Rapid-fire using a keyboard encoder???

[jhsmith321]

Does anybody know if it is possible to add the auto fire or rapid-fire options using a keyboard encoding? Is there any encoder that supports this function? Or is this only possible by using a hacked joystick with the built-in function??? Thanks!!!

[u_rebelscum]

http://www.groovygamegear.com/Page3.html#5

edit:
according to Heli-Tiger, this is just the standard keyboard repeat.  DirectX, and thus mame, does use this repeat, so this won't help you.  Sorry about the mistake.

[jhsmith321]

http://www.groovygamegear.com/Page3.html#5

Thanks for the link!!!!  By the way, does anyone knows if this encoder (keywiz) is better compared to the ipac, hagstrom or the mk64? and why is so???

[Tiger-Heli]

http://www.groovygamegear.com/Page3.html#5

Thanks for the link!!!!  By the way, does anyone knows if this encoder (keywiz) is better compared to the ipac, hagstrom or the mk64? and why is so???

Check previous threads - been discussed a LOT before.

MK64 is double the number of inputs and double the price.

Overall, I prefer the KeyWiz to the I-PAC, and I like getting four more inputs for less money.

The I-PAC has a keyboard pass-thru, supports USB and supports keyboard LED's if these are important to you.

[Tiger-Heli]

http://www.groovygamegear.com/Page3.html#5

It won't work the same, though!  I play Tiger-Heli a lot (could you guess).  I used to use a cheap gameport joystick with a rapid fire switch option. It worked in MAME.  Now I use a cheaper gameport joystick without an autofire button.

However, if I just use the keyboard and hold down the fire button, it shoots about twice a second.  This is also the way the actual arcade game worked.  This is with a standard keyboard (not an encoder) so key repeat is enabled.

You would need a special circuit to make the encoder do what you want.  It seems like I saw a thread or maybe one of Stephan Hans' pages talking about it, but I can't remember. . .

[rampy]

Instead of futzin up hardware... is there a cheat option in MAME that would be more appropriate?

Just a passing thought.

Rampy

[u_rebelscum]

http://www.groovygamegear.com/Page3.html#5

It won't work the same, though!  I play Tiger-Heli a lot (could you guess).  I used to use a cheap gameport joystick with a rapid fire switch option. It worked in MAME.  Now I use a cheaper gameport joystick without an autofire button.

However, if I just use the keyboard and hold down the fire button, it shoots about twice a second.  This is also the way the actual arcade game worked.  This is with a standard keyboard (not an encoder) so key repeat is enabled.

You would need a special circuit to make the encoder do what you want.  It seems like I saw a thread or maybe one of Stephan Hans' pages talking about it, but I can't remember. . .

Oh, you mean keywiz is talking about the standard win32 API keyrepeat that directX ignores, so mame ignores too?  Opps, my bad.

[grafixmonkey]

I can give you a schematic that will give you adjustible-rate rapid fire on a button, in the hardware itself, but it'll be more wiring than you're probably used to or want to do just to get autofire (involves a couple chips, capacitors, and pots).  You could probably do all the wiring in an hour, and then stick the circuit between your button and your IPAC.

If you want to actually build it, then PM me

[grafixmonkey]

Someone did message me asking for a schematic, so I'm in the process of making one.  Problem is, the datasheets for the rapid-fire chip don't say whether the output pin is high or low when it's disabled, so I have to build it before posting the schematic.

However, I can say that the final design will be able to fit two rapid-fire buttons on each chip, and may require a second chip that can handle 6 buttons (i.e. 3 rapid fire chips to 1 of the second) so to make all of your Street Fighter 2 buttons rapid-fire you'd need 3 or 4 chips (3-4 hours of wiring!) and some cheap extra components.  Making six buttons rapid fire would cost approximately $5-ish if you get parts from a component supply and not Radio Shack.  Add the cost of the board you put them on and it's $12-ish for soldering them on and $30-ish for using a protoboard (faster, easier, changeable, so you can test before it's final or add features later).

[1UP]

I'm interested in this.  Keep us posted when you have a circuit!

[jhsmith321]

Thanks grafixmonkey!!! as 1UP said, please keep us posted when you have the circuit or the documentation, that would help us all a LOT!!!

[grafixmonkey]

Ok, it's done!

It uses the 74629 chip, which can be replaced with a 74124 chip changing two pins from ground to 5v.  (the 74629 is newer, and it may be hard to find a 74124).  You can fit two buttons per chip.  If you need to use a 74629, the pins of the chip colored GREEN should be wired to ground and not 5v.  These pins control the frequency range you can control, and the relationship is in reverse for the older chip (ground means tiny range instead of largest range).

I recommend getting the flavor of chip marked "74LS629".  This stands for low power consumption and usually has been made more recently.  See notes about 7400 chips at bottom.

If you only have one button on the chip, then the pin colored RED must be wired to 5v and the pin colored BLUE must be wired to ground, and all wire connections colored blue can be ignored.  Even if you only have one button on the chip, the second capacitor should be in place.  If it's not there or the red and blue pins aren't wired to 5v and ground, then you might get a "crosstalk" from the unused half of the chip at an unknown frequency.

When this is built, the frequency of the rapid-fire can be controlled using the frequency control knobs (pots) so it might be nice to add these to an accessible part of the control panel.  You can also test the circuit before attaching it to your IPAC by touching an LED between ground and the wire labeled "Blue Button --> IPAC" or "Orange Button --> IPAC".  Then the LED should remain on when the button is not pressed, and when the button is pressed and held the LED should instantly go off (no lag between button press and first rapid-fire tap) and begin flashing.

A word of caution because I know someone will try to do this:  YOU MUST NOT PUT OTHER BUTTONS IN PARALLEL WITH THIS CIRCUIT ON THE IPAC.  Doing this with the design as it is would cause the chip to pump out as much current as it can trying to drive the ground wire to 5 volts, every time the parallel switch is pressed.  To make other buttons operate the same IPAC pin, a second chip can be used, or a bypass switch can be installed.  This circuit can also be modified (with an extra chip per every four buttons) so that a switch turns rapidfire on or off for any group of buttons.




A note about 7400 chips:  this whole library of logic chips starts with the numbers 74, and are often called "7400 chips".  In the case of the 74124, the 124 part labels it as a "Dual Voltage-Controlled Oscillator" and if you find a Texas Instruments databook and look up 74124 you get all the details on the chip.  Each specific 7400 chip comes in flavors, such as "74124" and "74LS124" and "74124A" and "74LS124A", and usually have the letters 'SN' at the beginning (i.e. "SN74LS124A" or "SN74124").  These extra letters change details like how much power the chip draws, or how much current its pins can take in or pump out, but don't change the function.

[grafixmonkey]

Thought you might like to know what it looked like when I was done building it.  Here are some pics.

You can see the logic probe reading a 'high' signal when the button is up, and a nice shot where the camera exposure caught the probe switching from low to high and back with the working rapid-fire.

You can also see what will soon be my new 4-player cab, my hacked AT computer power supply that I use as a voltage source, and my dinner which was very yummy.

Notice the picture on the bottom where you can see the component I used as a voltage control knob.  This is the style where you would tune it once, and leave it in the cabinet.  You may want one that's accessible on the control panel.  Because this circuit is guaranteed to fire once as soon as the button is pressed (and then keep rapid-firing), you can tune this to quite low frequencies of rapid-fire without getting lag.

[grafixmonkey]

One final note.  When you turn this circuit on for the first time, put a finger on top of the chip while you turn the switch on.  If you notice the chip heating up rapidly, turn the switch off immediately and check to make sure that pin 16 is on 5v and pin 8 is on ground and not the reverse!  

It is very common even for people in electrical engineering to accidentally reverse the power to the chip, and if you do this the chip will quickly get hot enough to melt itself, the protoboard it's in, and anything touching it, including your finger when you try to figure out what that smell is.  

The chip might not function properly after being reverse-powered, but if you turn the switch off quickly the chances are better.

[grafixmonkey]

More drama.  I tried to connect the IPAC and am having difficulty finding a 5v source on the board.  The LED header seems the best source for it, but the pinout on the Ultimarc web site does not match what's on my IPAC4.

Testing, hoping to god I don't see anything start smoking...

[grafixmonkey]

And the beast works!!

I love that feeling.

I had trouble because I didn't have a power supply inside the PC the IPAC was being plugged into that I could tap.  (MAME pc is out of commission   )   So I used the ground pin on the IPAC's array of button contacts, and a 5V pin I found on the LED header, and it works great!

The circuit draws 40mA though, so no more than two of these should be plugged in to the IPAC itself or I'd be worried something bad would happen to the PC's usb port.  Tapping ground and 5v wires in the computer's power supply really should work though.

Here's a nice little action shot of it working!

[jhsmith321]

THANK YOU SO MUCH!!!!!!

[1UP]

One question: how easy would it be to put in a toggle switch to switch the circuit between rapid fire and normal (bypassing the circuit completely)?

Also, it doesn't look like you've marked the polarity for the capacitors in your schematic...

Thanks

[grafixmonkey]

polarity doesn't matter here, if it ever does (never was clear on why certain kinds of caps had a polarity), and it's easy to build a toggle.  What kind of toggle though?  you could do:

* flip a switch under the cab that flips one button between rapid fire and normal,

* flip a switch under the cab that turns a whole bank of buttons rapid fire or normal

* have a pushbutton on the control panel (possibly illuminated) where tapping the switch would toggle a bank of buttons between rapid fire and normal (and the light could go on for rapid fire mode)


those are listed in order of increasing wiring difficulty, the third being a good 1-3 hours for someone decent at wiring chips to assemble turbo-toggles for a two-player street fighter 2 layout.

--EDIT--
Scratch that!  I just remembered a chip that makes the third option just  as easy as the second.  And you'd do as much wiring to put an individual switch on each button too, so all the options are about the same time-wise.
--/EDIT--

Anyone know why capacitors have a polarity?  because I sure don't.  Every theory I've learned about them shows them as a completely symmetrical device, like resistors.  My EE classes didn't even mention the polarity of caps, I found out they had polarity when I went to the EE supply store once and bought some.

[grafixmonkey]

Did you guys want me to post a circuit for the on/off toggle or no?

if so tell me which option you're interested in, cuz I don't really want to design all three.

[1UP]

Did you guys want me to post a circuit for the on/off toggle or no?

if so tell me which option you're interested in, cuz I don't really want to design all three.

I really only want one rapid-fire per player, so I was thinking more like having an on-off pushbutton on the panel that would let you turn the turbo on right there.

I think that some capacitors are polarized.  The ceramic disk-type caps are not, but the metallic barrel-shaped (electrolytic?) caps all seem to be polarized, and usually have a lead marked as positive.  I remember as a kid, I used to play with a large capacitor by charging it (holding the leads to a 9-volt battery) for a few seconds, then touching the leads together to create a spark.  I remember that it only charged the cap if I put the positive lead on the right terminal on the battery.

[grafixmonkey]

interesting.  I'll see if mine are like that, but they always seemed to charge both ways even though they were polarized.  99% of the time capacitors are for modifying a signal rather than storing charge, and their voltage will oscillate positive and negative (like in the rapid-fire circuit).

In theory capacitors are just plates of metal separated by a dialectric material, in some geometry that makes them fit in a small enough package.  Maybe some stronger dialectrics function better in one direction than another or something.  Now I'm really going to start trying to find out why they're marked like that.

So you want a pushbutton to tap-on, tap-off the rapid fire?  I'll adapt the circuit from the "shift key" topic.  Since it does the tap-on, tap-off part, all that remains is to have that control a mux for each button you want that to turn on/off.  

So, it ends up being one chip for every two tap-on tap-off buttons, one chip for every two rapid-firing buttons, and a second chip for every four rapid-firing buttons.  Not too bad for wiring, even for rapid fire on a whole 2-player 6-button layout.

[1UP]

Also, what's a good value for the frequency control pots?

[grafixmonkey]

I've got this thing almost done for you guys...   Got the toggle-button circuit toggling just now!

The frequency control pots just have to be any variable resistor that can be hooked up to 5 volts without drawing more power than it's rated to handle.  power = V^2 / R = 25/R in this case, so if you have a 1/4 watt pot, then its resistance must be at least 100 ohms (that would get warm, but would not light on fire or burn out.)  For power consumption reasons I recommend at least 1,000 ohms (1 Kohm), probably ideally in the 3 Kohm to 5 Kohm range, but the only difference is the amount of power it will waste.  (don't go above 6 or 7 Kohm! then you'll have trouble getting current to the chip at all.)

Of the three kinds of knob I showed on the circuit diagram, the first kind has a little twisty-knob post that you can mount to a panel and put a nice little knob on.  The second gives you extremely accurate control over your firing rate, but the little flathead screw must be turned dozens and dozens of times to go from one end of the range to the other, and is kind of flimsy compared to the other two.  The third is my favorite, a simple phillips screw that you turn about 270 degrees to go from 0 volts to 5 volts, and it's sturdy and low profile.  But if that screw turns itself half a degree or so without your permission, your firing rate will noticeably change.  

Now for updated project pics!  (am I posting too many pics? I don't have any web hosting...  you can ---smurfette--- at me if I am...)

The little filmstrip action on the right shows me tapping the button and the thing toggling between each shot.  (the big one is so you can see what you're getting into!  )  Instead of the red light going on and off, I wanted to have the illuminated pushbutton light going on and off, but I can't find my power transistors and had to use that little tiny transistor you see.  (little bulbous metal package w/ three pins).

[grafixmonkey]

It's finished, here's a pic - I'll turn it into a diagram for you later.  Don't worry it's not nearly as complicated as it looks.

#edit#
Oh yeah and the LED is washed out in the pic so you have to look close at the bottom two.  (Didn't feel like playing with light and exposure.)  Look near the little blue box thing, that's where the LED is (the one that was shown toggling on and off in the previous post.)  You can see in the top pic, my finger is off the "pushbutton" and the signal going to the IPAC is high.  In the middle pic the toggle LED is on, and I hold down the pushbutton, and I get rapid-fire.  Then I toggle it in the bottom one, and the LED is off, and I get a straight low signal going to the IPAC when I hold down the pushbutton.  

You know what, this'd be fun to add to an insert-coin button!    

[1UP]

Thanks, I'm looking forward to a schematic!

[grafixmonkey]

And it's here!  

Every chip in this circuit has capacity for more than the 'minimum' use (one toggle, one rapid-fire button)  so I tried to show how one would go about adding the second, third, etc. parts of each one.  The whole thing also consists of three major component - a vibrator (not that kind), a toggler, and a selector.  (figures B, A, and C respectively).  Rather than have a huge mess in a single circuit and try to exlpain how one would add the unused parts of the modules, I divided the diagram into those parts and marked which wires from one connect into another.

Diagram A:  toggler.  A single chip can handle two toggle buttons.  (one for each player?)  I showed both buttons hooked up, and the two wires (TURBO #1 and TURBO #2) that would go to two different selector circuits (figure C).  

Diagram C:  selector.  Each chip in figure C can only select one way at a time for all four sets of inputs.  It works exactly like a four-pole double-throw relay, or four double-throw relays in parallel.  So to have two players each able to turn on/off rapid-fire independently, you need two of these.  Each chip can handle four buttons at once.  The buttons with a green shaded background would require a second Diagram B circuit to provide their inputs.

Diagram B:  You've seen this one already.  There's a minor change.  In addition to the two rapid-fire outputs, there are two more outputs that come straight from the buttons before they are made rapid-fire.  The selector circuit chooses between the rapid-fire signal and the normal signal, based on what the toggler circuit gives it.

[grafixmonkey]

And here's a block diagram example of how you would use this to put together a two-player street fighter 2 panel with each player having a toggle that turns all of their buttons on or off of rapid-fire.

You should be able to figure out how to make any combination of rapid-fire buttons from this.

[1UP]

Holy crap!  That's a lot of wiring...  Why the second chip just to switch from turbo to normal?  Isn't it possible to just use a dual-throw switch of some kind to route the IPAC directly to the pushbutton rather than to the auto-fire chip?

Also, you said that when you press the button to fire, the logic state goes from high to low, but doesn't the IPAC work the other way around?  i.e. no signal is sent while the button is not pressed, then a signal is sent when the button is pressed?  I'm confused!  

[grafixmonkey]

wow this one really went long...  bear with me

The IPAC reads a button press event when the voltage on one of its pins goes to zero.  That's why you have a ground loop between all your buttons, with the other side of the switch going to the IPAC.  When you close the switch it grounds an ipac pin, and when you release a resistor on its pcb pulls the pin back to 5v.  I know, it's so tempting to just assign anything that's "activated" or "good" or "positive" to logic 1.     I did it too when I submitted diagram A as a shift-button toggle back a ways.

You can use a double-throw switch in place of diagrams A and C, yes.  You'd just put the center pole to the IPAC, and put the signals RF#1 and Button#1 to the other two.

The chips have the advantages of being a button you can smack just like everything else on the panel, and being able to control several buttons with one switch which you can't do with a double-throw.

Or you could replace diagram A with a double throw switch, running the center pole to diagram C as the signal labeled TURBO, and the other two poles going to 5v and ground, and that would also control several buttons at once.  (but would not be a button you can smack     I guess I just like smacking buttons!)

It might also be simplified by using a single oscillator (B) that oscillates all the time, and connecting its output pin to the Com pin of all the buttons you want to rapid-fire.  But two things would happen:  the chip would have difficulty driving more than 4 buttons (too much fanout) and would need a helper chip for this, and you would not be guaranteed a button-press the instant you tap the button as you are here.  (each button would wait to press until the oscillator went low again, when they would all press at once.)

You saw the picture I posted of the circuit - that protoboard had one each of circuits A B and C.  It's not really all that bad.  The 6-button street fighter layout would be though.  It was just an example.


What's great about these circuits is that each has a modular function - you can use circuit A to toggle just about anything, diagram C to switch between any two things, and diagram B to make anything...   um...  vibrate.  For example, you could have two separate, independent button configurations - say for one config, the center two joysticks of a 4-player board would be players 1 and 2, and the outer two wouldn't be connected, and for a second config they could be labeled 1,2,3,4 in order left to right - you can accomplish this using circuit C, without needing the shift function.  In which case, you could have four configs you can toggle between by flipping the shift key and your switch in different combos.  Or you could cause diagram B to make a deep "BeyYOOooooo" over a speaker every time a coin is inserted (with a transistor and capacitor in place of its voltage knob.)

[1UP]

OK, so I can skip the second chip if I only want 2 auto-fire buttons?  That's pretty much what I was thinking of in the first place.  I used to have a NES arcade stick that had a really cool setup:  Buttons A and B each had a separate "pop-up" button to turn on the turbo fire, and also a separate knob to control rate of fire.  You could still slap the buttons to toggle the autofire, but they were basically toggles themselves, the type that stay down the first time you hit them, or stay up when you hit them again (which is good because you can tell whether turbo is on or off by looking at the button).  That came in handy with Super Mario Bros, just set the jump button at the optimum frequency, then when you get to a coin box, slap the turbo button, then hold down the jump button and drain that box dry!  

[grafixmonkey]

haha, yeah those were the days.  I had that same controller, too.  It's sitting in my living room right now.  Level gaining in Rygar was my favorite flagrant abuse of turbo.  My brother and I found a way to prop a 2x4 and a weight against the joystick button, holding down turbo fire forever.  Then we'd sit poor old Rygar down in a corner where enemies only came from one direction, and leave him there all night.  When we came back, he'd be all beefed up for some spikey-Yo-Yo ass whoopin'.

Yeah, you can skip all but the circuit with the 74629 in it (i.e. skip making A and C, just make B), and use one double-throw switch per player, to switch the button between rapid-fire and normal.  If you can find one of those pushbuttons that stays down until you tap it again, you can use one of those, but my guess is you'd be stuck with a two-way switch to flick.

[Dav]

OK, so I can skip the second chip if I only want 2 auto-fire buttons?  That's pretty much what I was thinking of in the first place.  I used to have a NES arcade stick that had a really cool setup:  Buttons A and B each had a separate "pop-up" button to turn on the turbo fire, and also a separate knob to control rate of fire.  You could still slap the buttons to toggle the autofire, but they were basically toggles themselves, the type that stay down the first time you hit them, or stay up when you hit them again (which is good because you can tell whether turbo is on or off by looking at the button).  That came in handy with Super Mario Bros, just set the jump button at the optimum frequency, then when you get to a coin box, slap the turbo button, then hold down the jump button and drain that box dry!  

I did one a long time ago with a 555 chip and ran the output to a 4066.  You could pulse up to 4 buttons that way.  It's a very simple circuit,takes about 4 other components, you can find the plans on the net.  I think I used the datasheet for the 555 to get the plans.  It's not any easier to make that this design, but you can get all the parts at radio shack.

[grafixmonkey]

ok since people are going on about rapid fire again, I'm going to bring this thread to the front of the list again, and add something:

I just saw a 4-pole double-throw relay at Radio Shi...  I mean Radio Shack.  This could be used to switch four buttons between normal and rapid-fire without the need for diagrams A or C.  Diagram A would become a switch you'd have to flick back and forth, diagram C would become the 4-pole relay, leaving only diagram B to be built.

Also, I thought about it and you really would only need to build one or two of diagram B - if you attach its output to the COM output of all the buttons you need to be rapid-fire, I believe it would work ok.  It would need amplification in the form of a line driver though.  If there's anyone still interested in this, reply and I'll show how to add the line driver and run multiple buttons with the relay idea.

[GamingGreg]

grafixmonkey,  I would be interested in seeing the circuit you described.  I'd like to have that very thing in my cabinet .

Also, does anyone know if the MAME developers have though about implementing "rapid-fire" in software?  Since MAME emulates the hardware inputs to the ROMs, it seems like this could be done fairly easily.

[u_rebelscum]

grafixmonkey,  I would be interested in seeing the circuit you described.  I'd like to have that very thing in my cabinet .

Also, does anyone know if the MAME developers have though about implementing "rapid-fire" in software?  Since MAME emulates the hardware inputs to the ROMs, it seems like this could be done fairly easily.

It's been done is some derivatives, but the MameDev line on this has been:

Mame is about accurate emulation.  Adding rapid fire not only isn't emulation (that would a simulation instead), but it also would add a "feature" that was not in the original game, making mame so it's not as accurate as without rapid fire.  Therefore, official mame will not add rapid fire.

Of course, I am not on the MameDev team, and I'm just relaying what has been said.  FWIW, physically challeged people has asked for rapid fire before, and have received basically the above responce.  MameDev might change their mind about this, but I doubt it for the near future.  Until then, you'll need to find a derivative that does include rapid fire (I don't know which ones do).

[grafixmonkey]

I don't think Mame should have rapid fire in it either.  Mostly because it's a project for reproducing and documenting the original hardware for posterity.  Also because games assume it won't be there, and many take holding a button down to mean something useful in the game.  Also because most of the rest will auto-fire for you if you hold the button down anyway.  It's not all that useful a thing to have, so I'd rather them work on whatever other cool stuff they're working on.  

Custom versions of Mame which have been modified with rapid fire and then recompiled and re-released, however...   Feel free!

And the circuit is above, in this thread.  Diagrams, and pictures.  Unless you're speaking of the relay version.  I can post that later today.