pressed, there shouldn't be any input into the 9 pin anyway. Switching off the 9 pin, everything went back to normal; the D pad responded with the correct four directions, and the keypad worked normally. My soldering work isn't the best, but I can definitely say that there are no bridges going on, so I'm stumped.

 

For now, I've given up on the idea of incorporating a diagonal switch. Too much trouble unless I can work out the keypad anomaly.

 

The diode based AND gate looks like this:

 

 

I haven't experimented with this kind of circuit, but since there isn't a 5v line to use coming out of the Intellivision controller port, it wouldn't be usable anyway.

 

As you say, there really aren't any proper 8-way joysticks, which makes sense since for any greater degree of control one would probably opt for an analog stick anyway. I don't know enough about the internals of analog sticks to know if they would be adaptable or not.

 

Given the choice of proper 4 way directions and not good diagonals or good diagonals but not so good 4 way directions, i'd have thought proper 4 way would be preferred? Problem is of course that games will be unplayable in either combination that need more sensitive movements but then those types of games were hit and miss on accuracy anyway.

 

What are the AND logic gates you saw? Other option is a proper 8 way digital joystick, however the only small one I know of is the one used on the GP2x which is far from good anyway as the directions are not proportioned properly.

[bacteria]

I don't see how the diagram could activate one connection (pin 9) instead of another one (pin 1), perhaps elaborate? Connections are by the nature hard wired. If it works, great, I want to use it.

 

No issue getting 5v, just put three watch batteries in there, would the circuit work off 4.5v instead of 5v or does it need 5v?

 

Interesting find about buttons 7,8,9 - perhaps that kills the idea of proper diagonals?

 

Tron Deadly Discs works fine as 4 way, actually, the game is far easier to play especially when lining up to kill the repair robot - quite easy now.

 

I think the only games that would suffer in real terms are the likes of Skiing as many Intellivision games are 4 way anyway or having poor diagonals don't matter. That's my findings anyway.

 

Interested to hear your further findings!

[+5-11under]

If it's helpful, the ColecoVision Super Action Controller has sort of an 8 way joystick. The diagonals have two switches each, physically separate from the main four. For instance, if you press the joystick in the NE direction, the N and E switches do not activate, but two other switches do activate. For the ColecoVision, these switches are of course in parallel with the N and E switches, so it works like a 4 way. However, if you were to properly hack the controller, the diagonals could be separated from the horizontal and vertical switches.

[Akito01]

http://www.hanssummers.com/electronics/misc/gates/index.htm

 

This is one link I found when trying to find circuits for AND gates. It explains the principle, certainly better than I can, since I found myself having the same reaction as yourself. Still, this circuit diagram comes up so often that I have to believe it performs as promised.

 

I should probably draw this out, but to elaborate on my initial thoughts, the purpose isn't to switch outputs from one pin to another, but to switch on a third pin (pin 9) if two other pins are turned on (Up and Right, or Down and Left, or example). The outputs from Up and Down are essentially split -one line goes to the pin that activates that direction in the Intellivision, and the other line would form one of the inputs into the AND gate that would be matched with inputs from Right and Left to determine in pin 9 should be activated as well.

 

I still don't understand why my switch solution caused this unintended affect with the keypad. I've gone over it and rechecked it, and my test unit is sound in terms of construction. Something else is going on that I'm missing. Perhaps it relates in some way to why the directional disc and keypad cannot be used simultaneously; it makes sense looking at the pin out matrix, but I don't know why it would affect those three keys on the keypad and not the others, or why it happens even when no directions are being pressed at all. I can't say much more without mucking around further.

 

The next step will be putting together a test keypad, which will be the most complex bit of wiring for this

controller, and I want to be sure I know what I'm doing before investing further in this project.

 

 

I don't see how the diagram could activate one connection (pin 9) instead of another one (pin 1), perhaps elaborate? Connections are by the nature hard wired. If it works, great, I want to use it.

 

No issue getting 5v, just put three watch batteries in there, would the circuit work off 4.5v instead of 5v or does it need 5v?

 

Interesting find about buttons 7,8,9 - perhaps that kills the idea of proper diagonals?

 

Tron Deadly Discs works fine as 4 way, actually, the game is far easier to play especially when lining up to kill the repair robot - quite easy now.

 

I think the only games that would suffer in real terms are the likes of Skiing as many Intellivision games are 4 way anyway or having poor diagonals don't matter. That's my findings anyway.

 

Interested to hear your further findings!