2-4-8-way switcher for 8-way-joysticks

[Luigi]

Hi.


You all might know the problems that occur when you use an 8-Way-Stick for 4-Way-Games. If you don't push the stick exactly in one direction and two microswitches (or leafs) are pushed the same time, the game character might stay where it is instead of climbing up a ladder (Donkey Kong) or go down (Pac Man and other maze games). The switching mechanism is often not that easy to use....most of them have to be switched from under the panel.

So....I just wanted to point you to this exciting piece for arcade cabs.



Two guys from a german arcadeforum developed a pcb that is installed between up to two 8-way-joystick. I don't know exactly how it works...but it works!! I could test a prototype on our last Party (Nov. 2009) and it was a big difference playing with 8-way-mode or 4-way-mode.

The mode can be switched with a button and a RGB-LED shows the actual setting. They did a test setup where a 7-segment display showed the mode....I gotta have to find the video they made and find out how they did it  

There are the following modes available: 2-way vertical, 2-way horizontal, 4-way, 8-way


The PCB kit (just the pcb with parts....not assembled!!!) can exclusively be ordered at

www.arcadeshop.de

LINK

Assembly instructions: http://www.arcadeshop.de/images/specs/elimitator_install_en.pdf

No, I'm not getting money from the sales nor am I one of the developpers. I just like it and wanted to share with you  

[boomstick]

Looks like an exceptionally usefull piece of kit.  May well have to order one to see what its like.

Nice find

[Ginsu Victim]

Ok....what exactly does this do?

[boomstick]

Ok....what exactly does this do?

From what I gather, you install this PCB between your P1&2 8-Way joysticks and your IPAC / Keyboard encoder.  The PCB has a connection for a microswitch that is used to change modes.  An RGB LED indicates the selected mode.  The Microcontroller filteres the inputs recieved from the joystick(s) and based on the selected mode, (2-way vertical, 2-way horizontal, 4-way, 8-way) passes cleaned signals to the IPAC.  

For example, if in 4way mode and the 8way stick hits Up & Right the microcontroller filters the input sending only Up or only Right.  Quite how it tells what was intended I'm not sure,   I guess it determines which microswith was activated first and sends that to the IPAC.

It really sounds quite usefull if it works, would save the need to clutter up your CP with dedicated sticks or building modular panels. It might not be for everyone, but I am more than tempted to order one and see.

Luigi, do you actually have one of these PCBs?  It would be interesting if you could provide some playtest info.

[Bender]

I wonder if this would work for an xbox or PS3 controller hack
I'm so spoiled with my U360's on my cab, that I miss not being able to switch modes when I use my console stick
To use my SFIV stick with Pacman Championship would be so cool

some one needs to review this thing
Seems very interesting

[bkenobi]

Maybe I'm just confused, but this is a bit of hardware that's sole purpose is to take the input from the joystick and algorithmically determine whether the stick should respond from your inputs. 

If you are looking for something to fix MAME, then why not just fix the MAME code instead?  MAME can be coded to do anything this can and probably more (since it's not running on a low powered controller). 

If this is for other systems, this makes perfect sense.  There are no other ways to do this type of thing outside of MAME that I'm aware of.

[:oj]

If you are looking for something to fix MAME, then why not just fix the MAME code instead?  MAME can be coded to do anything this can and probably more (since it's not running on a low powered controller).

You can use it with MAME, but the intended purpose was for standalone arcade systems (as long as the used joystick is digital and 8-way). The E-Limitator will be interposed between your joystick and a gaming PCB or an interface to whatever you want (XBox, Playstation, PC etc) and this makes this motion filter all-purpose (excepting analog and optical joysticks).

[CheffoJeffo]

You know, I think that's pretty damned cool -- certainly a nice solution for 48-in-1s and 60-in-1s.

 

[bkenobi]

You can use it with MAME, but the intended purpose was for standalone arcade systems (as long as the used joystick is digital and 8-way). The E-Limitator will be interposed between your joystick and a gaming PCB or an interface to whatever you want (XBox, Playstation, PC etc) and this makes this motion filter all-purpose (excepting analog and optical joysticks).

Ahhh, ok.  That makes more sense!

[Bender]

Oj, thanks for the clarification

are you involved in the making of this device?

[CheffoJeffo]

are you involved in the making of this device?

No, I'm not getting money from the sales nor am I one of the developpers. I just like it and wanted to share with you  

[:oj]

@Jeffo: Guess he meant me ;-)

I'm only the German supplier of this smart thing. It was developed by 2 other German guys, Bencao and Punkrockcaveman. All I did was spending the money for the very first production of a larger batch.

And Luigi knows this project, because he arranges arcade parties twice a year. Luigi and his mate Cpt. Kirk own a hall with lots of vintage (dedicated and a few universal) cabs. This is the place for even this parties, where many fans of arcade games meet each other, talking and gaming together and sometimes handicrafts will be introduced.

[CheffoJeffo]

@Jeffo: Guess he meant me ;-)

Got me ... 

[saurian333]

Maybe I'm just confused, but this is a bit of hardware that's sole purpose is to take the input from the joystick and algorithmically determine whether the stick should respond from your inputs. 

If you are looking for something to fix MAME, then why not just fix the MAME code instead?  MAME can be coded to do anything this can and probably more (since it's not running on a low powered controller).

Surely you are correct, and this would be possible to do in MAME.  However, that's more overhead for your CPU (not much, granted).  The more low-level you can process an input, the better, and it doesn't get much lower than simply filtering the original electronic signal.  The only level lower than that is to physically prevent the switches from being pressed, as with a restrictor plate.

I'm all for modding the code, of course, and I'd love to see that done (might take a crack at it myself, actually; that probably wouldn't take a very complicated algorithm).  Just pointing out that hardware-based solutions are almost always faster as well as more reliable.  In this case, it probably wouldn't be too noticeable on modern systems.

[bencao74]

hey,

 here're some pics from punkrockcaveman bartop setup. First post, hello to everyone!








http://bencao74.blogspot.com/2010/01/e-limitator.html

Bye,

Bencao

[purplearms]

I'll vouch for Bencao74 and this product. He's a solid modder and this would be tremendously useful for MAME players who want to keep flexibility of their joystick without having to physically restrict their joysticks with gates etc... 

[Bender]

MAME already does this for analog sticks don't know how it handles digital
but no other emulators or consoles can do that analog or digital
I can't wait to but one on my xbox360 stick
Still would like a review from someone here, or Maybe I'lll just bite the bullet and do one

Aghhh I just thought of something else this would ROCK for Tron: 4 way lightcycles and tanks, 8way for spiders and cones   just a press of a button in between!

EDIT: I bought a couple. I'll post my reactions when I get one set up

[Franco B]

Ahh so thats what it is. 

I had seen the teaser on the arcadeshop.de front page and I was wondering what it was.

[z]

Something I've been curious about is if anyone is doing this same action with a SPTT (single pole triple throw) switch.  You could also use a rotary switch taken out of an old printer or serial port switch box or even three SPST toggle switches.  To help here are a few drawings of the idea, typical application would be a Happs' Super with actuator set for 8 way action:
The first drawing is basic joystick wiring for joystick with cherry microswitches (sorry, haven't figured out a way to do this with leaf switches).  If the joystick physically allows 8 way action then that is what the output will be.

Second drawing is the idea of making the 8 way joystick only report one direction at a time by making the ground conditional.  Example:  With this wiring if the UP button is closed then the ground is removed from the remaining three switches so they can not send a signal.  Even if UP and LEFT are both closed the only signal sent will be UP, since there is no ground reaching the LEFT switch no signal can be sent from it.  If DOWN and RIGHT are both closed the only signal that will be sent is DOWN, again there is no ground reaching the RIGHT switch so no signal will be sent.

Third drawing is the arrangement similar to above except the triple pole switch has been added as a way to optionally provide ground to the three switches (DOWN, LEFT, RIGHT).  The switch is open so no ground is reaching these switches other than the conditional grounds, joystick would still function in 4 way mode.

Fourth drawing is the arrangement with the triple pole switch in the closed position, all four switches have direct ground so all can send signals, joystick would be 8 way action.  It is key that the 3 lines coming out of the multi-ground switch do not touch each other on the way to the joystick, they need to be isolated or your will lose the benefit of the 4/8 switch.


I have not tried this on my system yet, just something I have been thinking about.  There are definite limitations, won't work for sticks with leaf switches and would need to spend some time figuring out best order of switches.

* If this is in the wrong area please feel free to move. *

[z]

Ideas always hit after you post...I'll leave the original up to show the evolution of the idea.

Here's a wiring diagram that only requires a $2 SPST toggle switch to jump from 4 to 8 way action.  The arrangement of the switches is critical, when put in 8 way mode (SPST closed) both UP & Down share a conditional ground and LEFT & RIGHT share a conditional ground.  The conditional ground in this arrangement is not an issue since in theory you can not hit UP and DOWN at the same time or LEFT and RIGHT at the same time.  Same limitations as above apply, no good for leaf switches, etc.  Not trying to take away from anyone's offerings, only put this out here since this is a DIY board and curious if the idea would work.

[:oj]

This seems to be a brilliant solution at low cost, indeed. But it won't work with any Japanese joysticks (without swapping the switches at least) or any other joystick using microswitches without the NC pole.

[Bender]

DAMN!!!!!! That is Brilliant

I can't believe no one thought of it before
I love simple and cheap solutions!!!!

[bencao74]

@Z :

jep, sounds fine. But I think from game feeling point of view this solution is only the half truth.

I observed when playing Lady Bug that i'm taking automatically the diagonal directions when taking a corner.

Running down the maze - Press Down
Taking corner left -> Press Down + Left -> in your solution you stick at down. Hope I`m correct. Please correct me if I`m mistaken!!!

Problem is that you still have the 8 ways and not exact movements. E-Limitator filters and optimizes this movements. Hopefully Bender can say something about the play feeling after he received his pcb and parts.  :-)

Bencao

[Bender]

neither of these resolves the physical restrictor part of the problem, but with an octagonal restrictor is should feel ok
there is almost always a trade off for that kind of flexibility

[RandyT]

I don't mean to take the steam out of this thread, but "brilliant" is stretching it a little.

This is something that has been tried and discussed in great length here many years ago.  The conclusion then, and to my knowledge still, is that this is not really the best solution to the problem.

The reason is pretty simple;  The device can not know the true intentions of the player, as there is just not enough information being put out by the control for it to make that determination.  All it can do is "lock" a player into a direction until that direction is broken.  MAME called it "sticky" mode for that reason.

In the two examples provided, this solution is little better than the problem.  In Donkey Kong, he doesn't stop if you accidentally hit a diagonal, he just keeps going in the direction you are moving in.  If you happen to be going up a ladder, this might be useful.  If you are just getting on one, you might keep on going smack into a barrel, which is less useful.  

Same situation for Pac-Man.  Sometimes a player intends to change direction and hits a diagonal instead.  Sometimes the diagonal is hit in error, when the player had no such intention.  The device absolutely cannot know the difference, so it comes down to a "pick your poison" type of scenario.

If you like the "sticky" mode of MAME (and not everyone does) then this might be useful to you on other machines / emulators.  But no-one should expect miracles, or even see it as a reasonable alternative to "reaching under the panel".  The results of each are vastly different.

RandyT



*edit* typos....

[:oj]

All it can do is "lock" a player into a direction until that direction is broken.

If you really think, the E-Limitator is such a simple ---steaming pile of meadow muffin---, then you're completly underrating or misunderstanding this project.

[Ginsu Victim]

No, I think Randy nailed it. I haven't liked what I've read and he pointed out the reasons I was thinking.

[RandyT]

All it can do is "lock" a player into a direction until that direction is broken.

If you really think, the E-Limitator is such a simple ---steaming pile of meadow muffin---, then you're completly underrating or misunderstanding this project.

Well, there is only one other option.  It can also do the opposite and open up the actuation of the cardinals to make it very sensitive to direction change, which would really do little more than to exacerbate the problem of user "drift".

Possible logic examples are as follows:

Method A = if previous direction equals left and current direction = left/up then new direction equals up (makes changes more sensitive)

Method B = if previous direction equals left and current direction = left/up then new direction equals left (makes changes less sensitive)

The problem with both of these is that there is no middle ground.  You might be able to get fancy with a time-base approximation or some sort of averaging scheme, but at the end of the day, even those will just be guesses, and a large portion of the time they would guess incorrectly.  

The discussions about this should be archived on here somewhere.  I'll see if I can dig up the links.

RandyT

[z]

bencao74,
Your read of my arrangement is correct, there are definite flaws that would show up and it is limited in what joysticks it would work with.   The arrangements above and the 2-4-8 arrangement below will always come up short for smooth 4 way action on an 8 way stick when a diagonal is selected.  Like you I'm also looking forward to Bender's review and hoping the E-Limitator shows well.  Many people (like me) never get past the 'idea' stage, I've got a lot of respect for people that put the time and money into actually putting their idea into the marketplace  

Randy,
I follow your logic and appreciate the thorough explanation, you did a great job summing up the challenge.  Until the E-Limitator is reviewed I'm still leaning toward physical restriction (along with the other parts I have bought from you I purchased a couple of SUPER 4-way™ Precision True Restrictors from you last year  ).


I'm sorry if my concept sketches hijacked this thread, I hoped to add to the discussion, not pull the focus away.  Mods may move my posts if they need to.

Z

PS - Bender and oj, thank you for the kind words!

[:oj]

I'm sorry if my concept sketches hijacked this thread, I hoped to add to the discussion, not pull the focus away.  Mods may move my posts if they need to.

In no case! Your solution is a very tricky thing and it would be a shame to hold it back. And I think it was a very good example for pros and cons of different solutions for the same problem:
- if I had a DK cab (for instance) and cant get a 4-way stick, I probably would use your method
- in a multicade cab (with 2 players) which requires all kinds of controls (and even 2,4 and 8 way steering), I'd rather take the E-Li

[boomstick]

@ Z

I for one would like to thank you for your input and suggestions  .  I had certainly never considered tackling this problem using switched grounds.  I was all set to design a modular control panel for my (still to be built) cab so as to include a dedicated 4way.  Having seen this thread, the E-Limitator is a possible reason to reconsider, and i must say that I will be giving your method a try as well.  

At present, due to space and time, I have only built a rough desktop Control Panel. Rather than messing around with the limited space available, adding a simple PCB or implimenting your switched ground method seems like an interesting test bed as to what will be most appropriate for a future cab design.

While I agree that there is no true substitue for the real thing, and I may decide on the modular panel afterall, I think [personally] both these methods need investigating.  They may not be everyones solution, but worth a shot.

Cheers

[Luigi]

My apologies for my absence

To answer the question: no, I don't have one yet. But I could test it on our last party in november. Punkrockcavern had a prototype which we installed at our test station (didn't take photos nor did anybody else ). I didn't have much time for testing due to the number of people queueing to test it, too. PLUS the sticks in our test station are mounted almost vertically so playing there is not fun at all. BUT we hung up a Mr.Do-PCB and I really felt the gameplay improved when switching from 8-way to 4-way.

If you own a DK or Pac Man you won't install an E-Limiter if you only get your hand on an 8-way-stick but you will install a restrictor plate. But I'm really sure this piece makes sence in a generic cab with 8-way-sticks where you use to play different PCBs or even a 48/60in1.

I'm looking forward to have it installed in my cab and I'm curious about Benders review.

[RandyT]

I follow your logic and appreciate the thorough explanation, you did a great job summing up the challenge.

Thanks, but I think it needs to go a step further.  The idea with the switches in series is interesting, but in the end I think you will find it to be not very useful. 

As you have it configured currently;

Up=Up
Up/Right=Up
Up/Left=Up

Down=Down
Down/Left=Down
Down/Right=Down

So what this gives you is a stick that is hard to go from Up or Down to Left or Right, and very easy to go from Left or Right to Up or Down.  This would give it a heavy "preference" or weight it toward the verticals.  Changing the order of the switches isn't going to help.  It will just weight the stick toward some other direction, which is really not what you want to be doing.

I'm sorry if my concept sketches hijacked this thread, I hoped to add to the discussion, not pull the focus away.  Mods may move my posts if they need to.

As it is all related to the issue one will experience when attempting to do such things with 8-way digital sticks, it's probably good to leave them as part of the discussion.  They help to demonstrate why it is this methodology has been ignored for so long and considered by most to be something that is fairly ineffective.

RandyT

[bencao74]

I didn't have much time for testing due to the number of people queueing to test it, too.

This party was very motivating for this project. The feedback was positive and the testers noticed an improvement in game feeling. So this was the starting shot to go to market

But I'm really sure this piece makes sence in a generic cab with 8-way-sticks where you use to play different PCBs or even a 48/60in1.

jep, exactly!

@Z

Many people (like me) never get past the 'idea' stage, I've got a lot of respect for people that put the time and money into actually putting their idea into the marketplace

Thanks a lot, I've searched my blog for the first entry of this project. June last year the first prototype was done. time flies
Bencao

[:oj]

For current reason: Please note that prices shown in the shop already include German VAT. This will be substracted (whilst checkout), if your shipping address isn't in the EU. So the present price for orders outside the EU is 28.40 USD for 1 pcs (depends on the currency exchange rate too). Bulk orders start at a minimum of 10 pcs.

[Bender]

Just wanted to give a small update
I received the E-liminators a couple of day ago
They were very easy to assemble, if you are comfortable soldering plus there are great directions on the website
I took about 15 min to assemble and put together
I'll give a full review in the review section in a week or so after some more testing
My initial responce is very positive. It makes a standard 8-way feel almost as good as the maps for the U360 (but limited to only standard 2-4-8 way of course)
The issue with both those solutions is still the lack of physical restriction
But considering that, they work great I was able to get over 100,000 on Donkey Kong on my first run through, so I was very pleased, Played Tron and was able to seamlessly switch between 4-way and 8-way in mid game (now that is cool!) no jaggies for me on the light cycles!
Also some other good news for me is that the joystick I'm using for this is a hacked wireless XBOX360 controller which only out puts 3.3v and the board asks for 5v but everything works fine off 3.3v (I did not install the resistors on the LED because off the lower voltage)

[Necro]

So what this gives you is a stick that is hard to go from Up or Down to Left or Right, and very easy to go from Left or Right to Up or Down.  This would give it a heavy "preference" or weight it toward the verticals.  Changing the order of the switches isn't going to help.  It will just weight the stick toward some other direction, which is really not what you want to be doing.

Just thinking this through initially, I had the exact same thought as you but then I was thinking a bit more on it - there's a chance that something doing with timing of switch presses is being used to determine 'primary' button.  Since a computer is doing the polling, even milliseconds should be discernible and it should be possible to clearly determine what switch was pressed first and make that one only activate or switch to another switch if it's clearly activated,

Basically, there seems to be - based on how it's done - the potential for this to be much more than what you said and have a decent algorithm in it to determine what direction to prioritize under a variety of situations/combinations of presses. 

[boomstick]

Look forward to reading your review bender.  It will be interesting to her your experiences.  I myself ordered a unit which arrived a couple of days ago. haven't yet had chance to play with it yet  .  Hopefully the weekend will allow me a little time. 

[RandyT]

So what this gives you is a stick that is hard to go from Up or Down to Left or Right, and very easy to go from Left or Right to Up or Down.  This would give it a heavy "preference" or weight it toward the verticals.  Changing the order of the switches isn't going to help.  It will just weight the stick toward some other direction, which is really not what you want to be doing.

Just thinking this through initially, I had the exact same thought as you but then I was thinking a bit more on it - there's a chance that something doing with timing of switch presses is being used to determine 'primary' button.  Since a computer is doing the polling, even milliseconds should be discernible and it should be possible to clearly determine what switch was pressed first and make that one only activate or switch to another switch if it's clearly activated,

Basically, there seems to be - based on how it's done - the potential for this to be much more than what you said and have a decent algorithm in it to determine what direction to prioritize under a variety of situations/combinations of presses.  

Sorry, but I honestly believe you are mistaken.  I was referring to the switch wiring method in my post, not a "computer" (microcontroller), which I am all too familiar with capabilities-wise.  Regardless, there is not much difference in the microcontroller method, other than that one can selectively maximize or minimize the weight given to the cardinal directions.

The plain and simple fact is that a controller can never "know" the intentions of a player, without having a substantial amount of input to base a decision on.  With a switch based stick, there are 8,  and only 8, pieces of information from which to decide.  From those eight pieces of information, only 2, other than the information already in hand (current direction), will ever be important for considering the intent of  the player in 4-way mode.  Cardinal Direction A + Cardinal Direction B, and Cardinal Direction B.  

If you know of some method, other than the ones I provided above, to accurately discern a player's intentions from a paltry two pieces of information, then perhaps sharing your idea would be more beneficial than inferring the possibility of some magic computer stuff that logic can't support.

We also want to make sure that results are quantifiable and not attributed to a "honeymoon" period associated with a personal expenditure.  We all know what MAME does to the controls in "sticky" mode.  If this piece of hardware can do more than that, we should probably get it working on stock market tips .

RandyT

[bencao74]

They were very easy to assemble, if you are comfortable soldering plus there are great directions on the website
I took about 15 min to assemble and put together

Thanks for the kind words about the documentation and assembly instruction

But considering that, they work great I was able to get over 100,000 on Donkey Kong on my first run through, so I was very pleased, Played Tron and was able to seamlessly switch between 4-way and 8-way in mid game (now that is cool!) no jaggies for me on the light cycles!

sounds like a improvement in gameplay. For my work this qualitative comments are very motivating
But we`re currently working on statistical analysis to measure the quantitative improvement in different games, too.  For this statistical analysis we're always open for extending our population. 100,000 points is a very good score for me. I would categorize you in the pro players. This category of players is hard to find for a representative population . So if you don't mind, you could send me or jan a mail or pm, if you want to attend this investigation. If anyone else want to attend, don`t hesitate to send a pm.

Hopefully we can release some descriptive statistic, soon. Depending on the factual basis we could think about some inference methods, too.

Also some other good news for me is that the joystick I'm using for this is a hacked wireless XBOX360 controller which only out puts 3.3v and the board asks for 5v but everything works fine off 3.3v (I did not install the resistors on the LED because off the lower voltage)

very nice

bencao74