Building TILT detectors for Visual Pinball?

[sofakng]

Anybody know how to build a full set of tilt detectors for Visual Pinball?

I've seen a tutorial (maybe from here?) that used a vibration detector from Radio Shack but that didn't detect direction of tilt or anything like.

After looking on vpforums.com I've seen one guy who made front AND side tilt detectors so if you hit your control panel to the left it would tilt left, etc.  However that was over a year ago and you must pay to post on vpforums.com so I can't even ask the guy how he did it...

Did anybody here make anything like this?  I'd be VERY interested in building something like this but I have no clue where to even start...

[NoOne=NBA=]

I don't have any links for you, but a bunch of mercury switches, set for different directions and placed in the bottom of the CP area, should do it.

I used some in a controller I built for my Atari 2600 years ago, and they worked pretty well.

You will probably need some kind of adjustable plate for them to ride on, so that you can get them biased properly though.

[RayB]

IME, tilt detection is simpler than you think.

Real pinball machines don't care what direction you are moving the machine (why would it matter?) All that matters is knowing that the machine is being rocked.

To detect this, they use a "plum bob" mechanism. It's really simple. All it is is a metal rod, with a weight at the end of it. This hangs down through a piece of metal that has a hole in it (maybe an inch in diameter, or less).

The hanging rod has power to it, just like a button does, and the ground is attached to the metal piece with the hole. By rocking the machine this causes the plum bob to swing. If it swings hard enough to touch the edge of the hole, then that completes the circuit (just like press a button causes contact of two metal parts, thereby causing electricty to flow through).

The heavier the weight of the plum bob, the harder it is to cause a tilt.

So there you go. Shouldn't be hard to just make one of these up out of all sorts of parts.

Or you could order one from Happ:
http://www.happcontrols.com/index.html?http://www.happcontrols.com/amusement/pinball/pinball.htm!

[mp2526]

I may be wrong as I haven't looked at Visual Pinball, but I assumed he was talking about not detecting tilt as in how a real pinball would, but rather virtually pumping the table by detecting real shaking of the cabinet.

[NoOne=NBA=]

I think he is wanting "nudge" control here.

An alternate method I've used on some of the pinball controllers I've built is to add an additional button on each side.
You can set them up so that each button is like hitting the machine on that side, and hitting BOTH buttons together is like hitting it from the front.

That setup worked especially great on the Atari 2600 Video Pinball game, where you could actually steer the ball using the nudge function.

[Darkstalker]

"Real pinball machines don't care what direction you are moving the machine (why would it matter?) All that matters is knowing that the machine is being rocked."

But unlike a real table, you need to map the tilt direction to inputs so Visual Pinball can emulate it.

"That setup worked especially great on the Atari 2600 Video Pinball game, where you could actually steer the ball using the nudge function."

Yea, I can get infinate points on my Atari doing this...Well, until I get tired of playing anyways.

Back on subject,  is this CP in an arcade style cabinet?  If so, it's going to be really hard to actually shake or move the cabinet if it is built sturdy (Like it should be).  If you have just a CP (like a HotRod or XArcade) then you'll have to be careful you have the CP level when you are playing.  If you sit with the CP in your lap, moving your legs around could cause a tilt also, or even handing it to another player.

I think NoOne=NBA= has got it right that adding buttons would be the best way to go.  Not very realistic, but it will at least work properly.  

[DanteBK]

Back on subject,  is this CP in an arcade style cabinet?  If so, it's going to be really hard to actually shake or move the cabinet if it is built sturdy (Like it should be).

I think it depends on what you're going for. I can't imagine how to make it possible to do left, right, AND front nudges ... but using the vibration detector idea on the front wall of the cabinet, if the cabinet's hinged like Project Arcade, should provide enough wobble when kneed to do one of the nudges (this is a theory, I'm just now starting assembly). Just figure out which one you use the most and set the detector to push that button.

[NoOne=NBA=]

You CAN'T do separate directions with a vibration detector.
That is why I suggested the mercury switch idea.
If you get the sensitivity adjusted properly, it doesn't take much to give you a reading on them.

The only downside I can see to the mercury switches is that they would more than likely need to be debounced to prevent you from getting a reading the other way after the way you wanted.

The vibration detector would work on any game that just has a generic nudge on it though.

[RayB]

Well hold on here... How does Vpinball implement "nudging"?? Is it analog or digital? How the software implements it will make a big difference on how the input device should be set up...

[NoOne=NBA=]

Don't know.
Never played it.

[Pilot143]

what about something that would detect tilts per direction, kind of like an upside down joystick?  have a hanging plumb, with 4 independent grounds... one for each direction.

[sofakng]

Visual Pinball uses keys for nudging ('Z' nudges the table from the left, '/' nudges the table from the right, '<spacebar>' nudges the table forward).

As for my particular setup, I'm follow the Project Arcade book (LuSiD's design) so my control panel will be attached directly onto the cabinet.

I know the vibration detector wouldn't work because it doesn't detect directions (eg. it wouldnt know that I want to nudge the table LEFT instead of RIGHT or FOWARD).  The mercury switch idea sounds interesting but I've never used anything even close to that before.

So, can you describe that more in detail, or are there any other solutions to this problem?

I realize I could use more buttons on the side and simply have them nudge the table, but it would add a lot more realism if I could implement physical nudging...

[Darkstalker]

"what about something that would detect tilts per direction, kind of like an upside down joystick?  have a hanging plumb, with 4 independent grounds... one for each direction. "

I think some terminology clarification needs to be made.  RayB has it right:  On a real pinball table, you install a plum-bob device (http://www.happcontrols.com/amusement/pinball/95032800.htm) that detects if the table has been shaken or tilted past an acceptable point.  If the plum-bob makes contact and grounds out, the machine tilts the ball and freezes the controls until it reclaims the ball in play.  A nudge is anything less than a tilt.  I appologize that the last section seems really bonehead, but it brings me to my very important next point:  How can you detect the difference between a hard nudge and a tilt?  How can you tell between a hard or soft nudge for that matter?

Maybe I'm being overcomplicated or anal-retentive about it but it seems like a really hard, if not impossible, thing to emulate properly.

[Darkstalker]

Snuck your post in before mine...heh

"I realize I could use more buttons on the side and simply have them nudge the table, but it would add a lot more realism if I could implement physical nudging..."

Being realistic, if you have a full 100+ pound solid MAME cabinet you're not going to be doing much, if any, shaking of the cabinet.  I think nudge buttons would be the best way to go, and setup VP so if nudge buttons are pressed 'X' times within a 'Y' time period it triggers a tilt.

You could put multiple vibration detectors on the cabinet, and knock on one side or the other to get a nudge, but you could get "bleed" to another detector.  Mercury switches detect tilt, and like I said before I doubt you'll get enough movement from a MAME cabinet for them work properly.

Besides, do you really want to promote people beating the snot out of your MAME cabinet by giving an excuse/purpose for it?

[Xiaou2]

 The tilt bob works almost like what is needed to detect nudges.   The tilt bob is sensative adjustable by changing where the weight is in relation to the outer sensor ring... so that it takes move travel to touch the connection edges.

 You could probably make a second tilt bob that has 3 seperate edges for detection - and then adjust the sensativity so that its more sensative than the tilt bob.

 I think there another method as well... by using a slam tilt switch idea - which is usually seen on coin doors.  Basically a leaf switch that has a metal weight on one end.  The larger the weight - the easier it will be to nudge it to make contact.  
 
 Youll need to place the nudge and tilt sensors on the appropriate cabnet sides to get the best sensing.  Im not sure how much effort it will take to shake up a cab - but a have seen some people really rock some ms pacman cabs quite hard with only the joystick.

[patrickl]

I once built a tilt dongle for my Palm Vx. This uses a 2 axeis accelerometer. See Tills Palm Pages
for schematics. These have two axis so you can detect both a front bump and left right bumps. One nice thing would be that you could configure how sensitive you want the "nudge" to be. You wouldn't have to tilt the cab a whole lot for the tilt sensor to still register your nudge (as with mechanical switches).

You would need some translation electronics and as I recall these sensors are not cheap either.

[RayB]

Sorry but a tilt "bob" with 3 seperate detections is NOT going to work! It will often swing, make contact with one side, then swing back and make contact with the other side! I'd even bet it could sometimes make contact with all three sides before it slows down enough.

And I agree that rocking a video cab is a bit... uh not a good idea, nor is it as easy to do as a pinball machine.

[Xiaou2]

so make 3 seperate bobs - that are only limited to one direction each.

 imagine this...  

 a steel marble sits in a plastic tube.  the tube is slightly angled upwords... and at the end, has contact points - or a switch.

 when shaking the machine... it will roll the ball up to the contact or switch... the roll back down.  

 believe it or not - shaking a machine isnt that hard.  ive witnessed it a lot when i worked at the arcade.  
 
 however... one could make a seperate lower mini pinball panel that was mounted in such a way that it pivoted a bit - using rubber surround to return it to center... and the above switches for nudges.

[patrickl]

The point is that that ball will bounce around and "click" both left and right switches (perhaps even several times) if you nudge the machine sideways (hard enough).

[Effayy]

I once built a tilt dongle for my Palm Vx. This uses a 2 axeis accelerometer. See Tills Palm Pages
for schematics. These have two axis so you can detect both a front bump and left right bumps. One nice thing would be that you could configure how sensitive you want the "nudge" to be. You wouldn't have to tilt the cab a whole lot for the tilt sensor to still register your nudge (as with mechanical switches).

You would need some translation electronics and as I recall these sensors are not cheap either.

Oh wow.  Sorry for hijacking this thread for a second, but with that tilt dongle, does that mean you could set it up to make noises a-la Ziggy on Quantum Leap?  

Oops. Sorry. I"ll take off the geek-hat now.

- FA

[Darkstalker]

"so make 3 seperate bobs - that are only limited to one direction each.

imagine this...  

a steel marble sits in a plastic tube.  the tube is slightly angled upwords... and at the end, has contact points - or a switch."

That is more or less the concept behind a mercury switch, but instead of a ball it is a bead of mercury in a glass reed switch with two diodes inside.  Here's an explanation with pictures on an unrelated topic:

http://www.shed.com/articles/TN.switches.html

The problem with this is, they detect tilt (Angles, not pinball) so you would have to lean the cabinet over for them to work.  As I said, even heavy shaking of the cabinet wouldn't generate enough lean to close the circuts, or all three would be going off with the amount you would have to turn the sensitivity down.

I did think of a mechanical switch type that MIGHT work, but I think you would have to engineer something to do it unless someone knows of something like this already.  Basically, think of it as a contact plate with a spring loaded mallet on an arm with the contact on the end.  The spring would keep the arm in and the contacts open.  Mounting this on the left and right of the inside of the cabinet, when you smack the cabinet it would cause the mallet to swing open and close the contacts.  The spring would have to be strong enough to prevent chatter (bouncing), but weak enough that you can get the arm to move without too much effort.  Here's a really crappy picture I made that will hopefully make more sense:

[patrickl]

I just saw a accellerometer PCB with a serial output. The PIC converts the ADXL data to serial codes. With a bit of knowledge of PIC coding I guess you could change that to switch the 3 nudge switches. The whole thing comes for $45 or $50 assembled.

Maybe if you ask them they can even do a version custom for Visual Pinball machines. I'd pay $50 for that.

[Soulcon]

How about if you substitute your normal Wheel and caster system under your cabinet with Spring loaded shock absorbers? Ofcourse they would have to be heavy duty to support the arcade and still allow play, but now you could utilize your Mercury switchs or the less exspensive Balll and tube system? Just a thought that I think would protect your Arcade and of course an overreacting tilter?

-Soul

[RayB]

Honestly, this is all too much for a rather small feature. Does the video pinball even handle nudging very well? (I mean, does it handle it well enough that you'd even use it?)

I know I wouldn't want to rock a cabinet with a monitor and other precious parts in it (hard drive).

How about just using slam detectors like they have on coin doors. Mount one on each side, and one in the front. You could tilt by hitting the side of your cab rather than actually rocking it...