I have a few ideas for original pinball games, but decided it would be better to work my way into that instead of leaping headlong into the abyss. To that end, I bought a beat up Tag-Team Pinball playfield. It was mostly populated although a few things had been scavenged. This will be my test bed for learning to make a new game. Here is the plan:

1. Strip out all lighting. I plan on using addressable RGB LEDs, so the old GI and matrix has to go.
2. Rewire solenoids and associated direct switches to be controlled by a custom driver.
3. Make a custom controller/driver board based on OPP by @openpinballproj.
4. Mount the playfield in a temporary cabinet.
5. Clean and repair the playfield and its parts to get it back to playable condition.
6. Create a new game using the Mission Pinball Framework from @jabdoa. I am not trying to directly copy the original gameplay, but will use it for design cues.
7. Build a coffee table cabinet and mount a display in the instruction card spot since there will be no head.
8. Get this all done to show at the Golden State Pinball Festival in May. GoldenStatePin

Stick to the plan, chums.

IMG_20190714_221512451_HDR (resized).jpg

I got all the lighting stripped out.

I did some switch matrix tests using OPP and MPF and everything seemed to work although the switches seemed inverted (active low). Strange, but I just changed the MPF config so that the result was correct.

I also got some initial testing of solenoids using my hacked together power board. It has 24V and 5V supplies as well as a power filter capacitor board from OPP/Mezel Mods.

IMG_20190803_171925072 (resized).jpg

IMG_20190811_222329523 (resized).jpg

IMG_20190807_225623741 (resized).jpg

@drsfmd, Yes, I connect the returns ("-" terminal) of the power supplies together at the supplies. The articles that jabdoa linked are great references.

I got the boards in today. This was my first time ordering from JLCPCB. They are 2-layer boards and they took about a week from order to receipt with DHL shipping from China to California. I might have got them sooner if I had stuck with green soldermask. They claim 24 hour fab time with green and 3-4 days with any other color. But I wanted red...because I wanted red.

I got the minimum order of 5. They are 7.75 x 4.25 inches.

They look quite good. Via drill hits are spot on and soldermask alignment is barely short of perfect. The soldermask is clean even on nearly tented vias. The silkscreen is clear with no smears or spatters. I have been spoiled by ENIG and ENEPIG plating in the past, but the HASL on this looks very good. The only place where the solder plating is not level was a spot on each of the corner mounting holes. I can live with that.

The third picture shows the tiny surface mount FET portion of the solderpaste stencil that I lasercut cut today. The FETs are only about 3mm square. The stencil is made from a clear polyester sheet so it it kind of hard to see. I had to use the Digikey ruler to have something for my phone camera to focus on. Sprechen of Digikey, I also got my parts order today. Maybe tomorrow I can solder a board if I have time.

IMG_20190909_165441655 (resized).jpg

IMG_20190909_165722573 (resized).jpg

IMG_20190909_121751606 (resized).jpg

In terms of OPP functionality, this board is roughly equivalent to:

- 6x solenoid wings
- 2x switch wings
- power filter board
- 2x interface wings

Or to put it more absolutely:
- Power filtering/switching
- 24 solenoid outputs (trying some new surface mount FETs)
- Up to 48 direct switches (can also use an 8x8 switch matrix)
- Raspberry Pi hosting option
- I2C and servo output via Pi
- holds two processor boards
- OPP chain ports for expansion
- 4.25 x 7.75"
- Power connectors: 0.156"
- Switch connectors: 0.1"

Is this board ready for sale?
NO! It is a prototype. Let me do some testing and iteration. Besides, OPP is in the process of moving to a new processor: https://groups.google.com/forum/#!topic/mpf-users/QVQsO6JjID8

I soldered a board today. I got a little lazy and only soldered the connectors I need for this game. I can always add them later.

There was some interest in knowing more about how I soldered the surface mount parts, so I added some pictures for that too.

1. I line up the stencil I made and tape it in place along one edge.
2. I lay down a bead of solderpaste on one end of the surface mount area.
3. Using a razor blade or putty knife, I squeegee the paste over the stencil and confirm that all apertures were evenly filled with paste.
4. Using tweezers or a vacuum pick up tool, I place all 110 surface mount parts. Then I set it on a hot plate and wait for the parts to reflow.
5. Clean off the flux and inspect the joints because surface mount soldering is done!

IMG_20190910_221726005 (resized).jpg

IMG_20190910_121557286 (resized).jpg

IMG_20190910_121937765 (resized).jpg

IMG_20190910_122101530 (resized).jpg

IMG_20190910_131733247 (resized).jpg

IMG_20190910_134833097 (resized).jpg

I make the stencil from the F.Paste layer which should not have any through hole pads on it. That is a good clarification that the hotplate was only used for surface mount parts and needs to be done *before* through hole soldering.

Yes, "windowing" of the solder paste is generally recommended for large pads. The FET footprint from Nexperia specifies these windows and the KiCad footprint already had these.

I paid $20 total ($11 for boards, $9 for shipping) for 5 boards that were 7.75 x 4.25 inches from JLCPCB.com. I am not sure what their upcharge is for scoring. Their quoting system is pretty straightforward like most Chinese fabricators I have seen, so it should be simple to get an estimate.

I powered up the 5V section of the board and connected to it via MPF. The main lesson so far is that the yellow LED and 1kohm resistor to ground on the same pin as the onboard button is enough to force the processor into bootloader mode at power up. I removed the two resistors for COIL_0-3 and COIL_1-3 to allow the processors to boot. Everything seems to be talking and reacting well to my MPF configuration.

I have some wire crimping and playfield soldering to do before I connect the 24V supplies to the board.

FET footprint (resized).png

Rather than building a legit cabinet at this point, or using a playfield rotisserie, I decided to make a temporary box. I plan on locking the playfield in this box. Since it has no bottom, I can work on the bottom of the playfield by flipping the box over. I am not done yet, but I think it is a good use of discarded plywood.

IMG_20190919_224047379 (resized).jpg

IMG_20190919_224119774 (resized).jpg

IMG_20190919_230845597 (resized).jpg

I plan on stripping the playfield down to clean it. While I am at it, I will scan the playfield. The playfield is not in great shape, so I may retouch the scan in order to get an overlay printed.

Today I followed the instructions here to modify a scanner to mosaic scan large objects: https://mpetroff.net/2013/09/scanner-modifications-to-scan-large-documents/

I used a Canon LIDE 100. As a test, I scanned my daughter's oversized coloring book. It took 4 scans. The image I attached had to be scaled down to meet pinside posting requirements. The actual image is 600 pixels/inch and is 90 megapixels. You will have to trust me, the actual scan looks fantastic.

IMG_20190922_165053216 (resized).jpg

IMG_20190922_165123717 (resized).jpg

IMG_20190922_165140807 (resized).jpg

IMG_20190922_165200176 (resized).jpg

FindingDory01 - FindingDory04_small (resized).jpg

Could somebody (SebKubi ?) with a Tag Team send me a quality picture of the guy lying on his back in the ring closest to the flippers? The flipper gouged out the artwork on mine.

I completed a stitched scan. It is not seemless, but it is still pretty good for blending 15 scanned images for a total pixel count of around 300MP. The one posted here is heavily reduced in resolution.

I have not verified that the holes and artwork line up with this scan yet. I need to do that since I am planning to laser cut a PETG sheet as a playfield protector to prevent further damage while I slowly work on the playfield scan touchup.

TagTeam02 (resized).jpg

TagTeamScan (resized).jpg

Almost ready to do a test cut of the PETG playfield protector...

TagTeamScan_PETGpreview (resized).jpg

@semicolin, thank for the photo. That is what I needed.

I am not sure how other people do their playfield protectors. I have the protector undersized a bit so it can float around and accommodate positional variance between machines. I will keep this up to date on my protector progress.

I will likely not be screen printing anything. My plastics are in poor condition and badly warped. The center one above TEAM is the worst. I could use some scans if you could provide them. I have digitally printed vinyl decals in UV ink and then adhered those to PETG plastic in the past. I used it for both the plastics and the playfield artwork. The PETG was nice because I can laser cut it. That is what I did for this 1969 Williams Suspense.

IMG_20190517_141543286 (resized).jpg

IMG_20190517_141554028 (resized).jpg

IMG_20190517_141628082 (resized).jpg

I cut out a test protector out of newsprint. Alignment looks pretty good all around, but I forgot an opening for the rollover switch above the upper flipper. Also, the shooter lane needs to be cut back a bit more to match the routed path in the wood.

Right now, I have a lot of clearance for every post and feature. I am concerned that the protector will have so much float that it may wander around the playfield a bit. Add a little grit to that wandering and it may start etching the protector or the artwork. To combat that, I am planning on making one post at the top of the playfield a tight fit. Another post at the bottom of the playfield will have a tight fitting slot. With those two references, it should be held in place but allow for some variance in post placement.

IMG_20191006_222805326 (resized).jpg

I fixed (hopefully) the PETG protector file.

-Added slot for rollover switch above upper left flipper.
-Extended shooter lane cut and added anti-splay tabs.
-Changed an upper post to a tight fit and a lower post in the right slingshot to a tight slot.

TagTeamScan_PETGpreview2 (resized).jpg

After another newsprint cut to verify the changes, I cut the playfield protector out of PETG. Unfortunately, I stored the PETG in a roll and it had a bad tendency to curl. I put it in between a piece of glass and plywood and put that in a 55degC oven for a few hours. After that, it was perfectly flat again. I left the matte protective film on for these photos so you can see the playfield protector better.

IMG_20191009_204307381 (resized).jpg

IMG_20191009_204243545 (resized).jpg

I made some solid progress this past week...along with a couple mishaps.

I have reinstalled the playfield posts using #6 machine screws drilled through the playfield and nylon lock nuts. Any of the posts that were intended to support playfield plastic mounting, spinner mount, etc. I put the machine screw in from the backside. That way, the extra thread sticks out past the lock nut and I can mount things to it. Many of the post holes were stripped or very close to it, so I had to do something.

I did a lot of soldering and rewiring. I converted all autofire devices to direct logic switches rather than power switches. I installed some arcade buttons for both flipper and start buttons. There were a number of switches that had broken in half, but I go enough pieces together to finish everything.

I wanted to do a quick test with my controller board and MPF before I had everything back together. I hooked it up, flipped the power switch, and POOF...all the magic smoke was wisping up to my face. Oops. Somehow I burned out one of the PSOC boards. These are cheap, but are also discontinued and hard to find. After some sleuthing, I found that the original game had a switch matrix switch on the same switch stack as the end of stroke switch, which had 24V on it. This would not have been a big deal if I had a flipper installed. With no flipper, the flipper pawl, was pulled but its spring and shorted the two switches out. In case anyone was wondering, no, the PSOC cannot handle 24V. You can see the black pock of failure on the main IC in the image below.

I programmed another board, installed drop targets, cut out that pesky switch that shorted everything, and installed some flippers. I was a little more cautious this time during power up. I was able to run a "game" in MPF through the actual hardware. All the switches checked out and I got all solenoids firing.

For all the testing I had been bypassing a power switch on my controller. I decided to test the power switch separately. It always seemed to be on, no matter what I did. Back to the schematic. Some of you might see my mistake in the image below. I put the transistor in backward! Pins 2 and 3 should be swapped. As it stands, current can always flow through the body diode of the transistor. I do not need this capability right now so I am planning on shorting this transistor. I might dead bug a part on there for the sake of testing at some point.

This next week I hope to get this mostly reassembled and actually get a ball bouncing around to score some points!

IMG_20191018_215935205 (resized).jpg

IMG_20191020_165328732 (resized).jpg

IMG_20191020_165428195 (resized).jpg

Capture (resized).PNG

After a little solder hacking today, I forced the surface mount power switch transistor to fit on the board with the correct pinout. It is not pretty, but power switching works now!

IMG_20191021_173343668 (resized).jpg

IMG_20191021_174018764 (resized).jpg

IT LIVES!

I finished a new power/electronics panel and added a coil show to the attract mode of my MPF game. It just cycles through all 15 coils so I can adjust the length of the coil pulses to change coil power.

IMG_20191021_222125137 (resized).jpg

IMG_20191022_230544728 (resized).jpg

After getting some solenoids to fire, it was time to stress the power system and driver board a little to see how they would fare. I modified my attract mode coil show to cycle through all solenoids every second. Also, the flippers are fired simultaneously at 3 Hz. I thought this was a pretty mean test that would reveal potential issues. It was LOUD. The video does not do the sound justice.

The solenoid drivers all look good with only a 5 degC temperature rise at worst case--the lower flippers.
Thermal image taken during solenoid test
The NTC thermistor is getting quite hot at 128 degC, but it is also supposed to get hot under heavy load. It is intended for inrush current limiting, so as it heats up, the resistance goes down, which is what I want. This is an intense load, so I expect normal operating condition to be much lower temperature. On the next rev of the PCB, I think I will try to improve the copper planes around the thermistor to better dissipate the heat.
The thermistor getting hot...it's okay

I am glad you guys enjoyed the video. I had to retake it because the first take startled me so bad I shook the camera when everything activated.

@openpinballproj, I already have no thermal reliefs on the high current planes. I think I can get the cooling effect I want just by tweaking the planes that are already there, but I might also add something to the bottom layer.

@jabdoa, that is good to know about the "max_wait_ms" command. I had not seen that anywhere in the documentation before.

I am shifting my focus to mounting some WS2812 RGB lights throughout the game. I am using them for both inserts and GI.

I made some simple laser cut holders for the surface mount LEDs for the inserts. They are made from a laminated MDF board I had lying around. It was 0.1" thick.

The LEDs are intended to clip in, but I will also tack them down with a dab of hot glue to lock them in. These stay quite cool, so hot glue should be fine.

I still need to install these for real. The installation I show below was just for a test and was not optimized. I am using the screws and screw holes from the original lights.

I also need to figure out how to mount the bayonet LEDs for GI. Some of the locations have a perfect size hole for a tight friction fit, but other locations have a hole that is too big.

IMG_20191026_233659183 (resized).jpg

IMG_20191026_233855178 (resized).jpg

IMG_20191027_162803364 (resized).jpg

Thank you! I thought that is what it was since every large hole is right next to a small hole. Almost all the red sleeves on my machine were missing or rattling around in the wrong place.

Good point, Jan. I plan on running some insert LED boards the next time I try JLCPCB's assembly service.

I mounted the electronics panel in the back of my temporary cabinet. I also installed 22 of the RGB LEDs to test them out. The result: some LEDs reacting to the drop targets and a pretty boring attract mode show!

IMG_20191029_234153456 (resized).jpg

IMG_20191029_213802326 (resized).jpg

I got all the lights installed...

IMG_20191110_213753628 (resized).jpg

No worries. I am extremely patient when someone is about to give me something. I appreciate your help.

Me too! I would never have attempted this without MPF. It makes this whole thing possible for me.

I was able to play a game! Well, 1.5 balls worth of a game. Unfortunately, my right flipper linkage broke. Clearly it is time to rebuild the flippers.

Here is the second game played with left flippers only. Not everything is installed yet. There are also a number of adjustments to make like pop and slingshot sensitivities, coil strengths, playfield angle, and a bunch of config files left to write. Oh, and sound! It is a start, though.

IMG_20191113_234959475 (resized).jpg

Replaced the old and worn bushings, plungers, and coil sleeves. Now the flippers are smooth and strong!

A little too strong, actually. At the same time, I had bumped the voltage up 10% on the power supply. I need to dial it back down because I was getting a few too many airballs. I have no glass on and the ball hopped out a few times.

The pops and slings are dialed in now, too.

IMG_20191116_174918101 (resized).jpg

IMG_20191116_174851558 (resized).jpg

I have been making slow and steady progress with the MPF game development. I still have a ways to go before it is feature complete and the bugs are worked out. I have the start of sound and music.

Here is an explanation of my current game development progress.

Thanks, Jan. It will be a little overkill for this game, but I want to get as much practice as possible before starting on my first original game.