Hey, all! I made a thing for electromechanical pinball machines, and here it is!

1976882_10152259952889371_1776935050_n.jpg

It does three things!

1 - it reads the scores from the reels, and when you score higher than the reels can display, it keeps counting!
2 - it saves high scores! With initials and everything!
3 - it provides 3.5mm accessibility sockets on the bottom of the cabinet and lets the flippers be activated with low-voltage signals, so that you can still play pinball using foot pedals or eyeblink detectors if you haven't got the use of one or both of your arms!
4, which is not a thing, and which is actually kinda my favourite - it doesn't do anything else! The machine still plays and functions exactly as it did before! It's still an electromechanical game, it still works by means of big chunky clickyclacky relays and a score motor, and if the time comes to sell it to a purist, the game can be quite easily put back exactly how it was with no trace of the mod left behind!

The eventual goal is to cover its unsightly innards with a modified instruction card, with the LCD poking out. Eventually I'd like to use an e-ink display instead so it blends in better, but... baby steps.

It still has a few little quirks - EM machines are hugely electrically noisy and working around the interference they produce is proving to be a wee bit complicated - but with a little luck this'll be going out on location test within the next couple of weeks, and if there's any interest here on Pinside, I can make more!

I must say, it's really re-invigourated my Super Star! It saves the top ten high scores, and seeing the score I have to beat really adds to that "Just one more go..." feeling.

(it actually shows the score you have to beat during the game, right underneath your current score. Like, when you're just starting a game, it shows "9: CMJ 55,340" on the bottom LCD row. And then when you pass high score nine, it shows high score eight, and so on. When I was playtesting it a mate of mine said "That's evil," and then he played some more)

I've been working on this mod since November, often going quite nuts trying to chase down the electrical ghosts that tend to come up around big solenoids and motors, and seeing it finally working... feels good, man.

With an optimism that bordered on idiocy!

And when that didn't work, with lots and lots of swears!

I made a wee control board for each flipper, so that the old switches switch (with another line to the MCU) an SSR which toggles voltage to the old wires - the breakout for the accessibility sockets are on the same boards. The flippers twitch a little when the outhole kicks, so there's some interference issues to work out.

The score reels are read optically - I briefly considered just counting the AC pulses sent to the reels, but we all know these reels don't necessarily always show the number of pulses sent to them, and it was more important to me that the lcd show the actual score according to the reels, not what the score *should* be.

At the moment it's wired into the transformer's mains input through the guts of a cheap adaptor - I tried to power it off the 6vAC GI circuit, rectified and smoothed, but that wasn't quite enough for the regulator with the current draw from the LCD and reader boards and everything. Future revisions will use the 24vAC lugs on the transformer.

Thanks! I had to learn microcontroller programming for this, and that was... yeesh. I'd gotten so used to things I take for granted, like multidimensional arrays, or even associative arrays, or even not having to save big numbers one byte at a time to specific EEPROM addresses. It's given me a whole new level of respect for early solid-state pinball designers!

I think the next step is going to be gauging interest. Perhaps a Kickstarter thing. I'd like to sell these for pretty little - I was thinking around the $100-150 mark, or about $30 for just the flipper control boards. I designed it to use very few components, and the LCD is the most expensive part - followed by those solid-state relays, which... honestly, I kinda went overkill on those, using ssr's spec'd way higher than necessary, because they're the most crucial point in the system.

(to me, those are the most important bit. Pinball is the perfect candidate for switch gaming - there are only two switches. I'd like to sell flipper control boards at a price low enough that they can become ubiquitous, and people without the use of their arms can play as much pinball as they like, wherever they like. Hell, with a pair of eyeblink detectors and these boards, you could play pinball if you were completely paralyzed from the neck down, provided you had a mate to plunge the ball for you)

(heh - also I was worried that they'd weaken the flippers. If anything, the flippers seem just the tiniest smidgen stronger - probably because they're not struggling through those knackered old flipper button switches anymore!)

Anyway, to do it cheap I need to get some custom PCB's made (this one is made out of protoboard, and held together with solder bridges, insulation tape and hope), and like anything to do with electronics, they're cheaper when you order more. I also need to make a prototype that'll work with other machines - this one's explicitly for 70's single-player Williams machines with this sort of score reel.

So, um... if you want one, let me know.

See, that's why there are so many names carved into the side of my Super Star.

Video!

Not a very good video, I'm afraid. It's really hard to take video of - the camera picks up the screen refreshes, and makes it very blurry. I'll try again with a better camera next week.

(I also took another video of the initials-entry bit, but on preview, it was even harder to see than this one)

That's the idea. Like, if you wanna order a microcontroller, the guys at the supply house will be like "That's five bucks, or if you order a hundred, three bucks each." Gotta sell it cheap enough that it makes sense for people to put in a $500 EM game.

Would this help? http://www.adafruit.com/products/399

Well, I've got it 90% accurate. When it's 99% accurate, I'll consider it finished - but it's that last 9% that's proving to be difficult.

A four-player machine is gonna be tough, but do-able. I haven't any four-player machines to make a prototype for, so I guess that's the next order of business.

I've had it running in my basement for a few weeks, and for a basement-dwelling game, it'd be great. But this is going out on location at a restaurant soon, and so it needs to be a bit more bulletproof. Location testing should begin next weekend.

For now, have another picture.

CameraZOOM-20140411174829790.jpg

Okay, a four-player version is gonna cost a little more - we'll need a better MCU, with more memory and more input pins, and/or some sort of multiplexing arrangement. We'll also, probably, have to move to a four-row LCD, for best results. People who want a four-player version:

* If you had to choose one manufacturer for me to work on first, who'd it be? Gottlieb, Williams, Chicago Coin, someone else?
(I ask because the physical shape of the reel reader boards will have to be a bit different between manufacturers)
* How much - ballpark - would you be willing to pay, assuming I could get this up and running on four-player games?

Aye, you're looking at more like two hundred, for a four-player one. The screen's the biggest expense.

No, that'd work. Good idea. I mean, it'd be a bit weird, but it'd work. That might be a good compromise for home-use machines, perhaps with a method to manually enter scores if players two, three or four score higher than player one. I know I'm gonna have a hard time selling a $200-odd mod for $500 machines.

Hey folks, some updates for you.

First, the device has a name! I've taken to calling it Eemix. It was going to be called "EMX" for "Electro-Mechanical eXtender," but that sounded kinda late-90's and wanky, like you were gonna have an X-treme X-pression X-perience or something, but "Eemix" is fun to say. Try it! Eeeeeeeeemix! Yay!

Second, Eemix is now on location testing in my 1972 Williams Super Star at DaBuzi's restaurant in Finleyville, PA, where you will find what I'm quite confident are the absolute best meatball sandwiches in the whole wide world ever! Response from the public who've played the machine has been very encouraging - high scores are addictive! Old, cheap machines can make money on location again!

Third, don't rush out there today because as I type this, it's broken! Out in the wild, on a location test where it's subjected to the chaos and randomness of the general public, Eemix lasted five whole days before needing me to go out and fiddle with it. Hell, for a prototype thing that no-one's ever tried before, that's actually pretty encouraging! I was expecting it to break down MUCH faster than that. I'm gonna write a better service/debug menu, grab my toolkit and head out there this evening.

Fourth, inspired by this thread (http://pinside.com/pinball/forum/topic/please-help-us-buy-a-pinball-for-the-disabled), I'mma tell you how Eemix's flipper control boards work. Here's a picture!

eemix_flipperboard.jpg

Note that none of this is to scale - I originally designed my schematic on graph paper, this is just a wee ten-minute Potatoshop mockup.
The big chunky thing at the top is the solid-state relay - one of these little fellows: https://www.sparkfun.com/products/10636
The chunk at the left is just a set of screw terminals, one of these to be precise: http://www.radioshack.com/product/index.jsp?productId=2102861
The big long bars are resistors - Eemix works entirely on through-hole technology, so that the average pinball owner can do any necessary repairs with a standard soldering iron, so they're fairly large as resistors go.
The squares with holes on them are my attempt at portraying headers like these: http://www.digikey.com/catalog/en/partgroup/0-100-2-54mm-pitch-connector-headers/24994 Those aren't the exact ones I used, I got a big bag off eBay, but a header's pretty much a header.

Here are the connections!

1. AC power - the wire that you desoldered from the flipper button switch.
2. AC power - the other wire that you desoldered from the flipper button switch. These get switched by the solid-state relay, now - and yes, removing that big old switch from the high-voltage side of the circuit does make the flippers a little bit more lively. Yay!
3. Connect to one side of the flipper switch.
4. (optional) Connect to one side of the auxilliary flipper switch (a 3.5mm socket, mounted underneath the machine).
5. Connect to the other side of the flipper switch.
6. (optional) Connect to the other side of the auxilliary flipper switch.
7. 5vDC. This would be supplied by Eemix's power supply - if you're using the flipper control boards without Eemix, you'll have to figure out a 5vDC source. If you like, you can get 5vDC from your 6vAC lighting circuit - this excellent post will help! http://pinside.com/pinball/forum/topic/open-source-freeish-pinball-electronics-kit-gi-buddy
8. (optional) 5vDC. Why are there two 5vDC headers? Because you'll need two of these boards! The other 5vDC header is so you can daisy-chain one board to another.
9. Ground!
10. More ground, same reason we have two 5vDC headers.
11. This wire goes to Eemix, to tell it when you've pressed a flipper. If you just want the pinball-with-foot-pedals-or-whatever features, but don't care about high-score management or other Eemix functions, you can omit this (and the 100 ohm resistor) from the circuit.

So, to those who want it, what happens now?

Well, now it's going to keep breaking down or getting confused, and I'm going to keep going out and fixing it and doing software and hardware revisions until I have something that will keep running, flawlessly, for a few weeks at a time. This is a device intended for operators, designed to get classic pinball machines back out on location and earning money, so we're going to keep gathering data on all the things that can go wrong with it in an on-site context, and fix those things as they come up, until there's nothing left to fix. THEN, we'll start on making a production version.

In the meantime, I'm getting quite excited.

For a web-based system, you'd be looking at more horsepower than Eemix can have for the price I want to sell it at - right now we're using a microcontroller with memory measured in single-digit kilobytes. If there's interest, a Raspberry-Pi-based web-enabled "deluxe" version for people with very deep pockets might be possible. That'd be something to look at further down the line. One thing at a time.

Well, since you asked for an update, here it is:

Eemix lasted TWO WEEKS on location before failing, this time! Last time, it only lasted five days before it needed my attention! That's a 100% improvement!

Heheh. I used to think that inventions either worked or didn't, and that inventors were, like, fancy-pants clever people who got things right first time. Nope. Things break and then you fix them and they break some more and you fix them harder and you learn something new every time, and the fixes last a little longer every time, until you have an official Revision One.

Anyway, as a proof of concept, Eemix works, and it's fun, and so far at least, its bugs are minor. I'm about ready to start designing a stage 2 prototype. That means actual PCB's made by a real proper PCB fabrication house, not just scrappy chunks of protoboard.

Once I have a proper first revision, and it's in the Super Star and working nicely, I'll let it sit for a couple weeks to see if problems come up. If not, then it's Kickstarter time! Woo!

Update: Eemix continues to function fairly reliably. "Fairly reliably" is not really good enough for operator use - I've been figuring out various ways to improve its software and hardware, and a first-version "Proper" prototype is in the works.

Meanwhile, I picked up a cheap Tri Zone, because I was having pinball withdrawal. This game saves the highest score (as long as it's under 999,990), but I got to thinking... wouldn't it be nicer if it supported scores higher than six digits, and saved a top ten with initials?

So, um... anyone interested in a version of Eemix for early Williams solid-state games? This should be a hell of a lot easier than the optical, electromechanical-friendly Eemix. Given that signals to the game are low voltage, we could also take over the awarding of score-based replays too, and add features like adaptive replay scores.

Update-wise, basically I'm figuring out a better version that'll work with more machines, after gathering data on how the prototype functioned in the wild. My test machine is now a Williams Space Mission, and there's nowhere near as much room around the score reels to attach the optics - glad I found that out before I started producing the previous version.

I haven't had much time to work on it lately as I've been up to my eyeballs in day-job nonsense, but parts are on the way to make a better version.

I started working at an arcade, and then I started working at PAPA, and then I got a daughter.

The prototype broke after 3 years so some more redesigning is necessary, and that'll hopefully happen over the winter.