[Spoiler alert: The work-in-progress drawings get discarded and started anew in post #9. Don't get too attached to the old ones!]

Back in 2008, I really wanted to build a pinball machine, and a significant lack of knowledge and funding didn't stop me from trying. I brought into existence a shell of a game, called Volcano Blast!, themed around a medieval village with a superstitious king and fast-approaching doom. The cabinet was satin-black-painted construction plywood, the computer was nonexistent, and the playfield looked like this.

15aa073ba9c4e63d23693f139432efdb85a42b38 (resized).jpg

Obviously, I didn't get far. Most pinball construction techniques were foreign to me and most of the desired mechanisms were out of my financial reach, me being in high school at the time. Not that more money would have helped, though. The design I laid out in Future Pinball - which managed to survive on an old hard drive - really wasn't worth playing, in hindsight.

Although I still have the translite someplace, nothing else survives of this build, having been sacrificed to Lionman to resurrect a Swords of Fury from junk. I have no regrets about this. But Volcano Blast is going to live again!

Quick content notes: The new design, art, and story will be heavily anime-inspired. If you hate anime, drain this thread now. Early posts in this thread will show you a version of the layout that were since discarded.

While I built only a cardboard construction prototype of Undertale and called it good enough, this one will absolutely get a proper whitewood so every shot can be tested and confirmed to be truly shootable. Probably will start on the physical stuff in 2020 as I bought a few too many games this year!

Yes, I'll be getting around to that eventually - and thanks for reminding me I also need to release some resources from my previous build. VP design, laser cut templates for walls and scoops, Teensyduino source code for the controller boards, and probably more stuff. I can see the VP template being useful to someone as it includes size and shape outlines for flipper mechs etc.

Tweaked my design a bit last night and will be doing more tonight. Pics to come. Left ramp is now in place. Added a scoop as one destination from upper flipper, and will probably add the ramp Y-junction soon for the other shot. Top flipper's two shots are looking a bit like STTNG, actually, with the Borg ramp and Command Decision scoop.

Also, the amount of "Rule of Three" in this game is getting ridiculous:
- 3 flippers
- 3 jet bumpers
- 3 balls
- 3 saucers
- 3 ramps (if you count the short plunger lane ramp)
- 3 main characters
- 3 wireforms
- 3-bank of drop targets
- 3-bank of standup targets
- 3 paths to a mini wizard mode

Next pic as promised.
Screenshot 2019-08-29 22.25.11 (resized).png

UT's ramps turned out decent (albeit crooked, as like the rest of the game), and I realized something tonight: I should probably stick with what I know rather than design for something I can't promise. And a nice-looking and properly functioning Y-combining ramp entrance in plain view is something I can't promise. The compromise? Turns out I did have enough room alongside the ramp for a lane to the back of the game, and I'm fairly confident I can engineer a hidden ramp to accomplish the same task.

Still a little on the fence about how I'm going to control this game. Building a PC into the last game's cabinet worked very well as far as debugging, and the processing power required to drive the LCD. But the thought of running everything off dev board firmware - and creating a new game in this day and age that can boot to attract mode in under five seconds - is unbelievably enticing. The display would have to step from LCD down to a DMD made of LED panels, and the audio capabilities would almost certainly take a quality hit, but I do like the idea. Will have to experiment with it a bit and see what the Teensy 3.5's in my drawer are really capable of.

Hidden ramp is easy enough by copying and pasting. Actually constructing it IRL will be slightly more challenging, of course.

Polished up the nearby areas a bit. I re-angled the 3-bank of standups to keep the ball from banking off of them into the volcano scoop (top lane next to ramp), as I think that could get old fast.

Rule of three now gets even more intense as there are now three independent standups too.

For a game with a lot of predictable shots, there's a LOT of rubber here. Behind the orange standup, feeding the upper right flipper, and everywhere it's visually obvious. I think it will be a nice balance between flow and randomness. The bounciness also seems appropriate with some of the thematic elements.

Click the picture for a clearer version - Pinside seems to heavily compress inline PNGs.

Screenshot 2019-08-31 11.04.26 (resized).png

And now for something completely cringy.

So, that old "game" I built? Here's the Future Pinball design. It makes Magic Girl look complete in comparison.

volcano blast original (resized).jpg

Here are some comments, and I have no qualms about being savage about them because it was my own work.
1. You know those games that feel like they managed to fit a widebody design into a standard width? Well, this is a game where if you cut out all the crap and actually made it a standard body, it still wouldn't be fun.
2. What's with all these giant chunks of scenery? It looks like a bootleg version of WH2O.
3. Cutting it awfully close with the glass clearance, aren't we?
4. I guess there were supposed to be TWO volcanoes on this thing. Dunno about you but I can't see any at all. Seems like the sort of thing you should probably plan into any CAD design from the beginning.
5. I need to promise some sort of reward to anyone who can decipher the point of all the shots, most of which come out in the right loop.
6. There are like 4 diverters hidden under the chunk-hills. Not sure why any of them actually exist.
7. Why the Whirlwind ball lock? There are plenty of other places to lock balls.
8. You know that star triggers are difficult to clearcoat, right? And are we supposed to aim for those or what?
9. Please explain the point of the red standup in the left outlane.
10. I don't think you could intentionally make a single one of the shots on the left. They are ALL too narrow.
11. I tried to BUILD this!??

Minor changes, and I think I'm about ready to call this layout done - until I play the whitewood, that is.

Screenshot 2019-10-04 22.00.02 (resized).png

Two saucers on the left seemed weird. What is this, a Gottlieb ball lock? I replaced the upper saucer (behind the drop targets) with a captive ball, as really the only goal here is to drop the targets and then get inside. Once you hit the ball, the targets will reset, deflecting the game ball down into the one remaining saucer. Good enough IMO.

The scoop above the upper right flipper has been downgraded to a saucer. No need to complicate the subway layout just to add a second way to pass the ball from the left flipper to the right.

On the right, the shooter lane ramp is gone. I realized with my last build that it seems to add some randomness to the trajectory of the ball, which is a problem for something called a SKILL shot. Using a ramp with a curved edge was important for Undertale, because the lane feeding the upper right flipper could also be shot back the other way from the upper left flipper on that game. But with no upper left flipper, there is no need for a smooth upward shot, and thus a simple one-way gate is good enough.

Also added a vertical light shield layer in front of the back wall, which will hang from the rear glass channel (think Whirlwind). Should tie together the back of the playfield and provide a place for some status-related lights in the space on the left.

Tonight I just couldn't resist messing around in Photoshop to see what I could come up with for a display-worthy game logo.
vb dmd logo dots (resized).png

I'm liking it so far. Since I already built one game powered by standard PC components, I decided I want to challenge myself this time around, and attempt to use only a set of embedded controllers this time around. With the chips available in the present day, I'm fairly confident I can build something that is at least reminiscent of an early 90s Williams title that has been retrofit with a color DMD. To save on memory, display graphics will be limited to 8-bit buffers - 6 bits for color, bit 7 for "opaque vs. transparent" when drawing one buffer over another, and bit 8 for essentially a "color vs. fill next X pixels" instruction that will hopefully condense graphics memory usage even further. More details on this once I start prototyping the display controller!

As mentioned in the first post, it'll be a few more months before I build anything. I have at least three games needing shop work first.

Those of you designing games in simulators with the intent to build: Turns out, "sit on it for a few months" is a really good idea. Sometimes after playing your design for a while, it just gets old, and you realize it could be better.

What could be better about the design I started this thread with? Well, LOTS of things. After a month or two of looking at it, I realized it didn't flow well enough, nor did the isekai-under-glass even accurately depict the environment I was trying to build.

I did want to go 90s-retro on the display and internals on this build, but I think I'm going to need higher display quality to really present this game the way I want it. So, scratch the LED idea, we're going desktop-computer again on this.

A brief summary of the backstory:
- A high school is built around a dormant volcano. It's dormant. Nobody believes it will ever erupt, and nobody wants to, because it's a symbol of the school darn it.
- Among the ordinary student body are three girls - the founders and only members of the Pseudoscience Club - with inherent magical abilities to control forces of nature - water, earth, and electricity (and, after an accident with a clock, time itself). Introductions later.
- The volcano erupts, much to almost everyone's surprise, raining down wholesale death and destruction.
- Time-girl pulls a "Groundhog Day" and resets time. The trio tries to warn everyone, and fails; it's dormant, why would it erupt?
- So each of them has a plan to try to save the school from the volcano using various strategies, supplies, and magical abilities. You, as the player, are helping them make it happen.

So between three plans and a bit of time manipulation, there are actual story-approved excuses for everything we know and love about mode-based pinball gameplay. So how does one design a playfield depicting a volcano in the middle of a school courtyard...?

Like this.
Screenshot 2019-12-18 10.30.55 (resized).png

PLEASE note that the artwork is a very quick Photoshop paintbrush job done by hand with a mouse. It is in no way representative of the desired art quality and serves only as a guide to some of the colors and shapes that could be expected.

Rule of three remains in force, and nearly all of these were intentional:
- 3 playable balls
- 3 jet bumpers
- 3 wireform assemblies (if connected pairs are considered one assembly)
- 3 blooming cherry trees in the scenery
- 3 main characters
- 3-bank of drop targets
- 3-bank of standup targets
- 3 banks of targets total (scattered thin targets will spell B-L-A-S-T)
- 3 possible exits from the jet bumper area
- 3 possible exits from the volcano ball lock
- 3 flippers
- 3 VUKs
- 3 destinations from the shooter lane depending on pull strength
- 3 shots that can be redirected depending on context

This version also potentially flows MUCH better, including two ramp diverters for the shots that otherwise would be repetitive. The diverters will require some custom machining, as they are directly over the playfield slide rails.

Not going to do anything "EXTREME AND EPIC" with the construction of this one - in fact, cabinet design will be based on the older System 11 style, shallower play area with backbox neck. With minor modifications to the depth, I was able to cram the whole playfield and all its scenery into a volume only slightly deeper than my Swords of Fury.

The next step, before I consider trying to finalize any art or get any game-specific parts produced, will be a fully flippable whitewood. A sorely-missed step on Undertale and one I will not skip this time around.

Hmm, two rough parted-out playfields next to a fresh sheet of 20-1/4" x 46" plywood. I can't imagine what's going to happen next!

IMG_20191222_151553464.jpg

A local contracting company turned out to be the perfect diagramming resource, with a large-format printer that could produce a map of playfield holes and components in one shot. My spray adhesive was dried up, apparently, but I was able to smear some wood glue over the board and get the paper to adhere that way. Definitely not a perfect laydown as I did have a few wrinkles, but for a playfield meant only to test shots, I think it's good enough.

I will wait for it to cure (additional glue may be necessary) before I actually proceed to cut out any holes, but in the meantime I just couldn't resist placing a few components on top just to get a feel for how it will lay out. It's amazing how much bigger it is in person when you're used to looking at the design on a relatively small laptop screen.

IMG_20191223_222207107 (resized).jpg

Yes, 26 days have passed.

Yes, I'm slow and had other things to deal with and it was cold outside.

Yes, this game will use a 1980s Williams ball trough.

IMG_20200119_152907490 (resized).jpg

Definitely good to be test fitting things on an inconsequential whitewood, as I've already made a couple of modifications to the cut diagram by way of pen annotations on the paper. I can also take this opportunity to start cutting guide walls out of thin cardboard, as preparation for the laser cuts I will eventually order.

Will double-check the flipper gap eventually, but I'm fairly certain I have it perfect this time around. A few hardware components had to be substituted due to me not having enough of them, but it's at least physical enough that I can tell what fits and what doesn't.

Whitewoods made of parted out playfields never look good. Tarnished metal and yellowed rubber abound.

That said, it's still a super cheap way of proving a build will go together as it should, and so far, this build is definitely going together as it should.

IMG_20200120_223854937 (resized).jpg

Ran out of the correct length of wood screw, so some of the posts are attached with drywall screws, set off with washers to avoid splitting the top of the posts.

Basement lighting at 10:40 PM is not particularly good, but this should give you a fairly good idea how the components are laid out.

IMG_20200120_223903871 (resized).jpg

Once I have the basic mechs fastened down, I can hook this thing up to the transformer I have sitting on the basement floor and actually do some flippable testing!

Wow, this thing has been sitting here for a month. Life happens, along with other annoying things like being tired or unmotivated.

Temporary wires are in place!
IMG_20200222_181525445 (resized).jpg

I figured that the flippers, slingshots, and jet bumpers would be good enough for testing purposes. I have a few more mechs laying around (a ramp lifter and a VUK, in particular) that I will install in due time. Another feature I wanted to make sure to do right is the bank of drop targets, currently still depicting the Temmies from Undertale.

IMG_20200222_152124731 (resized).jpg

My only issue was hole size, it seems, as the mech installed quite smoothly and seems like it won't be giving me any issues to speak of.

I assumed my heavy transformer brick would be enough to drive the game temporarily, but my assumption was incorrect. A quick voltage test gave me 53 VAC through one pair of secondary taps, and I hooked it up. When I plugged it in, I was disappointed to find that the current really wasn't there. The slingshots barely kicked, the jet bumpers could push the ball up to about an inch, and the flippers couldn't even raise all the way. This transformer has a lot of secondary taps at mains voltage, so I think it's designed mainly for isolation, with a low-current 50V for some other purpose.

IMG_20200222_183943571 (resized).jpg

It was nice to know I didn't horribly flub any wiring, but I still needed power to test this game. I moved the whitewood to sit alongside UT, clipped some jumper wires onto coil lugs and ground braiding, and added a snubber diode to protect this very expensive power supply.

IMG_20200226_103855038 (resized).jpg

Being at least somewhat safety-conscious, it was also very nice to have a fast-acting emergency shutoff by way of the coin door interlock switch, particularly when my overly sensitive slingshot switches fused together.

Obviously, it's not much of a game when you have to operate the flippers by touching a wire to a ground plane, so I couldn't do much shot testing for now. I will probably build a temporary cabinet front, with flipper buttons, for that purpose. In the meantime, though, the bumpers were working very well, and it gave me the opportunity to see what improvements I would have to make to the courtyard area.
IMG_20200226_091829849 (resized).jpgIMG_20200226_091803406 (resized).jpg

My VP design tried to do some bumper micro-managing with precise rubber angles and an outside deflector. When testing this out in real life, I found that these methods were actually pretty ineffective at keeping the ball within the bumper area. With nothing to lose as far as playfield integrity, I pulled a few posts and re-mounted them in new positions, eliminating several of them in the process.

IMG_20200226_102237472 (resized).jpg

The post/rubber arrangement on the far left will be removed, as it will no longer be effective in that position. All told, this change removed three posts and two rubbers, which will also be a small but welcome cost savings. Whitewoods are VERY GOOD. These changes, along with a post holding a ball gate that will have to be moved 1/4" to the left, are very minor things that most people wouldn't even notice, but encountering them on a printed and cleared playfield would be a huge setback as far as looking professional.

Big things are coming.

No exact promises as to when.

I really need to fix my screen recorder.

Undertale was designed completely out of order, with a decent chunk of the software being written before the playfield was even ready for it.

Volcano Blast! will be no different.

Except for some well-hidden switch/lamp/coil test modes, my last build was very much lacking in the operator menu department. Not going to let that oversight happen again.

On this warm(-ish), sunny, quarantined day, I have fought the forces of procrastination.

Arranged a few walls out of cardboard to test shots. Cardboard doesn't hold up the best, but at least it's something. Tested out all the major shots - everything is hittable - and shot some video.

Since I haven't planned out any scoop mechs yet, larger holes are blocked off. The shot that ends up in the shooter lane will eventually be the Hot Springs scoop. Given that I hit it twice in the video, it will be a much easier shot than the extremely frustrating Library scoop on Undertale.

Have been making a running list of things that need to be improved on the design before it can hit a "real" playfield. Nothing major, mostly just tweaks. I did spend some quaran-time last week finishing the teardowns on the playfields pictured in post # 10, and will be taking some measurements as I noticed that my lower playfield components (apron in particular) aren't quite centered properly.

Thanks guys!

So, been messing with a few things to make sure they will fit as intended. Tackled a few of the notes, just a couple more minor tweaks left on the layout. Ramps have been updated with the actual "flat laser cut" style that I've decided to go with, the same basic idea used in Captain Nemo and now Legends of Valhalla.

I learned a couple days ago that Pololu is still taking and processing orders, and that's awesome. One thing I wanted to do was go back and revise my ramps to make sure they can be physically built - and, more importantly, solidly mounted. The initial design lacked a bit of foresight:
bad ramp (resized).jpg

The green highlight is a valid mount point, however it does put a screw head directly underneath the guide rail, which would make it a bit inconvenient to install - if not impossible, because it'll need to be a somewhat long screw! The orange highlight is a mount point with no actual mounting post planned on the playfield, and the red highlight is a mount point that literally has nothing valid to attach it to.

Yesterday I tweaked and reshaped the area, and moved a few posts around, to come up with this revision:
better ramp (resized).jpg

As long as all four mount points use low-profile machine screws to avoid interfering with the ball rolling, this should work very smoothly. Every post is either the type with a tapped screw hole in the top, or (in the case of the leftmost one) at least doesn't have rubber tension on it. This entire ramp should be able to to be removed from the playfield with only those four screws, without needing to remove anything else.

The VP design so far is below. Note the newly designed ramp exit points at the inlanes. This should be possible with just a bit of bending/riveting/spot-welding of very simple laser-cut shapes.
Screenshot from 2020-04-05 13-22-48 (resized).png

And finally... operator menu adjustments are 100% implemented for now!

Still making no claims about being an artist, but I've nearly finished my playfield art sketch. New design includes smoke, energy, more lava, and the hot springs episode.

Screenshot from 2020-04-19 17-53-44.png

Street level is nearly 100% complete as far as layout and mechanical arrangement. This game will have twin back walls - one at 42", as expected on a System 11-type playfield, and one at the back (46") to support the layout elements that extend past the main wall.

Upper level is mostly done - just need to finalize the school building plastics, find the best place to mount each piece, and finish up some bent-metal plans for ramp guides and switch holders. The red flasher in the upper left will be mounted horizontally to the diagonal wall in the final design.

Remaining under-playfield planning includes shape and position of custom PCBs, scoops, and VUK guides. Looking into ATMEGA328 programming - a delightfully cheap programmable chip, currently $2.08 at Mouser, full of GPIO pins. One goal will be to reduce the amount of wiring as much as possible compared with last time.

Now we're getting somewhere.

Lamp/switch/solenoid driver PCB positions.
pcbs (resized).jpg

Since board houses tend to (a) charge per square inch of board, and (b) print a minimum quantity of 3-5 boards, I took this into consideration. Boards are kept small and thin whenever possible, and duplicated whenever possible.

Pictured:
6 RGB+white lamp duo boards (trapezoids with one red dot and one white dot)
5 single lamp boards (the little ones)
4 double lamp boards (trapezoids in and above inlanes)
3 lamp driver boards (credit card sized rectangles)
3 flipper controller boards with additional lamp and switch connections (larger rectangles near flippers)
3 linear 5-bulb lamp boards (bottom center)
3 linear 3-bulb lamp boards (on lower target banks and leading up center ramp)
1 curved 5-lamp board (center right)
1 white+RGB+white lamp board (ramp entrance)
1 extended switch board with parallel bulbs (between bumpers)
1 solenoid driver board (top right)

Not pictured:
1 hub controller (will go in the empty space in the middle)
1 cabinet driver board (will go in cabinet and handle knocker, shaker, backbox GI, and all cabinet switches)

LEDs will be warm white, dimming incandescent-style with a bit of Arduino code that I think I've already perfected. Testing out board-to-board communication protocol next, then I will start figuring out how to program the ATMEGA328's that are sitting on my desk.

White dots are single (white) LEDs, and red dots represent RGBs. The flipper boards have one RGB light each; the two boards that have no need for it will simply leave that component off the board, and using breakout headers, drive external lights instead.

Also... it's hard to tell how really complex a design is, until you have a look at the solenoid table. Oh my.
sols (resized).jpg

Computer to hub will be through a Teensy LC's USB connection. The LC will track which devices ought to be turned on and off (USB input), and what signals come back from the switch boards (USB output).

The hub will then communicate to and from the other boards under the playfield. The 8 boards with onboard ATMEGA chips will use a unidirectional serial connection TO the boards that drive devices (coils/lamps), and a unidirectional serial connection FROM the boards that take switch input.

I've experimented with some custom serial communication code and came up with a system that I'm fairly certain will work well. Data can be simultaneously sent and received from any number of boards at a time, which will cut response time down to as fast as I can switch the pins without HF interference issues.

"How can you be so SLOW?" you might be thinking. "It's quarantine time and all the other homebrew projects are coming along nicely!"

I work for an essential business, cut me some slack!

That said, although I AM slow, at least I'm making some progress. Pretty much finished up the plans for the left ramp framework and the diverter assembly. The diverters needed to be custom mechs because all the diverters I've found have been standard through-playfield mounted, and this isn't going to work on a game with WPC slide rails blocking that spot. I came up with a design with an overhead linkage that will lead back to standard coil brackets mounted to the backboard. Having purchased all the local hardware I could, combined with mechanical bits from the parted out boards, all that remains will be the diverter arms themselves, which will be part of my laser cut order.

Pictured: a huge mess.

bitmap (resized).png

Specifically, my Inkscape workspace in which the ramp walls and diverter bracket are taking shape. Had to do quite a bit of planning to make this work, as evidenced by all the random bits and bobs used as templates and measurements. Holes are cut for rivets, zip ties for GI and switch wiring, the diverter linkage, and the J-type speed nuts that will hold the school hallway wing plastic on top. The extended-height portion on the right will serve as a shield to prevent the ball from sneaking off into no-man's land in the upper left corner of the playfield, where the building is taller.

Tabs are sticking out for mounting the floor (standard light shield plastic) and the switch that will turn the diverter on if the ball is to be diverted. Towards the center is a bracket that will help support the right edge of the floor in a couple of places, and left of that is a structural support that can be riveted to the floor where it may be weaker. The diverter mounting bracket also includes a tab to attach the return spring, and the whole bracket can be bent in the opposite direction to assemble the diverter for the right ramp.

New plan. Excuse the lack of art (as is expected at this point) and the blocky VUK scoop.

Screenshot from 2020-05-13 22-04-26 (resized).png

I know you can heat up a sheet of PETG and wrap it into a cone. You can also vacuum form it into a volcano shape - like I foreshadowed in the Undertale build thread.

This:

... isn't how it is going to happen.

Thing is, I don't have the skills or equipment necessary to construct a complex shape like that, and actually make it look good. It also would be difficult to repeat consistently, which would be a problem if I built more than one game - not impossible considering the will of the UP Pinball Collective.

So I took a page from the Special Force book (used as an example, but many games do this) and tried forming a volcano instead out of stacked light shields. Even without artwork, it turned out surprisingly well. This method gives me much more fine-grained control over shape and height, with the added bonus that it can be assembled fresh out of the mailer from the laser cutting service with no unusual prep work.

In other news, a package from Marco showed up. In addition to some hardware for repairing targets on an EM, I threw in a few random pieces that I knew I would need for VB.

IMG_20200511_205826229 (resized).jpg

Contents:
- Rubber stopper grommets for ramp return drops. With these in hand, I could measure the size needed for the holes they fit into.
- PAL nuts for flipper buttons. Why not.
- The remaining hardware for the custom diverters. Only the laser-cut metal blades remain.
- Start and Extra Ball buttons, in theme colors. Why not.
- Brackets and associated parts for the left eject scoop. Just to make sure I have what I need.
- The resistors I originally ordered wrong, required for programming my ATMEGA328 chips.
- A VUK assembly. I can use this to mark the holes needed for the volcano and hot spring mechs, and test-fit the entrance saucer+VUK combo mech.

Two mentions of laser cutting and no mention of the progress? Hmph.

path4512-0-0-8 (resized).jpg

As you should expect by now, it's a big mess. Fortunately, it is separated into four layers for plastics, a layer for metal, a layer for playfield holes, and a layer for all the random measurements and chopping paths and other crud that helps bring this to life. I have an ever-growing text file detailing each piece, particularly the metal walls, to help avoid forgetting those small details that don't matter until you realize you can't build it without them.

Once the files are completed and sent off, it's time to decide whether to do art or PCBs next. Or any of the many other steps in this crazy project.

All the cutting for my previous build was done by Pololu. Shredder565 had good success with SendCutSend on his TAF build, so I'm somewhat on the fence as SCS seems to have pretty good prices.

Planning out the metal parts has definitely been the most tedious part so far. Fortunately I've at least created a few templates for mounting tabs, light cutouts, etc. that at least help with the simpler parts like guide walls. The rest of it is more complicated stuff that needs a regular dose of measurement to get right.

walls.jpg

And there are ten more pieces I need to lay out after these!

While planning out the subway+VUK+lock system under the volcano, I ran into a troubling conflict of mechanical footprints as I was moving the lock release into place.

conflict (resized).jpg

Back in the Undertale project, I would have only noticed it after attempting to mount this repurposed kickback assembly in real life, and then gone into emergency damage-control mode to try and rearrange things and make it work.

Thankfully, I have learned a few things since then, and the footprint conflict is obvious right from the diagram. The fix for this conflict (once the mech is in its final position and angle) is a laydown GI socket with topside wire access hole, and either adjusting the position of the rollover or replacing it entirely with a rollunder gate.

Thanks! Tried to make something engaging. We'll see how it plays!

A fair amount of my crap is 2D as well. Inkscape is pretty much my personal sandbox, with a healthy dose of measuring real-world parts, drawing shapes at odd angles to figure out depth, and attempting to visualize 2D objects bent into 3D shapes. Learning real CAD would probably be very helpful, but now that I'm nearly done designing a second game without it, I'm not sure I would actually need it anyway.

(It occurs to me that if I did a design that really was complicated enough to be impossible without CAD tools, it would probably also be a nightmare to construct IRL.)

The 3D bits are done with the help of Visual Pinball, and that was a bit of a lifesaver for a few parts. New volcano in particular.

Nice! I might have to hit you up for Pi tips. If it's powerful enough (and I can avoid the issues I had with the 3B last time), I might drop the desktop core for the cheaper option.

Presenting, the apron artwork.

Definitely not inspired by any particular brand or era.

apron (resized).png

Plastic and metal both almost ready to send in for a laser cutting quote. Basically just trying to think of anything else I might want to put on the order. We all know I'm going to realize what else I need just a few days (if not hours) after I send in the order...

I just supported my local Pinball Life.
eh not much (resized).jpg

Yes, Marco will get plenty of love too.

Meanwhile, I sent out a quote request with my completed cut sheets for 0.036" stainless and 0.06" PETG. Within a day or two I'll see how much that will cost, and most likely get my order placed.

I had a large area of wasted space in the metal cut sheet, so I decided to use it to make an engraved Yooper Pinball Collective sign rather than waste the metal.
cuts1 (resized).png

The plan is to build a second, more playable whitewood using these components. Once everything is known to fit together as designed, the metal will be polished and reused for the full game. Plastics intended to be clear (airball shields and such) will also be reused, and the rest will be printed and cut new.

What are these packages I just got in the mail!?

IMG_20200615_203426460 (resized).jpg

A bunch of stuff from PBL!

IMG_20200615_212115362 (resized).jpg

A few duplicates shown. Two backbox latches, two lockdown bars, a bunch of surplus leg bolts, and also a pack of EM fuses. Just saving on shipping for multiple projects and hobbyists - I'm not building two Volcano Blasts....yet?

I showed off a similar photo when a batch of parts arrived for UT a couple years ago.

When I showed it off at work, one acquaintance had to take a second look because he thought it was a photo of drug paraphernalia at first.

But was he wrong, really?

And the metal arrived from SendCutSend!

IMG_20200618_203553702 (resized).jpgIMG_20200618_203603800 (resized).jpg

Thicker than I expected. Most of Undertale's metal was from Pololu's .036" option. SCS didn't offer this thickness; they had .048 and .060. I picked .048 and am very pleased. This will definitely hold up to a ball.

The extra whitespace didn't apply, so I ended up separating the cuts a little more and removing both the emblem and the plaque from the submitted cut sheet. Cheaper that way too. Since SCS doesn't do plastic, I ordered my PETG from Pololu, and expect that in....not too long?

Tomorrow, it will be time for another trip to the contracting company for a print of the diagram for whitewood 2.0.

That is a brilliant comparison that I never even thought of.
(For the uninitiated, TMOHS = The Melancholy of Haruhi Suzumiya, starring a girl who has god powers but doesn't know it.)

haruhi_anime.jpg

The main difference being in this case, they know about the time loop from the beginning and are voluntarily causing it.

Combos are good for points, extra balls maybe, and attracting the attention of one of the girls in particular who apparently appreciates the effort of combos. You, as the player, are the unnamed visual-novel protagonist who also attends their class. It remains to be decided whether you have eyes. Canonically, there could have been any number of time loops between morning and apocalypse, but the game will depict three, one for each anti-volcano plan.

There are four steps to each plan. I'm weighing the pros and cons of letting the player play any step vs doing it only in sequence. I might make it an operator menu adjustment and see what people enjoy more:
Screenshot from 2020-06-21 22-59-29.png

Friday I took another trip to the contractor's office.
IMG_20200620_151659193.jpg

And picked up two pounds of 1/2" hex-head screws at work.
IMG_20200620_153626262.jpg

And it's show time.
IMG_20200621_164711607.jpg

I have already discovered a small handful of pieces that will have to be re-drilled if I'm going to get them to fit - and one very large and expensive piece that is so horribly measured, apparently, that I'm going to have to redesign it and have it re-cut. Yay.

Okay, so here's the scoop.

IMG_20200623_193106298 (resized).jpg

And another!
I'm happy about the fact that this is one single piece, no fasteners or welding necessary.

IMG_20200623_221725364 (resized).jpg

Now for the actual "scoop." Priority right now is going to be getting all the metal pieces finished and installed, so I can confirm if everything fits or if anything else needs to be re-ordered. But in the meantime, things are looking pretty good.

Diverter brackets:
IMG_20200623_221901255 (resized).jpg

Subway tunnels:
IMG_20200623_221922240 (resized).jpg

Anything sticking below the surface will have to wait until I can depopulate the playfield and cut out the necessary holes. I felt it best to leave everything as a solid board at first, just so that all the printed lines are intact and I know where everything needs to install.

Looking over the topside parts, I'm going to have to give tonight's work a rating of not-too-shabby out of 10.
pf (resized).jpg

Thanks!

I do have a few things not pictured, but nothing especially high-end:

- Bench vise
- Really small rivet clamp that only works for one or two of the rivets in the design
- Center punch for applying all the rest of the rivets
- Drill press, used so far to undo a rivet that shouldn't have been
- Two pieces of angle iron, mostly used as an extension of the vise but also used as a somewhat-effective-but-very-inconvenient bending brake when I was desperate
- Increasingly tattered rag used to avoid damaging the surfaces too much in the vise
- Bench grinder that will eventually be used to finish edges and shine up surfaces

Rail bending in action on the Vise Extender (TM) a.k.a. two thin angle irons.
IMG_20200624_174059476 (resized).jpg

Not a one-size-fits-all solution to these wildly inconsistent pieces, but for the volcano VUK scoop in particular, it was perfect.
IMG_20200624_181245_01 (resized).jpg

I might run some rivets into those holes at the top to make it look like they're meant to be. The brace bends over and rivets to the other side, but a couple of stray rivet holes made it into the design on the bend-only side. That said, I might wait until it's in a cabinet and see how much vertical leeway I have - those holes might be a good place for brackets for a "smoke and ash" plastic over the volcano crater.

In the home stretch as far as the metal, with just a few pieces to bend. These scoops copied the design from UT (before I thought to model a one-piece scoop as shown in an earlier post) and went together just as easily, which is to say getting the tabs to fit smoothly was very frustrating! Pictured: The scoops cooling down on the concrete floor after being spot-welded.
IMG_20200624_221115035 (resized).jpg

The rear scoop is the left eject, with a cutout for the subway from the volcano. The front scoop is the volcano tunnel, with an opening for the "nonviolent" lock release in the volcano. The scoops are not identical; the left eject scoop is deeper by around 3/8 inch.

You know what's really tedious and not at all fun?

IMG_20200628_135708654 (resized).jpg
Constructing a System 11-era pinball cabinet without a table saw.
Also, constructing a pinball cabinet with a circular saw which itself is from the System 11 era, and which can no longer cut straight.

Called it quits for tonight after I badly mismeasured the two sides of the neck, cutting one too large and one too small.

On the other hand, most of my pieces are cut, and more or less decently squared off. Cabinet sides, backbox sides, cab front, cab bottom, b'box top/bottom, two backbox shelf pieces to be adhered together, backglass lock trim, and the front and back of the neck box. Remaining parts are the two sides of the neck, backbox rear panel, speaker panel, and perhaps some internal bracing.

IMG_20200628_163034829 (resized).jpg
The rails are the 51-5/8 version as used in Bally cabinets post-1988 until they switched to the WPC-style cabinet design. Perfect fit so far. Finally bought myself a router, so now I don't have to find and borrow someone else's in order to join the corners and install glass channel.

And the plastic arrived from Pololu! Still has the cloudy protective coating, but that just makes it easier to see how the shapes turned out. Although I don't have the spacers installed yet to give it proper depth, post #27 has now crossed into the physical realm:
IMG_20200629_173731989 (resized).jpg

Front panel is cut!
IMG_20200705_091857606 (resized).jpg
The plunger hole was cut after the photo was taken, and confirmed to fit.

Stages of cabinet construction are planned as follows:
1. Cutting.
2. Routing.
3. Drilling.
4. Assembly.
5. Sanding/filling.
6. Prep and paint.
7. Decals and trim.

Step two is done, three is in progress.

So far, everything seems to be fitting together PERFEC---
oh bollocks.
IMG_20200705_113814159 (resized).jpg

Yep, left a space on the plan to make sure the slide rail brackets wouldn't interfere with the playfield mechs, but just conveniently forgot that oh, there are also pegs sticking out of the cabinet that you also have to avoid.

A bit of hacksawing and wire-wheeling followed this realization.
IMG_20200705_122416362 (resized).jpg

Although it won't be quite as convenient to install now, shortening the nut does indeed make everything fit together as intended.
IMG_20200705_122758979 (resized).jpg

I'm getting better at assembling cabinets.
IMG_20200707_205159225 (resized).jpg

Mistakes were made, of course. Back corners routed wrong (luckily re-routing was an option, more or less), backbox back panel joint routed too deep, neck spacer not perfectly aligned - but everything can be addressed or worked around.

IMG_20200707_213652255_BURST000_COVER_TOP (resized).jpg

Tonight was a night of glue and little else for the next 24 hours. Pictured:
- Adding a mating brace to the right side of the backbox, to address a routing error mentioned above - far left
- Assembling the neck spacer after predrilling some holes to attach it to the cabinet - center
- The lower neck pieces being adhered together for strength and to create a wider target on which to mount the rear glass channel - under the heavy wood blocks
- The cabinet itself, clamped together with ratchet straps and heavy tools - back

Tomorrow I will be bracing the cabinet with 90-degree corner brackets and standard leg plates, assembling the neck and attaching it to the lower cab, and most likely test-fitting it with legs and trim. Backbox assembly and leveling will also be an important step.

Up on legs!

IMG_20200709_071331907 (resized).jpg

Spent a lot of time last night with prepping the cabinet to drop the playfield in, and didn't end up having time to put together the backbox (so no glue to dry for now). Tonight's task will be cutting some braces, routing the top and bottom surfaces of the backbox for T-moulding, and getting the backbox assembled enough to glue.

Playfield fits PERFECTLY on the slide rail mech.

Backbox not fastened down yet (balanced precariously for this photo).... but we're getting somewhere now.

IMG_20200709_232114847 (1) (resized).jpg

Not pictured is the two pieces making up the upper trim that holds the backglass in. Still being glued together, and those will be fastened in when dry.

Bottom of the backbox is shallower than it probably should have been - didn't think to compare the depth with the neck - but I've come up with a solution involving the speaker panel that I think will mask any obvious underhang.

Plastic test-fitting.
IMG_20200712_143356449 (resized).jpg

Not perfect, of course. A few pieces were designed to fit around bent metal walls, and some walls couldn't be bent as tightly as I wanted. But otherwise it's starting to look more like the plan.

Meanwhile, Phase 4 of cabinet construction has been completed.
IMG_20200712_103627367 (resized).jpg

Backbox hinging is near perfect, playfield slide rails nearly perfect, and generally everything fits just about the way I wanted. The only thing that really stands out to me is the playfield volume is about an inch deeper in the back than I expected. I'm not sure how this even happened, with all the measuring I did ahead of time, but at least I know I have enough clearance to do a Fire!-style smokey shield over the volcano crater.

Over two weeks now with NO updates! What could possibly be my excuse this time?

Well, I mean... I bought a house...

IMG_20200729_195642736 (resized).jpg

Still have to get a whole bunch of stuff moved and unpacked. In particular, the pins. VB work will resume after things settle down.

In the meantime, Jesse - audio enthusiast of the Collective - is working at composing the game's soundtrack, and made me some recommendations for the physical side of this game's sound system. Unlike UT, you'll have no problem hearing this game at shows.

I've spent so much time on house-related stuff, etc... have I done ANYTHING on this project?

Well...yes, yes I have.
vb playfield.svg (resized).png

I continue to maintain that I am not an artist; the front-and-center stars of the show were drawn by my girlfriend on a piece of high-tech artistry software, while the rest was done by myself in a combination of Photoshop and Inkscape. But aside from a few tweaks and a few more details I'd like to add, I think it turned out pretty good.

Our cast:
Hinata Akiyama, the science girl, leads the club and can cast energy spells.
Saki Kusachi can influence earth and stone.
Yumi Shimizu can manipulate water.

A handful of mechs are installed on Whitewood 2 now, with a few remaining to be transferred from the original and a couple more just needing their hardware fasteners. The shooter lane ramp has a bit of a ball-hopping issue I discovered with my phone's high-speed camera function, which hopefully won't be too difficult to resolve.

On the software side, I ported and rewrote the scheduling and event queue functions from Undertale, hoping to address some very rare "I'm just never gonna give your ball back" issues I've seen on that game. I've added more advanced sprite drawing functions as well as queueable effects such as fading, shaking, and sliding in or out of frame, testing it out on some "visual-novel-esque" scenes to make sure they are an authentic weeb experience.

It was also time to rethink that logo. Sure, Post #15's background smoke effect looked decent for real smoke. But real smoke has no place in a colorful anime.
Screenshot from 2020-12-04 19-22-14 (resized).png

The black space on the right is reserved for debugging and visualizing playfield lamps and switches.

I'll admit, I'm proud of my new logo. I designed a particle effect to work with only two colors (dark gray and darker gray), and configured it to update at a limited frame rate, as would be expected for typical anime movement style. The reduced load on the graphics system, due to the intentionally low frame rate, actually works so well that my hotheaded work PC's processor doesn't even break a sweat rendering it. And the lack of prerendered frames keeps the program size down as well.

These nylon snap spacers were a cost cutting effort by Williams in the late 1990s.
IMG_20210112_200141644 (resized).jpg

Don't design a new project with them. They suck.
I will be replacing these with mach.screw+sleeve+locknut on the final build. More solid, more reliable, less breaky as you struggle to squeeze them on.

In the meantime, I put myself on a standard of "try to work on VB just a little every day" - and accomplished a relatively important milestone:
IMG_20210112_200211358 (resized).jpg

This is a list of every set of hardware fasteners (as well as rubber rings) that mounts into the playfield. Though I did have a bit of hardware left over from my last build, and more from parts playfields, it still didn't cover everything. This posed a problem, with me not knowing exactly what else I would have to order to complete the whole assembly.

Now that I know for sure what I need, I can take inventory of what I have (polishing as necessary), and acquire whatever is missing.

Also pictured is a bag full of the cheapest playfield inserts I could find. Of course, they aren't the final ones, but they'll allow me to test-fit lamp boards to know for sure that the circuit boards will fit. Speaking of which, I really need to get moving on that PCB design.

Having finished the list of post-and-screw combinations, and the map of where they go, I set out to distill them down to a raw list of individual hardware fasteners.

Screenshot 2021-01-15 12.31.56 (resized).png

I think I may have realized why this thing sat for so long - it's a huge drain on motivation when you don't have all the fasteners, and you aren't even entirely sure which ones you need to buy. Fortunately, I now have an itemized list, so - following some tumble polishing and the weeding-out of the rejects - soon I can head down the hardware aisle and fill a bag with screws of all kinds.

While working on this task, I decided to rework another area.
IMG_20210115_070507048 (resized).jpg

The school entrance shot (upper left VUK) is straight through the courtyard (bumper area), but as I got down and looked at the angle, I got concerned that the post next to the bottom bumper could make the shot harder than it needs to be. There is also no effective way to get the ball into the bumper action from the right flipper, despite having a clear shot right through them. So I removed the post, removed a target that really had no rule plans (or associated lights!), and stretched the rubber ring a bit further back.

After a bit of experimenting, I found a rubber angle that will bounce the ball into the bumper zone if you aim too low. This layout also adds a bit of potential randomness as the bottom bumper is now more exposed to the rest of the playfield.

The metal cutting was done by SendCutSend's laser cutting service, then bent at home with some basic tools. Designs were done in Inkscape, measured using path ruler tools on an exported blueprint from Visual Pinball. Mistakes were made, but in general it turned out pretty well.

Best of luck on your project - hope to see it here sometime!

Having completed the list of hardware, as well as a quick inventory of everything I already had on hand, I took today's lunch break to go on a shopping spree for bolts and screws. Couldn't get everything - I have no local pinball stores, after all - but I did manage to nearly complete the list.

As far as the inventory on hand, a lot of it was pulled from those parts machine playfields in page one. Only thing is, they were in pretty bad shape. Lots of tarnishing and rust, very little shiny.

But everything changed when the tumbler nation attacked.
IMG_20210121_180757646 (resized).jpg

After following Pinside's advice on which tumbler (the gray one with the black bowl), which media (walnut), and which polish (Mother's Mag and Aluminum), I fired up the polishing device for the first time in my life, and man, have I been missing out. My restorations are about to get a leg up, too. Anybody on the fence about trying tumbling, get off that fence and buy one. Pictured above is 24 hours of hands-off work.

The mounting holes in most of my steel walls ended up being just a tiny bit too small - 1/64th inch, to be exact. It may be related to how SendCutSend processed the files - the hole edges look a bit faceted, as if my vector was converted to a polygonal shape prior to the cut. I corrected for this with a small updated batch (including a new upper right ramp wall and a deflector for the hot spring scoop), but most of my shapes still had the slightly smaller holes.

Getting the playfield together this way was a bit of a struggle, having to find screws on the small end of their design tolerance and even then sometimes having a lot of trouble forcing them through the holes. As I took everything apart for hardware inventory, I decided this would have to change. When the screws grab the metal that bad, it would be much too easy to accidentally scratch a final cleared playfield.

Despite the material, and any hardening that may have been caused by a laser, a 9/64" cobalt bit enlarged the mounting holes with no problem at all, although doing it for almost every single metal piece was a bit tedious.
IMG_20210123_105818116 (resized).jpg

I tested with an actual screw, a very stubborn one that took a pair of vice grips to extract from the hole. No binding issues at all.

As long as I had all the metal in close proximity to the bench tools, I also took a wire wheel to all the visible surfaces I could reach. Although a bit of manual polishing would be of further benefit, they're already looking quite a bit shinier - unpolished one on top:
IMG_20210123_112514272 (resized).jpg

Nexy doesn't care much about this project, but Clark urges me to continue working:
IMG_20210123_141726389 (resized).jpg

And work, I did - over this weekend I finished drilling all the holes in playfield prototype number two. With everything known for sure, I can update my hardware notes, and from there, my Inkscape illustration of where every hole needs to go and how big. Having assembled everything at least twice will also help by showing me where the playfield needs to be dimpled for wood screws, rather than taking a guess on a final assembly.
IMG_20210124_145434858 (resized).jpg

Unfortunately, mistakes were made, as always:
IMG_20210124_115355336 (resized).jpg

I also finished the tedious task of drilling 55 insert holes. The inserts are the cheapest ones I could find in bulk - opaque white 5/8" circles at a quarter a pop. Typical practices were followed - drill a clean hole to the depth of the insert with a forstner bit, then drill a slightly smaller hole the rest of the way through, giving the insert a shelf to sit on. These are obviously not the final shape and color desired for my design, but they accomplish the goal of being in the right place to appropriately test-fit lamp boards.

Once all the cutting and drilling was done, as far as I can tell right now, I sprayed it over with a layer of grassy green from a rattle can. By no means would I consider it a perfect paint job, but it seals in the wood a bit and gives the prototype a bit more color than the same raw wood and white paper that I've been seeing for all this time.
IMG_20210124_150641851 (resized).jpg

Given a day or so to dry, I will glue in all the inserts and hit it with a layer of clear. It's not going to be a good clear either, but hopefully it will make everything flat and smooth enough that play testing will be a pleasant experience.

The 5/8" bit hole seemed to fit the 5/8" insert decently well. That said, I would definitely suggest doing test cuts for any size insert you want to drill for, just to make sure.

Inserts going in.
IMG_20210125_174452968 (resized).jpg

Inserts in.
IMG_20210125_180150427 (resized).jpg

My forstner drilling was definitely not perfect. A few don't lay completely level and a few are a bit raised. Hoping that some sanding and clearing will at least make it not completely bad. Worst case is I'll just pull the most raised ones and grind out some wood or sand down the bottom edge of the insert. Generally not a big deal as, again, not the final playfield.

In hindsight, clear inserts probably would have been a better choice, but again, not the final playfield.

This weekend was busy.
IMG_20210130_211835939 (resized).jpg

Very busy.
IMG_20210131_190339193 (resized).jpg

The playfield went from flat and light, to not flat and very heavy. My choice of relatively low-density birch plywood definitely didn't hold up on the bottom side, but again... not the final playfield.
IMG_20210131_152044935 (resized).jpg

I think I'm going to seek out better pop bumper lighting, as this lights sit a bit crooked, apparently can interfere with the skirt, and don't light things up quite as bright as I would have liked. But it generally looks pretty good. Anybody going to hazard a guess why I chose black skirts, yellow bodies, and these particular cap colors?
IMG_20210131_163234493 (resized).jpg

Aside from a few parts such as the slingshots, I've nearly reached the limit of what can be built with the hardware I have, so I'll be placing another order soon. With enough playfield hardware from Marco to make up the missing pieces, and fresh new clear silicone rubber from Titan, I should be able to nearly complete the topside assembly.

After that, I'll likely go back and do some serious PCB planning. Those lamp boards aren't going to design themselves.

I didn't get a huge amount done today. But I made sure to spend at least a little time on it. It would be a sin to not make progress on Groundhog's Day, on a game themed around time loops.

Ball trough is now fully mounted, screw hole locations perfected. I was very happy to see that when using the autolauncher kicker with the small footprint, there is miraculously enough space to cram it in alongside a traditional ball trough AND under a playfield support rail without any positioning conflicts.
IMG_20210202_214607916 (resized).jpg

Thanks guys!

ALL under-PF coils have now been mounted. Still need a bit of hardware - Marco ordering will commence within a day or two - but I'm feeling pretty satisfied with my footprint pre-planning.

Pictured: The volcano scoop, with three balls locked inside!
IMG_20210205_213155691 (resized).jpg

I've changed my mind once already about what type of display this game will use, and now the supply chain made me change my mind a second time. ("It's not my fault! The supply chain made me do it!")

Up till now, I was planning on mounting a ~17" monitor partway behind the speaker panel, which would leave a DMD-shaped area of screen visible through the backglass. But while shopping for PC components on Newegg, I was rather surprised to learn that 17" monitors, with the appropriate mounts and connectors, can be surprisingly hard to come by for a good price.

But you know what's actually cheap and plentiful?
27" monitors.
IMG_20210221_152954488 (resized).jpg

I wasn't planning on it originally, but I do think this JJP-size screen will catch a lot of eyes once it's showing off visual novel art and anime girls. The photo above was a test fit, and I'm in the process of improving the mounting system to be a bit easier to work with.

The monitor is mounted through the VESA holes onto a piece of 1x6. After taking this photo, I realized I still need to cut access holes for the cables:
IMG_20210223_182341913 (resized).jpg

The board attaches to a couple of runners that I've mounted to the back corners of the backbox. I could have just measured the right distance to the back and mounted accordingly, but I felt that adding spacer blocks touching (and glued to) the back wall will improve durability somewhat. This change in backbox design meant the knocker would no longer fit in the upper right corner where I had it, so I decided to switch things up and put it in the middle.
IMG_20210223_183900847 (resized).jpg

Moving the knocker down from the roof allows service without needing to remove the small backbox lamp panel (above monitor), and I believe that slamming the knocker plunger into a solid 2x4 will produce a more satisfying thud than simply hitting the not-so-dense birch backbox roof.

Orders have arrived from Marco and Titan, awaiting installation. I was going to go with clear rubber, but a couple sizes I needed were either out of stock or not available in clear, so I stuck with traditional white. This is fine to me, as I always did like the look of white rubber. Marco was out of post studs - of which I need about two dozen more - so I'll be placing another order with PBL soon for those and for some deeper GI sockets for the backbox insert panel. I think I will stick with static lighting on this backbox, since most of the space is taken up by the monitor anyway. The bonus of this is that I can wire pairs of GI bulbs in series (running each at half voltage) and power it off the same 12v rail as the monitor. This means the only wires I need in the backbox are knocker 50v/driver, 12v/ground, and the HDMI cable.

What the VUK is this?

Granted, I manually actuated it for that short video clip. But this confirms what I very much hoped, which is that the hot spring VUK would fire smoothly and not take any bad bounces from rebounding off the wall.

Then I printed off a letter-size snippet of the playfield art, just to see how it looks in real life. Parts orders arrived and I spent a fair amount of time installing things. Aside from some plastics sitting freely to check their fit, everything here is fastened down.
IMG_20210306_210818858 (resized).jpg

I also swapped out the flipper buttons for springless versions, and they are now smooth as silk to press. The PBL product page warned that they "will not work in Bally/Williams machines that use fliptronic opto boards." Like Undertale, this game is using aftermarket Williams fliptronic opto boards. I decided "Challenge accepted!" and bought them. Worked fine.

The frick, Mouser? You almost gave me a heart attack!
part1 (resized).jpg

Turns out, ATMEGA328-PU and ATMEGA328P-PU are two different parts. Thankfully, they still have plenty of the correct chip.
part2 (resized).jpg

I was slightly concerned for a while that I had fried one of my test chips (not that they're particularly expensive). Thankfully, plugging everything in today confirmed everything was fine, and after taking a moment to remind myself how to program them, I now have a two-dollar chip blinking a bright white LED with a PWM fade as smooth as an incandescent bulb. At last I can throw in my communications code and start turning these things into playfield controllers.

First thing to be completely DONE on this game is... the backbox lamp panel.
IMG_20210424_111401718_HDR (resized).jpg

Pictured: Clark inspects a corner bulb with curiosity, wondering why it's a warmer color than the others.

I decided my backglass would just be the game's logo along with some smoke and burning cherry blossoms - I'll post the art when I consider it done. Since the main eye-catching details will be drawn on the large monitor instead, I didn't feel the need to do anything special with the lighting. In fact, I decided, why not just leave it always-on, and wire the bulb sockets in series pairs, so the GI can run off the same 12v rail as the monitor?

The backside wiring needed a bit of extra thinking to accomplish this, but it worked out.
IMG_20210424_111315987 (resized).jpg

Why so extreme? Just another case of "but I wanna tryyyyy." I like the idea of having just a bundle of six wires (plus an HDMI cable) going up to the backbox. It feels nice after the dozens of wires you have to deal with on the older games that make up most of my collection.
YEL/GRN: 12v supply for GI and monitor
BRN/BRN: Speakers
BLU/BLK: Knocker

A "Blast" from the past, from the Undertale thread:

Well, my friends, I think it's safe to say I learned how to use KiCad. Submitted for your approval: the entire Volcano Blast boardset, less the PC mobo.

IMG_20210803_182624539 (resized).jpg

Pictured is about $165 worth of stuff, a surprisingly good deal IMHO. OSHPark does $5/sq inch and sends three copies of each board (which I actually had to ask around to confirm because apparently I lacked reading comprehension). So I did a lot of design tweaking to make sure as many boards can be duplicated in multiples of three as possible. There are only three small boards left over (two Teensy-LC Hub Boards and one Single RGB board).

This was an exciting order to get in the mail (as well as the four separate orders of parts from different suppliers with which to populate the boards). I'm especially happy with the flipper driver board (bottom left, and two stacked near the right). It's the size of a credit card and includes five solenoid drivers - two for authentic Fliptronics functionality, one reflex device (slingshot or jet bumper), and two coils free for anything else. It also throws in a ULN2803A driver chip that directly drives eight lights. The auxiliary solenoid board (far right) is a slightly shorter variant of the flipper board, supporting any five solenoids and replacing the lamp drivers with breakouts for more switches.

The single Teensy Hub board houses the delightfully cheap Teensy LC, along with a lamp/logic power connector and a set of header breakouts that can branch off to each board that acts as a controller. The remaining boards are (primarily or otherwise) playfield lamp boards. Some of them, such as the Duo+RGB board (top row) and the Trio Lamp Board (stacked in bottom center) include ATMega328P chips that handle lamps on these boards and others connected downstream.

Perhaps you may wonder what actually took me so long since my last update. Well, besides annoying time-wasting things like real life, there was a lot of testing to get done behind the scenes before I could even start designing these boards.

The first thing was board-to-board communication. The Teensy has a built-in USB board, which straightforwardly solves upstream communication to and from the PC with code I can directly reuse from UT. Getting Teensy to have a conversation with an ATMega was slightly more challenging. My earlier proposed comm protocol ended up being a bad idea, so I mixed it up a little and threw in another pin for upstream clock.

Now the Teensy board will set on/off for the next bit in a batch to all downstream boards at once, then fire off a clock signal (to which the boards respond by sending back their own next bit, from switches). It can then read the results sent back and move on to the next bit. Once all bits are sent (I figured 32 ought to be enough for anyone), it pauses for longer than a max pulse delay, resetting all boards to expect bit 0, and restarts from the top. The entire process of sending and receiving 32 bits takes less than 10ms, guaranteeing no noticeable delays in computer response to playfield events or vice versa.

While perfecting communication, I got a sobering but welcome reminder that the ATMega328 is an 8-bit chip, not 32-bit, and it does NOT have integrated floating point. Meaning that my "incandescent fading" code for lights, which was initially written with decimal intensity levels in mind, was absolutely devouring clock cycles. After a total rewrite that paid much more attention to data sizes and bitwise operations, everything started running a lot smoother, even PWM code.

There are plenty of other features I worked on as well, including light dimming at half brightness (useful for adding more RGB colors for that handful of lights, as R/G/B are handled as three individual lights), reflex devices with built-in protection against stuck switches, most of the features expected from Williams Fliptronic flippers, and switch debounce. I wanted to make sure I'd physically tested every feature on a breadboard before committing anything to PCB. This definitely took some time, but it gave me peace of mind.

Finally, I'm proud to report I flipped a ball around! One of the last tests I did before laying out the final board designs was fill up the solderless breadboard with transistors, program a 'Mega for two Fliptronic flippers and two reflex slingshots, and wire it up. I don't have especially good video footage of this test, but it felt AMAZING to slide the glass back onto the game, plunge a ball onto the field, and flip it around with mostly finalized hardware and firmware. I like to think that the science girl, Hinata Akiyama, would be proud of the progress.

Speaking of which, the other major time sink has been the writing of Volcano Blast's entire backstory as a light novel! Currently I have it written all the way up to the point where the game begins, covering the origins of the girls' elemental magic, lots of little details that will be thrown into art and callouts, and justification for a bunch of shots and gameplay mechanics. The game is basically good for all audiences but the light novel is a tad bit more mature-oriented. I will be making it public closer to when the game is finished.

Well, here's a bit to get you started! I'll probably release everything eventually, but here's the schematic and at least a board screenshot of the flipper driver board, the one I'm the most proud of. The mounting holes use the MountingHole_3.7mm footprint, and are exactly the right size to put a standard playfield wood screw through.

Screenshot from 2021-08-03 19-50-16 (resized).pngoutput.pdf

Edit: For some reason, the pins on the lower right of the ATMega in this schematic are all getting connected together in the diagram. Ignore that.

Okay, that's interesting. I welcome any and all such unsolicited advice - even for things too late to change, it's probably something I can apply in another project. Given that this was literally the first eleven designs I've ever done in KiCad and there are probably many tricks I don't know, it was indeed helpful for the future. Thanks!

Fall - and cold weather - is coming, and for those of us without a dedicated paint shop, that means one thing.
GET MOVING ON THE PAINT. Time's running out.

Given this time deadline, I decided it was a top priority to get this cab painted, regardless of how done the playfield is, just to make sure I'm done needing the warm paint shop called "outdoors" before I lose access to it. So, I pulled the playfield out, stripped the cabinet of all trim and hardware, and brought it out to the garage. First order of business was sealing up the rough joints, which I did with some fiberglass filler.
IMG_20210913_195317818 (resized).jpg

I wasn't sure where to get glass fiber additive, so I took the recommendation from a playfield repair thread, and thickened up the mixture with sawdust instead, as I had a whole bin of the pine variety ready for just such an application. I'm not entirely sure how well this concept turned out, as the resulting coverage definitely had some hills and valleys, but after a laborious evening or two of sanding, the cabinet joints looked and felt pretty good.
IMG_20210913_195327245 (resized).jpg

After a thorough wipedown of the cabinet (during which I managed to forget the exterior, spoiler alert), I laid it on a rolling cart to inspect and/or admire.
IMG_20210916_192838323 (resized).jpgIMG_20210916_192904967 (resized).jpg

The backbox is not ready for painting; there were a couple of visible plywood voids left, which will take 24 hours to cure after filling, but the cabinet seemed to be ready to go after a brief modification to the plunger mounting hole so I could move the ball shooter 1/16" to the right.

Confident enough in the lower cabinet to move forward, I retrieved the rattle can primer and shot a layer.
IMG_20210916_201132162 (resized).jpg

It's decent. Nowhere near a HEP job, but at this point it's not too bad for only being my second cabinet build ever. There was some dirt/cobweb or something stuck to the left side, which I'll have to take out in the sanding process. The rear side of the backbox neck is definitely not perfect - a tad bit lumpy as I had a lot of trouble sanding down the fiberglass - but in general it doesn't look too bad.
IMG_20210916_201203204 (resized).jpg

The flat square cardboard box in the upper right of this photo is my speaker/vent grille, which I finally managed to secure by way of Grainger:
https://www.grainger.com/product/GRAINGER-APPROVED-Sheet-5PDH0

I chose a gloss black as a base, which I'll be applying all at once when cab and b'box are fully primed. I'm aware that gloss highlights the defects, but I think if the non-decaled areas end up looking bad enough in plain gloss black, I'll try throwing on some hammer texture on those surfaces. Currently awaiting a quote on the cabinet art decals. I think I'll wait to reveal the art until it's fully installed; it's no 90s Williams package but I think it turned out decently well.

Backbox primed, at least the surfaces that matter the most.
IMG_20210917_192901636 (resized).jpg

Both cabinet and backbox use the primer sparingly, to reduce material costs and decrease the amount of work time required to - oh who am I kidding, I bought one can of spray primer and knew I was gonna frickin run out.

The speaker panel didn't make the primer cut list, as I feel it wouldn't matter so much since most of its outer surface area will be hidden behind speaker grille.
IMG_20210917_201349960 (resized).jpg

I'm still happy I managed to find the stuff, but it was a serious challenge to actually get it cut. I'm not sure what the appropriate tool would be for that job, but I was stuck with a pair of aviation snips, and it was, um, lots of fun. However, I did successfully get all the pieces cut out to the desired sizes. Pictured, bottom to top: Speaker grille, bottom front vent intakes in cabinet, top rear vent exhausts in backbox. The square piece alongside them is the grille for the cabinet speaker.
IMG_20210917_201327177 (resized).jpg

With priming complete, I think we're on track for a Saturday of painting.

Cabinet is painted. Photo came out slightly blurry, but I'm probably not going to retake it. After all, the blur hides the flaws.
IMG_20210919_190452350 (resized).jpg

Saturday saw the first coat, today was all about sanding and recoating. Like Undertale, I chose the brush-on variety, and given that even the second coat had some runs, I think it was probably a bad choice and I'll probably never use anything but spray for a cabinet again. But I think I can work with what I've got here.

Backbox:
IMG_20210919_190539626_HDR (resized).jpg

I used a much smaller brush to make sure the tiny details at least got some paint - mainly the button recesses and the edges of cabinet holes. This photo makes the edges look particularly bad, but I feel like it looks a bit better in person.
IMG_20210919_190525298 (resized).jpg

All that remains are a few other minor bits. Bottom of cabinet and backbox, mainly. I chose to paint the bottom surface of the cabinet because this seals it up and keeps finger oils from seeping in from moving the game - or even being seen at all, really. It will be interesting to see how the finished cabinet bottom looks, with the contrast of the silver carriage bolts holding on the lower speaker and tee nuts holding on the shaker motor. The whole thing should be fairly well sealed up, with all the speaker and vent holes covered up with freshly painted grille:
IMG_20210919_190511730 (resized).jpg

The finish is more even than the picture makes it out to be, as I chose flat black for the grilles but had just sprayed it recently.

IMG_20210919_190548688_HDR (resized).jpg

Paint is done - and not a moment too soon, as this morning I had to throw on a jacket on my way to work.

IMG_20210921_193900845 (resized).jpg

The speaker panel's paint looks a little rough, I'm assuming because I didn't prime it, but once covered with the grille I don't think it will look too bad. All the generally visible surfaces were already painted as of yesterday, but in the evening I finished up the bottom surfaces of both the cabinet and the backbox. All told, I used up almost an entire quart can of paint on this, with just a little left over that will probably be used for the playfield backboard.

Painting is done and dry, cab and backbox brought back inside to warm up!

IMG_20210922_202419723 (resized).jpg

The edges were left unpainted so that the relatively thick paint wouldn't clog up the routed channels that the T-moulding is supposed to fit into. This will all be covered up once the decals are installed and the moulding is in place, with red on the sides and silver across the top edge.

I'm really happy with how the speaker panel ended up turning out. Even if it's not totally perfect, everything measured up well enough. I didn't think quite long enough before drilling the lower screw holes in the grille, and had them right in the middle of where the panel is routed out (and thus only 1/4" thick) where it meets the backbox floor. However, simple fix: drill pilot holes all the way through and use those five screws to help secure the panel itself in place!

Local sign shop can always be counted on to run a few days past estimated date of project completion... but their colors are gorgeous.

IMG_20210930_145722320 (resized).jpg

Super excited to get my cabinet put back together with shiny new side art. Full cab art package reveal once it's all buttoned up again.

Phase Seven of cabinet construction is complete, and I think it turned out pretty darn good.
IMG_20211001_220948879 (resized).jpg

Application process was pretty standard, and those who have applied cab decals before are already familiar.

Decals laid flat. (That's "Volcano Blast!" in Japanese.) Nexy examines my painter's tape and approves its use.
IMG_20210930_182047818 (resized).jpg

Test fitting on the backbox. I thought I laid out the artwork on a canvas slightly larger than the cabinet panels for coverage purposes, but unfortunately I must have forgotten this step, as most of the decals were an almost exact fit.
IMG_20210930_182232395 (resized).jpg

To make sure everything went on straight, decals were done in halves. Place still-backed decal on surface, tape down one half of it, pull and peel the free half and cut away backing, stick it down.
IMG_20210930_214529694 (resized).jpg

Backbox left side fully applied.
IMG_20210930_183658002 (resized).jpg

Cabinet left side fully applied.
IMG_20210930_225309079 (resized).jpg

Remember how I said the decals were an exact fit? Well, that came back to bite me on the right side of the backbox, when the decal shifted as I taped it down and it ended up going on just a hair crooked. MEASURE TWICE PEOPLE

T-moulding went on pretty smoothly as well, along with the standard trim package. I will want to polish a few things up, such as the lockdown bar receiver, but otherwise it's coming back together well.

Thank you!

The logo text was drawn out on paper with jerky hand movements, then scanned and traced in Inkscape. I like how it turned out.

Fresh new-ish apron for post number 100!
IMG_20211005_205627757 (resized).jpg

Already had an apron, the Riverboat Gambler piece you've been seeing on the whitewood. Just needed to spruce it up! I removed the operator sticker from the apron and trimmed off the shooter lane prongs from the shooter gauge (since the autolauncher fork holds the ball and would physically conflict). A spray can of Goof-Off made short work of the apron screening, but I would have to paint over the shooter gauge artwork; it did not succumb.
IMG_20211003_182747089 (resized).jpg

After a thorough cleaning involving Goo Gone and a garden hose on jet mode, I shot both sides with gloss black rust-blocking enamel. Like most paint, it started off with prominent orange peel texture that smoothed out as it dried.
IMG_20211004_174240127 (resized).jpg

Alongside the cabinet decals, I also threw in the apron stickers on my decal order. After a bit of trimming to remove the faint edge lines, they were ready for a perfect System 11 ripoff.
IMG_20211005_185002349 (resized).jpg

Since then, I designed a new sticker for the operator adjustment switch bracket, which I'm sure I will post once the bracket itself is polished up. Around the same time as the apron painting, I hit the disassembled lockdown bar receiver with a can of metallic silver, which also came out looking nice.

And this is how the game sits today.
IMG_20211015_073216384 (resized).jpg

I picked up some cabinet guards from PBL to protect the interior paint as I dropped in the playfield, and indeed they worked well for their purpose. I didn't fasten down the apron or shooter gauge, but I laid them in place to check the color scheme and it seems to come together nicely. The playfield slide rail hardware had been tumble polished, but what really gave it smooth action was a thin layer of grease. No more horrible screech as I pull the playfield forward for service!

At this point, I will be giving the project a rest for a few weeks. Besides needing to do some cleaning (as probably evidenced by the photos), there's only three weeks left until Midwest Gaming Classic. VB wouldn't have been ready in time for it, but I have another popular homebrew project that needs some love and maintenance before it hits the show.

So expect a necro of the Undertale thread soon!

The last few days, I took a mighty plunge and soldered a lot of through-hole joints. A LOT.
IMG_20211127_195153144 (resized).jpg

With card table space finally cleared up, I arranged everything approximately as it will appear on the playfield. Admittedly I'm expecting a bit more wiring than I'd initially hoped with this PCB arrangement, but in general I don't think it's going to be too messy-looking.
IMG_20211127_210717875 (resized).jpg

Whether I get the playfield pulled out tonight and the boards mounted is uncertain. Left a piece of padding sitting in front of the game and my fluffy baby melted on top of it. Removal of playfield would disturb his beauty sleep.
IMG_20211127_205840177 (resized).jpg

PCB modules now installed on the game. Well, all but two of them.
IMG_20211128_144801819 (resized).jpg

That would be because some idiot decided to assemble the two single-RGB boards with the header pins facing the wrong direction.

The rest of last night was spent on a spreadsheet of devices and connections, making sure everything on the playfield was accounted for and listing every wire connection that must be made. It took quite a while but will be very worth it, both for assembly and for addressing in software.

One pop bumper switch had to be moved, due to a positioning conflict with a lamp+driver board. I'm glad I waited to wire everything up, because this could've been hard to move depending on how I may have chosen to route the wires.

IMG_20211128_141321955 (resized).jpg

Remember the lit playfield photo? It was made by hanging thin wire from lamp sockets just to see how things would look.
Not this time.
IMG_20211205_195550864 (resized).jpg

This is after soldering down what I hope to be the final lamp wiring. Seven of the eight lamp drivers are used on each of the lower two flipper controller boards, and now the lamp wiring for both of them is done. Well, except the GI socket that illuminates the Hot Springs sign, and the trough light. But at least there are hanging connectors for when those components are added/wired.
IMG_20211205_195839284 (resized).jpg

I actually had a bit of an oversight when I was doing the left side, and included a bit of extra GI sockets in what was originally intended to be a very short string. But I'm considering this to be one of those mistakes that's good in the end, because it's still a reasonable number of sockets and it actually frees up a lamp driver so I can move the orange flasher (by left scoop) to a more sensible driver board.

Most of the wiring could be done from a normal seated position. Clarky-boy liked that.
IMG_20211205_151955751 (resized).jpg

Playfield lighting and communication channels are nearly all wired up, with the exception of jets, flashers, and four lamp boards for which I didn't have enough connector housings. Order is on the way.
IMG_20211212_205633813 (resized).jpg

I think I will run wires towards the bumpers but leave them hanging for now. Bumper lamp sockets are not fun to deal with, so it would save me frustration to just leave them disconnected until the final playfield assembly is underway. Flashers don't need to be wired immediately - especially since the sockets are not yet mounted topside - but I will leave some Molex connectors hanging near their future access holes so they can be easily plugged in once installed.

Somebody photobombed me at bottom left!
IMG_20211212_205658254 (resized).jpg

A few zip ties have been left loose so I can finish up those missing connectors. Once that's all tightened down, I can start on the second wiring harness: switches and solenoids. Given that the switches can't be placed neatly on circuit boards like the lights, I expect this part of the wiring to be a bit less smooth of a process.

Thanks!
Yes, there are a couple of subway branches. In the top right (top left in the wiring photo, of course), there is a skill shot drop hole in the outer orbit. The center ramp has a lifting mech, and shooting the ball under the ramp lands it in the second branch of that same subway, with the ball ending up in the volcano from either shot. From there, it can either be fired upwards to "erupt" via the VUK, or that repurposed kickback mech pushes it out of the VUK and into the volcano tunnel scoop (also hittable from the playfield, to the right of the ramp), where it makes its way down the longer bent subway and eventually emerges from the left scoop. This sets you up with an upper left flipper shot either into the hallway (eventually reaching the VUK at the far back of the playfield) or into the hot spring, which is the long-ish scoop you can see at the left side of the photo, that sort of subways into a VUK that feeds a ramp back down to the right flipper.

....man, there's a lot going on here.

BACKGLASS.
IMG_20220212_022329348 (resized).jpg

Not so quick process getting this one. I finished up the art for this months ago - last modified date on the final art file is September 25. Sent it to the local sign shop with instructions on how to do the somewhat unusual layering (includes a black masking layer to improve contrast). I said it wasn't a huge priority to finish up right now, because I was busy getting ready for MGC anyway. So it got a little delayed.

Okay, maybe a lot delayed. December 9th I finally gave them a call and was told it would be another couple weeks. January 5th....almost done. Two days later, their printing guy called me up. The vinyl backing sticker was done. Just one problem. The sheet of glass I'd brought in from the local glass shop was the wrong size!

Fortunately, it was in its original packaging, with the original receipt (giving the correct dimensions), and the glass shop made it right free of charge. But since it's tempered glass - special order only - it was another few weeks before I finally got it in.

...Fast forward to last Friday, and at long last, Volcano Blast has its backglass. Trim went on with some very careful rubber mallet taps, aside from the top piece which decided to crack into pieces. At least I had enough left over, from trimming the sides to fit, to cover the corners and center, which is really all I needed to protect the top edge effectively anyway.

Thus completes this game's backbox, more or less 100%. I might swap out the speaker panel screws with black ones to camouflage them.

Nearly two months now with NO updates! What could possibly be my excuse this time?
Well, I mean... I bought a house...
IMG_20220407_133642942 (resized).jpg

Some of you may recognize this from the game room thread, but quick summary of the story behind it. My grandfather moved into a home a few months ago and had to sell his house to pay for it, which sucks horribly. What sucks worse is it's the house he built from the ground up in the 50s for both his family and his photography business. On top of that, my mother was tasked with putting her childhood home on the market and was entirely heartbroken about it. After stewing over the situation for only a few days, I derailed the entire plan by selling my own (old and honestly quite poorly laid out) house, and buying Grandpa's instead.

The room pictured was the main photoshoot area at one point, then converted to a computer room at some point during the System 11 era. Half of it will remain a computer room - although with a fair amount of counter space kept free for board work - and the other half will be my private arcade. Yes, the windows will be blocked eventually. Still quite a big mess of unsorted personal belongings in the way, but once that's all said and done, I'll be able to get working on VB once again. The house also came with a workshop area and a very nice selection of shop tools (drill press, table saw, vices, bandsaw...) so I think that will be a big improvement on both homebrew and restoration work in the future.

Settled in decently well for right now, enough that I can at least prioritize doing VB work when I have free time. I love the new game room, there's enough floor space to set up sawhorses as a poor-man's rotisserie, allowing space to walk around it while also blocking access to two perfectly good games!
IMG_20220530_161949504 (resized).jpg

So, I've been working on the wiring, which has been waiting since before I moved out of my old house, and definitely not because I've been procrastinating at all, no sir. Fortunately, with everything planned out in advance, coils were super easy to do wire routing, and switches are going to be up next. Not all coils are done yet - still need to finish up power/ground and build a couple of assemblies to drive the diverters - but that stage is nearly done.

As far as the coils I'm using, I had a ton of them sitting around - but, being sourced mainly from System 11 games, most of them were 23-800. Which is fine for a few applications, such as the autofire and the volcano VUK (easily the deepest VUK on the game), but why put extra processing into doing high voltage PWM when you could just install the correct coils in the first place? Anyway, I wanted to build some mechs with fresh new/restored parts, so I used that as an excuse.

Ew:
IMG_20220627_205949620 (resized).jpg

Since the new arcade/workshop has that countertop, I cleared off a bit of space and set to work, with new coils from PBL and some nicely tumble-polished brackets and hardware.
IMG_20220627_210254497 (resized).jpg

And voila, two fresh new slingshot coil mechs, built to Williams WPC specification.
IMG_20220627_211728742 (resized).jpg

As long as I was at it, I rummaged through the organizer and found that I had enough restored parts to build a new outhole kicker. Since the one currently sitting on the whitewood is pulled directly from a parts playfield, tarnished and full of dirt, this will be a nice and quick replacement once I turn the playfield back over to add Molex plugs to the few topside coils.
IMG_20220627_221908584 (resized).jpg

Solenoids are wired up!
IMG_20220703_224806249 (resized).jpg

Well, aside from the ramp diverters and outhole kicker, which are mounted topside, but at least they only need plugs. I definitely didn't waste any zip ties by tightening them too early or bypassing them with wires, no siree! Yes, flipper mechs are a little ugly and EOS switches not quite mounted right... oh yeah and drywall screws holding the return springs... yes, this is all temporary.

The only playfield wiring remaining is switches, then I expect to be able to properly play this whitewood. There aren't any holes cut for rollovers, so I'll just mount those switches near-ish and call it good until it's time to swap over to the real playfield. Also waiting on two magnetic reed switches from @Sonic, which will cover the two lock positions in the volcano subway.

The standard lower third is now fully wired up!
IMG_20220710_204006698 (resized).jpg

I had a bunch of colored zip ties I got for cheap, so I mixed it up a little. Wire colors are kept pretty consistent throughout:
Red- 5V supply
Orn- GPIO to switches
Yel- 12V supply
Grn- Switch ground
Blu- 48V supply
Vio- Board communications
Brn- Speakers
Blk- Drivers to solenoids (and common grounds because I couldn't find thick enough green wire for those)
Wht- Drivers to lamps/flashers
IMG_20220710_204024196 (resized).jpg

Future wiring will probably come from Wire-Bot, but I stocked up on all this before rockwell opened up shop, I think!

Plenty more switches to go, but this covers the most important ones of all. I'll need to do some cabinet wiring before I can properly flip this, but so far everything is going together pretty solidly.
IMG_20220710_203934901 (resized).jpg

Second draft of the light novel is coming along as well!

Milestone reached: Under-playfield wiring is completed at last!

Well, in theory, anyway. Knowing my history, it's entirely possible I will have messed something up or otherwise need to make changes to this. But everything is solid, wired up according to multiple guides I made up, and it matches all the electrical elements that I've tested on breadboards.

Up next is some cabinet wiring to finish up, then I'm hoping I will be able to drop the playfield in and play this thing at long last.
IMG_20220717_232440212 (resized).jpg

That's exactly what I dealt with on my first build. Had a massive rat's nest of wiring, full of EMI issues, and ended up chopping everything out and redoing it system11-style in time for its first show. I'm hoping my planning has been good enough that I don't have to deal with that situation again!

GUYS. I DID IT.

Getting this to work was... interesting. How, you may ask?

So, the wiring went pretty decently. It's not totally done (still need proper line input rather than an extension cord running through the hole that the power button will go through, etc), but I have the voltages I need. 5v for logic, 12v for opto switches, and 48v for flipper coils. I finished(-ish) that job Sunday night, and immediately wrote up some basic firmware to at least let me test out the new flippers. Tried to burn it to an ATMega328P... communication failure.

Why on earth would I get this problem when I had one working great last year? What changed between then and now? I spent a huge chunk of time last night messing with it, trying to figure out what I was missing or what was not connected or configured properly. It was about 10:40 PM when I finally realized the problem. The clock speed fuses weren't set, so the unconfigured chips were running too slow to keep up with the USBTinyISP programmer.

I'll have to dig through the notes and remember how to set the fuses, but I did find the one chip I'd previously messed with that already had the fuses set, so I programmed it with my quick Fliptronics firmware and dropped it into one of my driver boards. Took a deep breath, fired up the power supplies, and pressed the flipper button.

Nothing.

Also tried triggering the slingshot, also covered in firmware. Also nothing.

I didn't know what it would take to fix. I figured tonight I'd have to open it up again and diagnose my wiring. Check the low hanging fruit first, like make sure it's getting the 5V and ground.

And then I got out of the shower, thinking about it, and I realized I had missed the lowest-hanging fruit of all. I was trying to test it with the coin door open. You know, the door with the high voltage safety interlock.

So anyway, that video is the first proper test of a Fliptronic-style controller in custom firmware, on a custom flipper driver board that happens to be the first complete KiCad design I ever made. I'm happy now. You can also see some lit GI (driver controlled), as well as the reflection of the fully lit backglass logo. Backbox is considered done at this point.

Okay, WHERE does the time go, like WOW.

Guess I was busy doing other things around the house, although I'll admit procrastination was also involved. And scouring the workshop for one single component with which to test a theory, and only today remembering where it was.

Perhaps an explanation is warranted. That video was very satisfying to record and post and show off. And then I tried it again when I arrived home from work, and promptly had the board lock up after a couple of solenoid pulses. Same problem on the left side, random resets and lockups. I wondered if memory corruption was happening, but was directed back to plain old lockups when I set an LED to blink continuously for diagnostic purposes.

It must be so painful to watch someone teach themselves electronic engineering...

Anyway, this disgusted me quite a bit, after all the time I spent planning out wiring, grounding, and routing, and I did a number of online searches trying to dig up the reason why a microcontroller with a wide voltage tolerance would be so unreliable in the game. As it turns out, that wide voltage tolerance only matters when choosing what to steadily run it at. Tiny dips or spikes, like you get when you suddenly burden your power supply with a coil firing, will still cause it to reset or lock up.

So, just in case anybody else is designing a board with a microcontroller, these are the steps leading up to the final fix:
- Set all unused GPIO pins to output, but keep them turned off.
- Add a 10K resistor between 5V and the reset pin.
- Add a 0.1uF ceramic cap between VCC and GND (and I should do the same with AVCC but haven't yet).

The extra components I threw onto the back of the board since it was the only way to make them fit.
IMG_20220826_211842873 (resized).jpg

But it WORKED. Honestly, I was amazed, but I should have known because all the guides tell you to do it. I guess I was just used to testing on the bench without it and having it work fine. Of these three things, the ceramic cap was absolutely the main thing that fixed it, though the other two probably helped. DO NOT DESIGN A BOARD WITHOUT THESE COMPONENTS. I will remember this in the future.

Anyway, time for another hiatus. The component I had to dig for was the ceramic cap, as I apparently have only one of them in the house. Mouser order incoming for those and the resistors, a very commonly needed rating, of which I also have only a tiny supply.

Zener diode trick is interesting, I may need to incorporate that into my next design. (Since I'm really that unhinged.)

The resistor pulls the reset line up to 5V. Although the chip generally works as intended when sitting on the bench, having the reset pin floating means it could end up getting tripped unintentionally. I don't know if this specifically helped my issue or not, but all the AVR help sites recommended doing it.

Thank you! It's been a long journey, with a lot I haven't done before (despite this being my second build), but I think we're finally starting to reach the tipping point where things are going to come together pretty quickly.

Couldn't do board work last night as I'm not expecting the new caps until Wednesday, but I spent the evening on firmware, both for the individual controller boards (I'm calling them satellites) and for the main hub controller powered by the Teensy LC. Downstream communication is perfected, now I just have to solve an issue with switches and I can write up the final firmware routines.

From there, all control goes to the central computer!

For today's update, be prepared for a bombardment of tech details!

Some bits of the code from Undertale can be reused, especially the light sequencing routines, but much of the playfield interface code has to be rewritten for the new system. I can't really complain about this, because the new way is actually quite a bit simpler, at least in the PC framework.

In the old control system, a varying number of bytes would be sent down to the two Teensy controllers. The controllers received bytes in pairs, with the first pair being the number of the device to control, and the second toggling it on or off. To retrieve switch states, the computer sent down a blank data pair, which the downstream controller took as a request for switches and sent that data back up.

The new control system simplifies that a whole bunch. The USB interface code is designed to send 64 bytes of data at a time, so the computer just does that on every 60fps frame. Addressing? Nah. The byte order expected for all the peripheral devices is already known from the start. I just send down four bytes of data for each satellite board, thus 32 on/off bits for each playfield module. The Teensy hub controller sends data out to each module all in parallel at once, one bit at a time, then receives data back from them in the same way.

The boards use this data to turn devices on or off, with some of them programmed to look for other later bits and use them for dimming/PWM. I haven't specifically mapped out the bit arrangements for each satellite yet, but the functionality is there. As a board receives lamp/solenoid output data, it sends back switch states. The hub controller collects all reported switch states, but the computer doesn't get told about them quite in real time. The Teensy incorporates a one-frame data buffer (thus, computer gets switch data closer to 30 refreshes per second than 60), and uses this for error checking. A switch only gets reported as closed if the playfield module reports it that way twice in a row. This way, if EMI or connection issues cause a single clock cycle to mess up, it won't result in a phantom switch closure. This isn't so much of an issue for output devices, as accidentally turning on a bulb or coil for only 1/60 of a second would barely be noticed.

Yes, most of this is already implemented and being tested, so far with very promising results.

Perhaps you may wonder about safety features. What if the computer, or the Teensy hub, locks up while a coil is active? Well, there are two layers of safeguards going on in the communication system. The Teensy checks constantly for data from the computer, and assumes it will get a consistent stream of bytes even if nothing is changing on the playfield. If it doesn't see anything for 50+ms, it zeroes out the entire output array, which kills all lights and coils as soon as downstream communication finishes. A similar version of this same design is in force on the ATMega firmware on the playfield modules themselves, performing the same task on a lower level if the Teensy locks up and stops communicating downstream to them.

The ATMega chips locking up would pose a problem, but adding a few caps to the module boards seems to be solving all the reliability issues I had with them. Another order is on the way, as I've decided to add larger electrolytic caps on their main power inputs. When I hooked up the two left flippers, I had new lockup issues caused specifically by the power dip from operating two flipper coils at once, and adding a new cap to the driver board and a larger new cap to the main supply fixed those problems. If chip lockups do prove to be a problem in actual testing, I have a solution ready - there is one more free pin on the Teensy LC, and I could use that pin to drive a relay to cut 5V logic power entirely, which would force all the module boards to reboot. If this happened during a game, the lights would flicker off for a second, then everything would go back to normal. Because playfield state is constantly refreshed, players would barely notice.

Guys... we're finally making some progress. VIDEO BELOW.
First, a picture: The game with the lights on, except these lights are directly controlled by the computer.
IMG_20220924_222925960 (resized).jpg

Now for this video... after numerous issues with configuration and communication, I was finally able to bring everything together enough to get some very basic gameplay. Flippers, slingshots, and bumpers work (might need to double check an issue with the top two jets though), but most importantly of all... the volcano can erupt!
(Gonna have to turn down the power on that left eject solenoid, though!)
Super quick video: