MEEP: A Minimal Mini-VPin Project [Completed!]

[jeremymtc]

Current status: Completed! Finished photos in last comment.






Hot on the heels of my previous bartop project comes another one, this time a mini virtual pinball cabinet built in minimalist fashion from scrap wood and mostly recycled materials.

Friends and guests who've played the bartop have frequently commented that a pinball game would be a fun addition, and I agree with them. In the long ago past I had played around with Visual Pinball and VPinMAME, and I'd seen many of the wonderful machines that people here on the forums have built over the years, so I knew that it was an attainable goal and was eager to join the club.

This machine will be running a 21.5" 1080p LCD for the playfield, an 11.6" 720p LCD for the DMD/Scoreboard display, and powered by a Firebat T8 Plus/Pro Mini-PC with Intel N100, 16GB RAM, and 512GB SSD. Input is handled by an SJ@JX USB encoder, and audio via the guts of an old bluetooth soundbar.

The software being used for this project is Visual PinballX Standalone (the SDL port of VP) via Batocera Linux, with all of the console junk stripped out of it.

Once committed, the first order of business was to come up with a design for the machine. I wanted to use the same design language and finishing materials as on my bartop build both for aesthetics and economy (mostly economy!) I was able to reuse the template I had made for my bartop cabinet to define some of the shapes and angles, although the proportions are way different.

This cabinet will not have a backglass display, only a DMD hemmed in by speakers in the rise at the rear. The speaker/DMD panel will have an acrylic overlay which should meet the playfield overlay fairly seamlessly.

This was the non-scale design sketch I ended up with (please disregard the weird colors):



With a set design to build toward, I focused on the computer hardware and software aspect of the build for a couple of weeks. Once I had more or less settled the question of hardware and had a running prototype of the system in pieces on my workbench, I sketched out an internal layout image just to help keep my thoughts organized as I transitioned into building the actual cabinet:



Once I had convinced myself that I had enough to move forward with building the cabinet, I started planning out the dimensions using the various bits of hardware that I needed to accommodate. There were still a few unknowns, but only of the sort that become easier to figure out as you go, once you have the first basic cuts established. So with that, I picked out some wood from the scrap pile and got to cutting.

First mockup using the initial cuts below. The wood used for the sides, facia, and header is 1/2" plywood salvaged from a staircase template used for a previous home improvement job.



The cabinet base and rear panel were made from an exceptionally grody and waterlogged piece of 1/2" T-111 siding. I had to cut away a lot of material from the edges to get to the "heartwood", but not a big deal since I only needed a slice 11.75" wide for both panels. A ton of sanding took care of the rest.



I realized too late that I had screwed up a little when cutting the interface between the back panel and header panel (the piece that goes over where the DMD will live). I should have extended the header panel to cover the join, instead of having the rear panel butted against the header. It's an easy fix though, just need to cut down the rear panel a little and cut a wider header to fix it:



I had to trim the monitor bezel slightly for it to fit within the cabinet symmetrically, and it's visible in its approximate future location in the cab. The monitor uses the bezel to retain the LCD panel, so I don't really have any choice but to use it. The rest of the monitor assembly is decased, and will hang from the bezel which will be situated on ledges for easy removal.

That's as far as I got for today. I still need to do some wood filling, radius all the corners, do a seal coat with urethane, more sanding, and fix the header panel before I can move on to the next steps: cutting battens, ledges for the playfield monitor and DMD display, speaker and amplifier mounts, and all the rest of that. Will update as I go!

[markc74]

Very cool idea. I had sketched a design out remarkably similar a few months ago but stopped myself from going further with it until I finished the last pinball cabinet (6 years and counting!)

I was even going to use the same type of mini pc (n100) but found performance in vpx pretty poor so I'll be following to see how well yours runs.

Otherwise love the design so I'll be following 

[jeremymtc]

Thanks, Mark! I actually owe you a debt of gratitude - your Polybius Mini has probably been the single greatest design influence for these projects 

Yes, the N100 system is just about minimally adequate for VPX. Still, with v-sync and frames locked to 60fps at 1080p it works well enough for what I'm doing. I'll probably only ever have a curated collection of maybe 15 classic tables on this machine, and the only one I've tested so far that doesn't stay locked at 60fps has been Twilight Zone. In general though, this machine is intended more for use as a fun novelty at parties than as any kind of serious pinball simulator, so if there are tables that it just won't run that's unfortunate, but ok with me.

I do have a question for you, and/or any of the other experienced vpin builders: For monitor positioning, do you think it's better to have it mounted near-flush to the top rails of the cab, or recessed at a slightly greater depth? I had planned to have the playfield overlay mounted near-flush to the rails but with maybe about 1" separation between it and the playfield monitor.

[Alejo I]

This looks so cute.

[jeremymtc]

Thank you, Alejo!

Day 2.5 of the cabinet build:  I've made some progress.

The main box has been glued up, and the header panel/rear panel joint has been fixed. Now it's properly mitred and fitted:



The great Nigel Tufnel once boasted that his Marshall 'goes to eleven'. I may have him beat with this amplifier robbed from a cheap soundbar - this one goes to 'Chinese woman yelling at you'. When the volume knob is turned to whatever the arbitrary maximum value is, the user is rewarded by what I can only guess is a severe scolding in a language foreign to me

I made and fitted a little acrylic mounting plate for the tiny amplifier board:



And bored a hole in the rear panel for access to the control knob:



This is the aluminum lockplate/palmrest, which was salvaged from the broken lid of an old Dell laptop and cut to size. The radius at the front was fortuitously part of the original laptop design, and I matched it for the radius applied to the front corners of the cabinet. Also visible in this photo are the monitor mounting ledges, which I had initially glued in the wrong place:



...And this is me gluing the ledges a second time, in the correct position. I had stupidly neglected to account for the 1/2" thickness of the bezel the first time around.



I roughed out the front control panel button layout. This panel is not permanently installed as of yet since I might want to make some layout changes, but being able to install some buttons there will allow me to confirm whether the plotted positions for the flipper buttons are workable with regard to inside clearance. The current layout is a stacked coin and start button on the left, exit in the center, and plunger on the right.



Radius' (radii?) were applied to all of the corners, and some amount of wood filling has been done, but there's still a lot more to do. It's starting to look a little bit like the drawing!

[markc74]

Thanks, Mark! I actually owe you a debt of gratitude - your Polybius Mini has probably been the single greatest design influence for these projects 

Thanks! Bizarrely I'm actually working on that today to upgrade it to a Pi4 and a 10" screen - it was supposed to be a Pi5 but that doesn't work well with the controller hat I'm using (annoying!) but it's a great little toy and I don't really play high end games on it anyway

I do have a question for you, and/or any of the other experienced vpin builders: For monitor positioning, do you think it's better to have it mounted near-flush to the top rails of the cab, or recessed at a slightly greater depth? I had planned to have the playfield overlay mounted near-flush to the rails but with maybe about 1" separation between it and the playfield monitor.

On my Flip cabinet I've mounted the monitor flush but I'm not sure I like it like that. It definitely makes the tables feel flatter and lack depth. Could be that I need to tweak the angles in the table but I'm not comfortable enough with VPX yet not to screw everything up!

Anyway - progress is looking really good!

[jeremymtc]

Thanks for the input on the monitor! That was exactly my thinking on having it recessed to give it a little more apparent depth and appearance of parallax, especially since this one's 1080p and the PC driving it is no graphical powerhouse. I've played with the Z-depth and view angle/view distance settings a little bit in VPX, but find that the results are just a little weird-looking on my setup. I think that maybe they're tailored more for use in VR or for headtracking than for a physical cabinet, where the cabinet itself gives some perspective effect at least as far as how the shape of the playfield appears.

If you do want to mess around with the perspective settings, you can do it without fear of ruining anything. Just delete the table's .cfg file (just like in MAME) and it'll regenerate a default.

Glad to hear that you're still tinkering and having fun with your mini cab. It really is a neat little toy!

[jeremymtc]

I haven't made a ton of progress on this cabinet over the past week, mostly because I'm waiting for delivery of a few parts that will necessitate woodwork before the cabinet can be cleaned up and finish work can start. Nonetheless I've chipped away on some things.

I've decided to go with a different LCD display for the DMD panel than originally planned. The original one was chosen simply because I already had it on hand and would have been workable, but it was a compromise - it would have needed to be mounted vertically, and there were some scaling issues with it within the PinMAME/ES/Batocera software when oriented in portrait. All of that was rendered moot when I shorted something on the LCD controller while relocating the input panel during a test session and the panel went into an irretreivable burn-in mode. A happy accident, as far as I'm concerned as it forced me to buy a more suitable display. The new one is a 7" 1024x600 HDMI panel which can be installed in its normal orientation and shouldn't run afoul of the scaling problems that the old one did.

I had been thinking of how I wanted the lockbar to operate for a while. The easiest solution would have been to simply screw it down into an internal crossbrace, but I really didn't want to have any visible fasteners anywhere. I came up with a very simple solution which looks like it's going to work out fine - the concern here is not for security as in a real commercial cab, but just for fast access and a way to keep the playfield display and glass locked into place securely:



The lockplate is bonded to an interior insert (really just a crossbrace) which carries a stud/carriage bolt that can be engaged by a camlock mounted through the bottom base of the cabinet. The insert prevents the assembly from being able to walk or twist relative to the cabinet, and the cam lock prevents it from being lifted upward.



The playfield-facing edge of the insert does two things: it acts as a longitudinal stop to locate the playfield monitor, and it has a slot cut in it which accepts the front edge of the playfield glass:

 

The carriage bolt was countersunk through the insert piece prior to gluing, and has a nut and washer on the underside to keep it retained tightly and help prevent it from torquing or walking if someone tries wrestling with the lockbar. The two nuts locked together at the bottom of the stud are what the pawl of the camlock will engage with to prevent it from being pulled up vertically:



In the cabinet's final form there will be edge molding running in a continuous strip around the cabinet profile and rails, and the lockbar will actually sit on top of the molding, adding 2mm to its height. The material I'm using for the molding is fairly soft and has some squish to it, so when the lock bar is engaged it should have a pretty positive clamping effect but with some 'give' to it.



I've installed some placeholder buttons to test, check and verify that everything in this crowded area of the cabinet can coexist peacefully. Switches and wiring loops are present so that there are no (or more likely, "fewer") surprises later.



I've also added a couple of small, pivoting ledges into the rear of the cabinet to retain the playfield monitor. These double as ledges for the playfield glass to perch on to give it some separation from the top of the monitor. The molding trim which will mount to the rails has a 2mm lip which overlaps the edge of the rail and will prevent the playfield glass from coming free of the cabinet. For removal, the monitor slides forward with the lockbar removed, and once the monitor bezel passes the center pivot of the moveable ledges can be tipped up a few degrees and pulled away. The playfield glass will be masked and painted to "hide" the bezel and ledges, and all of the theoretically visible area of the insides of the cab will likewise be painted black.



The playfield glass (acrylic in this case) will need some additional support down the length of the playfield, and I'm still deciding how I want to do that in a way that doesn't interfere with removal but stays invisible. One option is to glue stringers down the edges of the underside after masking and painting, but I'm not super enthusiatic about gluing to a painted acrylic surface. Another option would be to mount stringers to the top of the monitor bezel either by glue or self-adhesive velcro. I'm leaning toward the velcro option, as it would make them easily removable if there are any potential interference issues with the edge trim during playfield/monitor removal. Maybe some of those little plastic 'peg and hole' shelving supports might even do the trick..?

[MiteWiseacre]

I too like to use up material around the shop 
Looking good, how do you plan to finish it?

[jeremymtc]

I too like to use up material around the shop 
Looking good, how do you plan to finish it?

Thanks! It's a fun challenge to try to construct something cool from literal junk   

Regarding finish, that's a very timely question as I just came in from the shop after spending the afternoon on this blasted thing. Exterior finish is woodgrain vinyl, with acrylic overlays and black PVC edge trim - in finished state, it will appear similar to my "Project Jetsam" bartop.

Progress has been made!

During a lengthy delay while waiting for parts there wasn't a whole lot that could get done on the cab, but there were a few little things that I took care of. I filled and redrilled the holes for the flipper buttons, mainly for aesthetic reasons, but also because I noticed that my various press-ganged test subjects always placed their hands on the upper set of flipper buttons rather than the primary lower ones, and I figured that some additional depth/distance was called for. I also threw caution to the wind, and drilled the cooling fan, power button, and AC cord passages despite not having all of the hardware necessary to determine their fit in place. It's worked out well in the end though.

I also finished filling voids in the plywood and sanding, applied a coat of urethane varnish, and sanded it back level to prepare it for the surface vinyl application:



I ended up going with shelf supports to support the playfield overlay midway down its length, to prevent it from sagging while allowing for easy removal of it and the playfield monitor. This worked out great:



I finally received the 7" LCD panel to be used for the DMD, which had been the main blocker to progress on the cab. I couldn't finalize the playfield monitor position, playfield overlay dimensions, or lockbar setup until that piece was in place, as the design of the cab calls for each of those parts to "lock" the others into place for assembly and disassembly. To mount the speakers and DMD in the minimum of space allowed I concepted this setup to be built from acrylic sheet remnants and wood:



This 1/4" carriage bolt (countersunk through the top of the header panel), wingnut, and washer, hangs the assembly in place. A small bamboo dowel on the right indexes the speaker/DMD panel level relative to the cab. The mounting block for the module is slotted where it accepts the carriage bolt so that the assembly can be pulled free just by slightly loosening the wingnut, once the playfield overlay is removed. A lot of the cab's electrical and power junctions reside behind this panel, so quick and easy access for maintenance was important to me:



And this is what the lower half of the module looks like when fitted to the cab. Some furring strips were hot glued to the interior of the module to provide support for the outer overlay/reveal panel, and that panel is held in place by adhesive u-shaped edge trim which adheres to the furring strips. The HDMI and power cables for the display are also routed through the interior of this panel.



...And this is what it looks like, as an assembly. There will be steel mesh speaker grilles fitted into the reveal panel from the inside, but I don't have the grilles in hand yet. The playfield overlay snugs up to the underside of the DMD panel assembly, and in the final installation the edge trim that meets it and retains the reveal panel will fit a bit straighter and more uniformly:



Vinyl fitted, fan grille attached, playfield now properly located, and all the various motherboard connection breakouts/relay/buck converter/led connection nonsense is done:



The buttons in place are mostly placeholders for the ones I intend to use, however those will also be LED lit. The cab is now in a playable state though:



A bit of a rant here: Please allow me to say on record that mapping controls in Batocera/Emulation Station and VPX Standalone sucks sweaty hairy donkey balls. I level the blame mostly at Batocera/ES - they try so hard to simplify the process for the typical 'console gamer' installation, and make so many assumptions about the hardware users will be hooking up, that they end up making it umpteen times more difficult than it should be to do something so simple as mapping an encoder board. Rather than use the tried and true convention of labeling inputs as "button 1", "button 2", "axis 1", "axis 2", etc., everything is this horrible consolified "North", "South", "East", "West" garbage to describe button inputs. Maybe I'm just old, or maybe I'm just spoiled by how it's done in MAME or any other PC software I've ever used, but Batocera's system is just terrible to use for anything that's not a standard gamepad layout 

Whew.

Paint was drying on the lockbar and playfield overlay when I took the photos above, but they are now done too. The last things to do in construction terms involve parts that I'm still waiting to receive: edge trim, speaker grilles, desk grommets to finish the exterior amplifier knob and AC cord passages, and the power, flipper and front panel buttons.

[PL1]

The buttons in place are mostly placeholders for the ones I intend to use, however those will also be LED lit.

Are you planning on using microswitch buttons for the flippers?

The downside to that approach is that the hysteresis of microswitches (top) is more likely to affect pinball gameplay than leaf switches. (bottom)



You might want to consider leaf switch buttons like GGG's LED friendly Electric Ice 2 buttons with True-Leaf Pro switches, X-Leaf switch assemblies that replace regular microswitches, or non-LED Class-X buttons with True-Leaf Pro switches.


Scott

[jeremymtc]

Hey Scott, thanks for the input on the "input" 

The flipper buttons won't be illuminated, but for now they are microswitch buttons. Leafs may be a future addition though, and Randy's stuff is usually the first I look at for quality parts. As it is, this machine is very much set up with hardware of the lowest common denominator kind in all aspects, so I'm not sure if infinitesimal gains in terms of precision and latency will be very noticeable. 

[jeremymtc]

Bit of an update here - the cabinet is now about 98% complete, needing only the addition of dust filters and exterior trim bezels for the intake vents in the base. I will try to get some good photos of it uploaded in the next couple of days. I will say that it has cleaned up remarkably well, and its function/performance has exceeded my expectations. Chuffed.

I haven't shown much of the internals of the cab yet, and now is as good a time as any for a rundown. Below are some work-in progress photos from an earlier date - please excuse the poor lighting and sawdust/debris, and keep in mind that wire management hadn't really been attempted at this point.



The PC powering this build is a Firebat T8 (Intel N100 Alder Lake) mini running what is basically a stripped down Linux installation. Although it's a very low-power PC with integrated graphics, after installing a lot of tables and doing a lot of playtesting I've found that it really works quite well at 1080p/60Hz resolution. If you're contemplating a build using one of these, I'd say that it works great so long as you're not trying to push 4K.

Some of the early testing I had done was with more recent tables (such as those from VPW) set up for 4K high-refresh monitors, and some of those ran pretty poorly - typically 20-30fps or so. Using some of the less hyper-detailed VPX tables as alternates, performance has been fine and totally playable with solid framerates and more than acceptable display quality.

This is a screenshot of btop (a system monitor application) running via ssh from my desktop computer, with the pinball cab running Monster Bash in a game session with multiball active. The little N100 4-core system is chugging right along at 16% utilization. The only time I see CPU utilization approach even 50% is during game selection and loading from the frontend; at the left of the CPU histogram the burst of activity shown is the game being selected and loaded:



The mini PC's cooling solution is pretty tightly integrated in its very small case, with an intake fan that draws air in from the bottom and exhausts through one side. It didn't make much sense to decase the PC for this build as it would lose the benefit of what's basically a fitted duct channeling air directly through the CPU heatsink. The PC has been oriented so that its exhaust is mostly inline with the cabinet's exhaust fan, but it's rotated slightly out of plane in order to keep all of the i/o ports more easily accessible.

I needed to break out a couple of functions from the motherboard in order for it to function seamlessly in the cabinet, but this was definitely made trickier due to the very tight packaging of components inside the mini PC case. I went with a 3-pin audio cable with the JST connectors snipped off the ends to solder to the motherboard V+ and ground contacts for the power switch, and to the +5V lead for the cooling fan header. A small hole was drilled in the case to allow passage for the cable:



The PC is powered by a laptop-style switching power brick, but this one is 12V rather than the more common 19.5V laptop power supplies. Since the PC only needs 12V, I was able to power the relay module used to run the cabinet's exhaust fan without stepping the voltage down before it reaches the relay. Instead, a buck converter was installed on the output side of the relay to act as a fan speed controller of sorts. The relay is triggered by the +5V signal coming from the mini PC's fan header, and output voltage to the cabinet's exhaust fan was adjusted to ~7.8V for best acoustics. System temperature hasn't exceeded ~65C in testing so far:



There was a need for multiple +5V outputs to power the amplifier board, DMD screen, and various button LEDs, and I spent some time thinking about how to manage those. The SJ@JX USB encoder is "capable" of supporting about a dozen LEDs on its own, but the cabinet didn't need quite that many. I considered using some of those encoder outputs to power the amplifier and DMD, but had some concerns about combined amperages and ground loops. Rooting around in a junk drawer netted a handful of small 1 amp USB AC adapters which were pressed into service to power the amplifier and DMD instead, and these are plugged into the power strip mounted into the rear of the cab.

In the photo below the button LEDs are plugged into the +5V encoder outputs, but I've switched those up based on the desired behavior of the system when different power states are taken into account. By default all of the USB ports are powered during suspend, and while there may be a BIOS switch to alter that behavior I've decided that it's actually easier to handle what I want to do in hardware instead:



Typical for this sort of installation, the PC is set to suspend/hibernate with a short press of the power button, and it shuts down with a long press. The power button is illuminated, and its conditionals for lighting pretty much follow what the encoder board outputs during suspend and power-off, so it alone is powered from the encoder. The control panel button LEDs instead receive their power from my fan control circuit so that they only illuminate when the system is active.

Most i/o to the cabinet is handled wirelessly with a bluetooth keyboard, RF wireless mouse, and a PS3 controller connected via BT to remotely navigate Emulation Station menus. I threw a 4-port USB extender into the cab mainly for convenient reach to plug in USB thumbdrives for bulk file transfer.

I hope to post photos of the more or less completed unit tomorrow or the next day 

[minorhero]

Just finished reading this thread, really nicely done!

Since I am currently finishing up my first arcade cabinet build, seeing a vpin build is very timely for me

Any top tips for someone thinking of going down a similar road?

[jeremymtc]

Thanks, minorhero! Very kind.

My number one tip for a vpin would be to use Windows instead of Linux. My #2 tip would be to plan your build around standard PC parts with a discrete GPU instead of a mini-PC. Basically, do everything exactly opposite to what I've done and you should be golden 

I had asked a similar question here in the software forum when I was in the planning stages for this one, but it ended up more a monologue in which I answered my own question and gave some reasoning for it.

Why use Windows? Because 97% of the visual pinball speaking world is pretty sure that you're using Windows, and that's where all of the software, utilities, features and toys are located.

Why use standard PC components? Because you won't be artificially constrained by hardware and have greater upgrade potential.

A third consideration would be to have clear target for your goals before budgeting for a machine. A 4K resolution build will cost (at least) an order of magnitude more than a 1440 or 1080p build due to the PC hardware and display required, and for that reason it doesn't really make a lot of sense to try to build a 4K mini - you're much better off going full size at that point. 

I don't regret the hardware or software choices made for this build one bit, and quite the contrary, I'm really stoked on this machine. But, it was more an exercise in learning and hacking which I find to be enjoyable in a kind of masochistic way, and it enabled me to build the machine for astonishingly little cost (less than $75 USD in direct costs, and well under $200 in total if repurposed existing parts are taken into account).

Hope this helps in some way! 

[jeremymtc]

Except for a couple of minor parts that will never be seen, it's finished! Here's some pics:


Playfield glare is not as bad as it appears in photos, but maybe I'll take some nighttime pics soon.




Functional lockplate with updated control panel






Volume control at top, AC cord exit at bottom, both finished with 35mm desk grommets





I've really enjoyed this build, even though it has been frustrating at times. It's a ton of fun in action.

Thanks for following along!

[javeryh]

This is beautiful. It's a really compact design too - perfect if you don't have room for a full sized machine. How does it feel to play?

You have to post a video of it in action - boot up, game select and some demos! 

[minorhero]

Thanks, minorhero! Very kind.

My number one tip for a vpin would be to use Windows instead of Linux. My #2 tip would be to plan your build around standard PC parts with a discrete GPU instead of a mini-PC. Basically, do everything exactly opposite to what I've done and you should be golden 

I had asked a similar question here in the software forum when I was in the planning stages for this one, but it ended up more a monologue in which I answered my own question and gave some reasoning for it.

Why use Windows? Because 97% of the visual pinball speaking world is pretty sure that you're using Windows, and that's where all of the software, utilities, features and toys are located.

Why use standard PC components? Because you won't be artificially constrained by hardware and have greater upgrade potential.

A third consideration would be to have clear target for your goals before budgeting for a machine. A 4K resolution build will cost (at least) an order of magnitude more than a 1440 or 1080p build due to the PC hardware and display required, and for that reason it doesn't really make a lot of sense to try to build a 4K mini - you're much better off going full size at that point. 

I don't regret the hardware or software choices made for this build one bit, and quite the contrary, I'm really stoked on this machine. But, it was more an exercise in learning and hacking which I find to be enjoyable in a kind of masochistic way, and it enabled me to build the machine for astonishingly little cost (less than $75 USD in direct costs, and well under $200 in total if repurposed existing parts are taken into account).

Hope this helps in some way! 

This is super helpful! Thank you! The final pictures look great, congratulations on an awesome build!

[jeremymtc]

Thanks fellas 

Yes, will try to get some decent video footage up soon.

[jeremymtc]

Heya folks. Here's some video footage of the cab in operation. Sorry for the long delay in getting this up - got called out of town and then broke the screen my phone, and I'm sure that the video quality hasn't been improved by the fact that I couldn't really see what I was recording.

YouTube Video

Cold bootup time is right about 40 seconds. Entering and resuming from suspend only takes a couple of seconds and is the normal mode of operation; it only needs a cold boot after it's been unplugged from the wall. All in all the machine is pretty snappy and runs well.

[javeryh]

Super nice. You should be proud. 

[DaOld Man]

Looks real good.

[jeremymtc]

Thank you, javeryh and DaOld Man! 

The machine has had a pretty positive reception with friends and family, and I'm happy with how it performs and plays. I think that I might go ahead and invest in some proper switchgear for it per PL1's suggestions now that it's proven itself in action. It's immersive enough for me to 'forget' that it's a digital representation of the real thing on occasion, and I often find myself trying to manipulate it with body english.

I've ended up changing the magnasave/nudge buttons to clone the flippers, as they've ended up being more trouble than they're worth in terms of user/guest confusion. Either people would inadvertently put tables into Tilt by using the wrong button when mapped to 'nudge' (I think that VPX expects an analog input rather than digital, so the "strength" of the nudge input is basically 100% when using a button), or accidentally cycling through lighting settings when mapped as magnasaves because many table authors seem fond of using those 'unused' inputs to adjust LUT on non-magnasave tables. I'm sure that in Windows it's easy to remap your way around those issues, but under Batocera it's such a nightmare that I don't want to touch the mapping settings ever again, so long as it just works. If I were to build this table again using Batocera, I'd definitely go with just the two flipper buttons rather than four.

I will definitely look at adding an accelerometer board to the machine in the near future to give it some analog nudge without the hassle of external controls, but I'm not totally sure of what compatibility issues might arise with regard to support in Standalone and Batocera quite yet.