Pinball Project
Posted in
Afternoon watch, 4 bells (2:28 pm)

For quite a while now, I've wanted my own pinball machine. But buying one is expensive. Shipping is expensive. Local (within 300 miles!) sales that happen are for uninteresting machines, or broken ones.

But wait! I like to make things. Why don't I make my own pinball game?? It's really just a bunch of switches and solenoids. How hard can it be? (/sarcasm) I have a great local makerspace, with access to CNC machines and lathes (metal and wood), laser cutter, 3d printer, etc etc. I know some things about electronics. I know woodworking. I have taken classes on machining. I can do this!

There are plenty of resources online if you want to rebuild/repair/reskin an existing pinball machine, but nobody, and I mean nobody does it from scratch. Probably because it's a pain, and maybe I'm setting myself up for failure here, but this is a learning opportunity for me.

You can buy pinball solenoids from $10-$35 or so, but if you want an entire flipper mechanism with a solenoid then the price goes up and up. Plus there are no guarantees that what I buy will even be correct for what I want to do. You can buy pop-bumpers, slingshots, and all the other parts. But then you're just assembling a pinball machine, not making one.

Instead of spending hundreds of dollars on solenoids, armatures, and what not, I'll just build them myself.

So this page will be my commitment to documenting my ups and downs.

Summer 2015
I've sketched several designs, and done some 3D modeling in Tinkercad. Version one was very simple:
Simple solenoid housing

Version two had one end designed to come off, and to be glued on. Not very interesting, and I'm not sure how good an idea that was anyway.

Version three was way more complicated, designed with a square base with mounting holes, an end to take a nylon sleeve, etc:
Third solenoid housing design

I started to print the v3 solenoid housing, but my printer filament wasn't coiled very well at the factory, and bound up part-way through the print, pulling the filament from the extruder halfway up the shaft. That was frustrating:
Bad 3D print of solenoid-v3

Plus, it seems way too big for what it's going to do…

28 September 2015
I got a little math help at work from a coworker, and wrote up a formula to estimate the thickness of the coil so that the solenoid spool ends would be tall enough to contain the wiring. Of course the estimate assumes perfect coiling and stacking of layers of wire which won't happen, but it's a ball-park figure. Basic calculations for my wire and spool diameter say spool ends of 1" diameter should be plenty high to test several different designs.

Here is the formula we came up with:
Solenoid size formula

30 September, 2015
I had an evening of one-on-one training on the metal lathe at my awesome local makerspace. I made a bushing to get the basics down. Then I went straight on to start building a solenoid from scratch. Using a 1" delrin rod, I lathe'd out a 1.750" long spool with a 0.500" center diameter and a 0.390" hole down the center (25/64, for a 3/8" steel armature to slide through). The spool ends are 0.100" thick:
As-build solenoid spool
As-build solenoid spool

Making the entire spool out of delrin means I don't need a center sleeve to reduce armature friction. It also means if it breaks down too much I have to replace the whole coil instead of just a sleeve. I'm on the fence about this design decision, but I'm also not going to worry about it right now.

I'm going to make one with 26 AWG magnet wire (mine has a diameter of 0.017"), and I want 600 wraps. Plugging these variables and my spool size into my equation:
a = 0.500 + ((0.034 x 600 x 0.017) / 1.550)
a = 0.500 + 0.224
a = 0.724

Just a little shy of 3/4", so plenty of room to work! Doing a little more math, I find that 1300 wraps is probably my maximum if I do a good job getting them real evenly spaced. More than that and I'll need a bigger spool!

Which brings up my next point: how to wind the wire onto the spool? I really don't want to do it by hand. Especially if I have to make a bunch of these. Plus if I sneeze and lose my count I'm up the proverbial creek and have to start over. So I'm going to build a winding machine.

I've been thinking about a winding machine all summer. Last July I bought this digital counter on Amazon:
Digital counter

2 October 2015
I found a replacement sewing machine motor and foot pedal on eBay for $20. This will let me control the speed of the coiling while leaving my hands free to guide the wire onto the spool. I tried to find some used sewing machines on Craigslist but they were all way more than the replacement motor. I also bought these normally open magnetic reed switches to use as a trigger to the digital counter. A small magnet on the shaft being turned by the sewing motor will trigger the reed switch each rotation, incrementing the counter. Voilà! Reproducible windings. This is totally going to work.

5 October 2015
Wow, the sewing machine motor is already here! I plugged it in and tested the foot switch, it's going to work great. I need to step up my game and get the winding machine framework designed.

15 October 2015
Well the rotation sensor I recycled from an old laserjet printer that I thought about using for the solenoid machine died when I was poking it with 5v trying to get it to work—it lost all its magic smoke. I probably hooked it up backwards or something. Fortunately the magnetic reed switch idea that I was originally going to go with will work, I had a pack of them arrive in today's mail. I breadboarded up the digital counter with a 9v battery and a reed switch, then watched it count up as I waved a fridge magnet over it. There were a few double-counts, but it probably had to do with how I was waving it around with my hand.

19 October 2015
I pieced together the electronics for the winding machine and soldered everything together. The digital counter is wired to a small perfboard that has manual controls for up, down, and reset. Additionally, I've soldered a pair of wires to the up counter with a magnetic reed switch so it also increments the counter. Here are some pictures of it:
winding machine electronics
closeup of controls for winding machine

23 October 2015
I had some extra hours at work to burn off, so I took today off to work on the winding machine. I grabbed some scrap wood from my garage and headed to Gizmo, my local makerspace. I machined a pulley out of some 1.5" aluminum rod that would fit the belt that came with my sewing machine motor. I drilled a hole through the pulley and glued a neodymium magnet to trigger the reed switch.
aluminum pulley for winding machine

I mounted the motor so I could figure out where to drill the holes for the rod and bearings.
sewing machine motor mounted

I used a metal rod salvaged from a dead laser printer. The flanged bearings were given to my by Marty, one of the coolest guys I have ever met, who also runs Gizmo with his wife.
rod and bearings for winding machine

The pulley has a set screw to hold it on the rod:
winding machine pulley set screw

I used the laser cutter to cut out some 1/8" baltic birch plywood to mount the digital counter and controls. Because of the tight tolerances, I had to remove the side tabs on the back of the counter and will hot-glue it in place:
back view of digital counter

Final assembly of all the parts except the magnetic reed switch, which I broke while putting everything together. Those things are fragile!
DIY winding/coiling machine

24 October 2015
I hot-glued a new magnetic reed switch to the wood behind the pulley, then soldered the connection to it in place. I plugged everything in and voila!, I have made a winding/coiling machine.

28 October 2015
I picked up some 1.25" aluminum rod and machined out some holders that will sandwich a spool with a center hole up to 1" in diameter. The parts are cut at 60°, tapped with 3mm screws. I'll add thumb screws later, and still plan to Plasti-dip the cone parts to increase friction and hold the spool tight.
Machined spool holders for winding/coiling machine

And here it is, all assembled
My homemade digital winding/coiling machine

5 November 2015
I drilled out some 1" holes for dowel (or my spare delrin rods) to hold the magnet wire spool while I wind. At some point I once again broke the magnetic reed switch that triggers the counter to increment. Those things are fragile! I clipped it off and installed another one (good thing I bought a bag of them!), this time I installed it in a big glob of hot glue. I carefully bent the leads to make it easier to solder to and to take leverage away from the brittle glass that encapsulates the switch. Hopefully this is the last time I have to replace the whole thing.

The other (bad) part about breaking the machine down to drill the holes is that since it's all wood, it never goes back quite the same. Version 2.0 will be made from metal, probably aluminum. I'm heading back to my metal supplier to comb through the scrap pile again tomorrow at lunch, it's only a couple of minutes from where I work.

So, with all that drama out of the way, I actually wound my first solenoid tonight. Here it is, getting started:
Homemade solenoid being coiled

And here it is, finished, with 500 turns of 22 AWG magnet wire:
DIY solenoid finished winding

And after I wrapped it in electrical tape:
Home-made solenoid

If you look at the first two images, you can see how nice I started out and how I finished. It's really hard to keep the sewing machine motor turning at a steady rate, even controlling the pedal with my hand. And once I made a mistake, it compounded the rest of the coiling. It wasn't terrible, but not nice and neat like I started. I think I'll come up with something to stop the pedal at a certain distance to always get the same spin rate.

Oh yeah, I also over-tightened one of the spools holders and blew out the threads. I had to drill another hole and tap it, so I just put the bit through the old hole and drilled through the other side. It was easier to line up that way. Note to self: don't wreck the threads again.

7 November 2015
I printed out some prototype flippers on a Makerbot Replicator 2. These took about 7 hours to print out.
DIY pinball flippers

18 November 2015
It's been busy lately, no progress to speak of. However, tomorrow I'm driving over to Gig Harbor for the FAST Pinball/Mission Pinball meetup!

22 Jan 2016
Well I've taken time off for the holidays but I'm starting back up again. The Fast Pinball meetup was awesome despite being such a long drive for me. I drove from Coeur d'Alene, where I work, over to Gig Harbor, a 5 1/2 hour drive (more with traffic). I took my solenoid winding machine. There was a fair bit of excitement over it. None of them had ever known anyone to manufacture their own coils. It made me feel special—or stupid for doing it all from scratch. But look at what I've learned…

So last night I had my one-on-one training with the CNC milling machine. This is where things are going to start coming together. I need to mill two solenoid parts: the metallic plug that links the solenoid to the armature it pulls/pushes, and the bar that links that whatever it's pulling/pushing. I have some ideas down on paper now that I think are pretty cool (thanks to Marty at Gizmo-CDA!) and plan on getting them milled in the next week or so. I have to draw them up in CAD first. I'm excited again, and glad to be back on the case. Once I have these parts, the last thing I need for solenoids is a frame to hold them together and to the back of the playfield. And at that point, I'll start needing control hardware!

23 Jan 2016
Today I spent some time drawing my armature and linkage up in CAD. I have a simple file for the solenoid plug, just one horizontal cut with the mill for a slot in the end. I also drew up three different linkage sizes: 1.5", 2", and 2.5". That should give me enough space to connect anything I want. I drew the tooling in CAD so I can mill a stack of four of each size, all at the same time. That's 12 pieces at once! Then I got distracted because it was music night at Gizmo so I put down the mouse and picked up my mandolin. Good times were had by all!

25 Jan 2016
I received two power supplies in the mail today: a 48v 6.7a supply and a 12v/5v supply. Between these and the PS-4N70R5R12 70v/12v/5v monster I already acquired, I'm set up for power.

8 Feb 2016
I made my tooling (for the aluminum linkage to a solenoid) on the mill. I drilled and tapped holes in it for registration in the vise on the mill as well as for positioning my material to cut out the parts I need. The milling program I wrote will create two linkages of three different sizes (1.5", 2", and 2.5" length), for a total of 6 parts, each time I run it. The original plan was to run four of each part, but given the size of the parts and the end mill I'll be using, two is safer. I can always load more material on and run the program again to cut out more parts.

11 Feb 2016
I spent some time getting familiar with the startup and shutdown procedures of the CNC mill at the Makerspace. I was unsupervised, so I decided not to try to cut anything solo, at least not just yet. I'd rather have someone with experience oversee my first couple of jobs. I worked on zeroing the axes and identifying the tooling's 0,0 when clamped in the vise. My results for 0-Y were different than my first time with supervision—another reason to not go forward quite yet. Here's a picture of the tooling.
DIY linkage for solenoid on CNC milling machine

28 Apr 2016
It looks like I forgot to update earlier this month. On 13 April 2016 I milled my first parts:
CNC milling my pinball linkage parts
My custom pinball linkage

Then last night I made some adjustments to my tooling (centering it in the vise to take some pressure off one end that was sticking out further) and milled two more bars of material for a total of six new linkages. I did this mostly on my own, getting double-checked on the mill before I ran my program to make sure I didn't break anything. I have a total of nine linkages in three different lengths now. This Saturday I plan on machining the armature that will slide in the solenoid and pull on these linkages. That part will be mild steel rod, with a slot in the end and a hole drilled for a 3/32" coiled spring pin to hold the armature and linkage together. Essentially what this means is that I expect to have something I can actually apply voltage to and watch it move this weekend!

30 Apr 2016
I milled my armatures out! Here is one with the slot being cut out:
Milling the armature for a solenoid

There were several considerations to be made, and in the end I chose to cut steel rods down to about 6" each, then mill an armature end on each side. The rod can be cut anywhere between the two ends to yield armatures of different lengths. The choice for 6" was so the vise in the mill can hold on to the steel rod. After getting my first one milled, I went on to mill 9 more rods, for a total of 18 armatures possible. The mill program uses a 3/8" end mill to face the end of the rod, then I switch to a key cutter bit to cut the slot out in two passes (one pass wasn't wide enough for the linkage to fit). Finally, another tool swap with the drill to spot the holes where the coiled spring pin fits in. After milling, I drilled those spotted holes by hand with a 2.4mm drill bit (someone absconded with the 3/32" bit). I assembled a full solenoid and tested it with spectacular results:
Fully assembled solenoid plus extra armatures

I hooked up a 48v 6.7a power supply to the coil with alligator clips and it nearly yanks the linkage right out of your hand! Next step: build a table and mount and assemble a flipper, then test how strong it is. I'll have to pick up some control hardware soon now. That said, the Fast Pinball guys will be at the Northwest Pinball Expo again in about four weeks, so I may just meet them there and talk about my needs in detail. Plus, their core controller is out of stock right now…

10 May 2016
I just brokered a deal on Craigslist to get all the guts of a 1974 Skylab pinball machine. No box, no glass, and unknown condition. But hey, I was just about to buy some nice 3/4" Baltic birch plywood for more than it's costing me for all these parts. I will pick it up on Thursday and post some pictures when I can. This will be a great way to test some of my parts and ideas and possibly get a few new ideas.

3 Jun 2016
I made my second trip to Tacoma to attend the Northwest Pinball and Arcade Expo. I had a nice long chat with FAST pinball, nailed down the hardware I needed to start, and placed my order!

14 Jun 2016
As I'm waiting for my hardware to arrive, I've been going through the documentation for FAST and the Mission Pinball Framework's upcoming 0.30 release. After ordering my hardware, I was added to the community and have been able to chat with the developers and other pinball makers. I'm really excited about getting started with FAST and MPF, they are a power-duo!

30 Jun 2016
I've had my hardware for a little while, but have been working at my makerspace on a robot project due for a parade on the 4th of July. I have continued to work on MPF though, and just this morning fixed a problem so the MPF media controller runs on a Pine 64 SBC (Single Board Computer). The problem was with the Kivy library, and running sudo pip3 install kivy --upgrade appears to have fixed it.
Mission Pinball Framework running on Pine 64 SBC

5 Jul 2016
I just ordered a pack of ten inductive proximity sensors that I'm going to test out my "invisible switch" idea. By inserting them into a recessed area of the back side of the playfield, they should be close enough to sense the steel ball roll over their position. I think this makes playfield design much more interesting and less cluttered. I hope it works.

12 Jul 2016
I printed out some new flippers that should fit standard flipper rubber bands like the bag I got at the Pinball Expo last month. They are designed with a hole in the end for an M3 bolt and a hole in the bottom for a 0.375" steel rod which I'll drill a hole into and tap to hold the flipper in place. These two took 1 hour and 9 minutes to print out.
3D-printed pinball flippers

20 July 2016
I became a one-day crimping champion yesterday putting cables together for the Fast Pinball hardware. Biggest tip I have here is pay for some good crimpers. Engineer PA-09 is real nice and will run you about $40. I bought cheap ones before I knew about this, save yourself the trouble and get good crimpers, they're totally worth it. The heart-shaped part of the crimping die forces the wings around and down to hold the wire and the jacket really tightly. There are crimpers that will do both in one shot, but unless you do this every day you don't need them.
Crimping power connectors
Crimping power connectors
Crimping power connectors

21 July 2016
I soldered some connectors onto the FastPinball LEDs that I have. I should do this for all of them, but there are a bunch and that will take time. I put the connector on, then attach the female part temporarily so I don't solder the male pins on in a way that the mate won't connect to it.
FastPinball LED soldered connector
FastPinball LED soldered connector

I also tested my inductive proximity switches at 12v DC, they are going to work great!

As this page is getting progressively longer, I've decided to move this project over to my pinball project wordpress blog from here on out.

Leave a Comment »

Leave a Reply