Next step in lighting this thing up is to fit some LEDs in where the old indicator LEDs used to be.  However, I didn’t want them to be just static.  After talking with a buddy about his electronics projects, I decided I would appropriate his project for use in my WSG.  Conveniently, his project was a K.I.T.T. styled LED tracer.  I’m not sure where he found the schematic, or if he designed something on his own, but I found a schematic that worked for me, and began breadboarding it.

The biggest obstacle for me was going to be figuring out how to step the power down from 9V to 6V, as I wanted to run everything (the lights and the sound) off of the same battery.  I dug around on-line, and found a schematic to regulate the voltage down to 6V (I can’t find the original source I was looking for, but it’s similar to the one found on this page), and luckily I had the IC needed to make it happen.

It worked just as I had hoped!

However, there was a dilemma.  The two LED banks on the existing box were a set of 8 LEDs and a set of 7 LEDs.  The K.I.T.T. schematic I found was only 6 LEDs.  I figured I could try and double one or two of the LEDs and see if that would work (and not kill the circuit).  I tried one, and it worked fine.  I would have used a center one, but since there are two center LEDs, I opted to double the first LED and the last LED, to minimize the obviousness of my duplication.  It worked swimingly!

However, there was still the matter of what I was going to do with the second bank of LEDs.  After some hemming and hawing, I decided to wire up seven more LEDs, meaning half of the 6 original LED spots would be driving 3 LEDs and half would be driving 2 LEDs.  I would take the additional 7 LEDs and arrange them in a random order, with two of them firing simultaneously (much like the K.I.T.T. layout).  I figured this would emulate the futuristic computers from 60s and 70s TV & Film, which would have been about when this T1 Test Unit was originally made.

I wired everything up (using Euro-style screw terminals, to allow me to easily randomize the order of the LEDs, without having to plug and unplug them from the breadboard) and it worked just as expected.  For a short while I toyed with the idea of hooking in a second set of controlling components to allow for the futuristic lights to be driven at a different speed (the speed can be adjusted by adding an additional resistor (up to 47k total resistance) on the 22k resistor (that is not hooked to the 1N4148)), but decided against that as it would mean I would probably need a third PCB in the enclosure, and I was already running low on space.  I will say though, it did look cool at a super fast speed!

With these steps done I will be adding 21 LEDs to this box.  My Weird Sound Generator is not only going to sound awesome, it’s going to LOOK awesome.  Now I just need to get it soldered up, and this project will be pretty much done!

I had twelve tasks left on my previous post, but have decided that the beverage holders belong in phase 2, so I’ve nixed that. Also, I’m going to leave the dual RCA jacks I previously hooked up, as is. That left me with ten tasks to complete, and as of today, I have completed 8 and a half of them. Of course, there were a couple tasks I forgot about when making my last task list, as well as a couple I just added while working over the past two weeks, either because they needed to be done, or I just had time to work on them:

* Mount PC power switch to back of machine.
* Setup Frontend software (Hyperspin) as well as emulators (MAME, Nestopia, Kega Fusion & Snes9x)
* Install cable holders to keep floating cables tucked away

While I wrote my last post, I was in the middle of patching the old holes that were left on the front panel (with the coin doors and start buttons). I patched them up, some of which were deeper than others, and let the wood putty dry a couple days before sanding it down.

Previously, the panel was screwed and glued to the front of the machine, but I wanted to allow this panel to be removed fairly easily next time I wanted to make a modification to the machine (plus it will lighten the machine a bit if I can remove it before moving it into my basement). So I needed to come up with a way to fasten the front panel without the use of glue.

Since there was already a small strip of wood on the bottom of the cabinet cavity that acted as a stop for the bottom of the front panel, I would just have to come up with a way to have the bottom of the panel latch on to that strip, and then I could screw in the top of the panel.  I figured this would be better than just using screws, since the way the cabinet was manufactured, there is no easy way to put screws in on the bottom of the cabinet after the fact.

I went to the hardware store to see if I could find any metal pieces that were shaped like the kind of bracket you would use to secure a door with a 2×4 (or for you video gaming geeks, the the S or Z pieces in Tetris), but smaller (approximately 1 inch in depth and no more than 2 inches in length).

Unfortunately, I couldn’t find anything small enough, so I had an idea.  I could use just a few small L-shaped corner braces, and join two of them together with a bolt and nut, making an S piece.  None of the smaller L’s at the hardware store had two screw holes on any flat surface, so I would need to butt two of the things together to ensure that they wouldn’t spin around.

I got home, made my S’s, and fastened them to the front panel and even though they were closer to an inch deep than I wanted (I was shooting for 3/4″), it turns out the full inch was better. Because of the angle of insertion of the panel needed that 1/4″ of wiggle room.  The panel didn’t fall out when I put it into place, but I tossed some (blue painter’s) tape on it in the meantime, to just make sure it stays in place before I permanently screw it down on the top-side.

It was around this time that I received my new FrankenPC case from Newegg.  I put the components into the cvase and realized that the fan I had needed a longer cable in order to be mounted in the spot I needed on the case, so I snipped the fan’s wires and added about a foot of wire in the middle.  I closed the machine up, and like butter, it fit neatly into the cavity on the bottom of the machine, next to the power supply for the monitor and sound system.

The only downer is that the case was made of paper-thin metal, so I stripped out about 3 screw holes while putting it together, but for less than $20, what do you expect?  Luckily, I won’t need to open up this machine very often.

My next big task was going to be tackling the control panel.  I still had a lot to do.  First step was going to be installing the hardware to be used for the joysticks.

My plan was to bottom mount the joysticks using screw in hex insert nuts, the kind you might see on pre-fab furniture from Ikea. I mounted them onto the control panel, and so far, so good.

Next, I needed to route out a channel for the T-molding.  This couldn’t have gone easier, considering I had never used a router that wasn’t used for data transmission.  However, the slot cutting bits that are used for the T-molding channel spit out a TON of sawdust.  Make sure to do this in a well ventilated area!  The T-Molding fit right into the groove (needed to use a rubber mallet to help it in) and to my surprise, it was centered!

Now it was time for the scariest part of the project: adhering the control panel overlay I had previously purchased from Arcade Overlays onto the control panel.  I didn’t want to have to order a new one, so with the help of my wonderful wife, we slowly started at one side, adhering and pulling the backing paper as we went.  as luck would have it, everything matched up, just as intended!

I now had to cut out 22 holes for the buttons and joysticks.  The overlay material is very thick, and rather than use power tools to cut through thick, but ultimately too thin for anything more than a knife, I opted to cut them out using an X-Acto knife.  This was a painfully slow process, and even though things looked good at the end, the blister on my middle finger was not so appreciative of the method I chose. I don’t know what a better alternative would be, but be prepared for pain if you go the X-Acto route.

With the control panel overlay now installed, I needed to begin the tedious task of transferring the controls from the mockup control panel to the real thing.  Much like when I first installed the controls onto the mockup panel, this was a slow process, and was only made more complicated by the fact that the wiring, while labeled and nicely organized, just got in the way.

It took well over an hour to transfer everything, but when it was done, things were starting to look real nice.

Since I had the control panel put together, it was a good time to mount the I-PAC and J-PAC to the underside of the control panel (which will make things easier if I ever need to remove the control panel.  This was a fairly straight forward process, however, one of the holes on the J-PAC was smaller than the mounting screws that came with it, so it was a little unnerving enlarging the hole on the J-Pac with my cordless drill, but I did it, and it looks good.

I then turned my attention back to the cabinet itself, for a couple tasks that I had recently realized I needed to do.  I installed my PC power switch, an inch or two away from the Monitor/Audio power switch, and luckily the switch was just deep enough to be mounted in this fashion, as I neglected to measure anything before drilling the hole.

Next I recognized that there were a few cables for the monitor that would get in my way any time I needed to move the computer case in or out of the machine, so I used a couple small, nail-in clips to fasten the wiring to the side of the arcade cabinet, making things easier for me in the future.

…and that brings you up to date with the progress on the Gauntlet project.  I am in the middle of configuring the front end software, and I recently decided I was going to re-key the arcade locks (on my own), so I’ll let you all know how that went, very shortly!

* Mount PC into cabinet (either in a case (easy) or free (more difficult)).
* Replace old locks with Hex locks.
* Modify control panel layout and create new mockup for brief testing.
* Assuming the new control panel works out, transfer this to permanent control panel on 3/4″ plywood, including cutting panel, drilling holes and spraying with black paint (in case there are spots not covered by the control panel overlay or t-molding).
* Install mounting hardware from old control panel to new control panel.
* Replace wire-nutted splicing of video wiring with insulated quick connects.
* Patch front panel (damage from old screws) and scrape off front panel fragments that are stuck to mounting strips.
* Install front panel to cabinet.
* Replace stereo RCA jacks on audio amp with stereo 3.5mm headphone jack and mount the jack.
* Install new mounting hardware for the joysticks onto the new control panel.
* Install/adhere the new control Panel Overlay to permanent control panel.
* Move controls from mock-up control panel to permanent control panel.
* Create replacement back access panel, including mount for PC power button and port for power cable to route out of cabinet.
* Mount fan on vent grate on back of cabinet to allow sufficient ventilation in cabinet (particularly for the PC).
* Install T-Molding on new control panel.
* Mount J-PAC and I-PAC2 to underside of control panel (if room permits) or next to audio amp (less desirable, as removing the control panel will be more difficult).
* Install beverage holders on sides or front of machine (This is certainly not a requirement, but I’d like it, just for the fact that it would make a spill less likely, and I need to put my beer somewhere… right?!).

So I began working on Friday, and started with something easy that had began bugging me; I swapped out all of the locks for the Hex locks.  I had actually removed the locks a week or two ago, so any time I needed to move the front panel, I’d get coin doors flipping open and whacking me in  the arms, face or shins, depending on where I was when it was being moved. I then moved onto making a new mock-up control panel, using an isosceles triangle button pattern for the red and green stations (rather than the more spread out formation I was previously using).  I cut out the control panel, again using the control panel overlay as a guide.  However, this time I bought a new Milwaukee hole saw to drill my holes (the whole saw kit I bought from Harbor Freight is fine for a hole here or a hole there, but not up to the challenge of 32 holes on the mockup and then another 32 on the real deal.  I had already put 64 holes with the Harbor Freight bit, and it was a pain in the ass).

While the new hole saw is a vast improvement, the hole cutting still went slow, as the bit invariably gets clogged with wood particles. I had the foresight this time to find a wire brush to quickly clean the teeth of the bit, which sped up the drill time a bit (no pun intended).

I test fitted one of the changed stations, and things felt alright, so I moved on to the final control panel.

I traced the outline mockup control panel onto the plywood, as well as all the button holes to make quick work of the transfer.  The ¾” plywood was a little more difficult to drill through, as the bit would get bogged down about halfway through.  What I ended up doing is drilling until at least the guide bit got through, and then flip the board, and drill in the other direction (I actually grouped them together, so I wasn’t flipping the board back and forth.  I did all of the front cuts and then all of the back cuts).  This cut down on splintering on the surface, and the bit seemed to go through the first ¼” a little faster this way.  I did have a little piece of the plywood break off between buttons, but I patched that up with wood putty and let it dry overnight.  I sanded it down the next day, and the panel was ready to paint (after hitting it with a tack cloth).

Painting was nice, as it allowed me to get rid of 4 miscellaneous cans of black spray paint I had laying around that were previously used on various projects, but not empty.  I put on several coats, and by the end had a nice smooth surface that I was confident would allow the control panel overlay to adhere properly.

Next I started working on transferring the control panel mounting hardware from the original control panel to the new one.  The existing hardware bolted through the entire control panel, so a bolt was visible on the top side of the panel.  Also, there was a small wooden spacer used, to lift the hardware off of the underside of the control panel by about an eighth of an inch.

Since I wasn’t going to pull the wood spacer off of the old board, I found a suitable replacement: old paint stir sticks (which were coincidentally from the now defunct Knox Lumber). They were almost the same dimensions, and required only minimal cutting on the top two inserts, cutting that was easily performed with a utility knife. Using a staple gun, I fastened them into place.

I didn’t want the bolts exposed on the top of the control panel, so I opted to use wood screws that were nice and beefy, and that would come within fractions of an inch of the surface of the control panel, ensuring the best hold possible.  I installed the screws, with the hardware, and it fit like a glove!

My last task for the weekend was to replace the wire nuts that I had recently installed on my video wiring splicing, with insulated quick connectors.  I decided to do this because if I left it hardwired, I wouldn’t be able to completely remove the control panel without clipping wires.  While I don’t plan on having to remove the control panel, for well, anything, I am going to want it removed for when I move the beast into my house.  The cabinet will fit through the doors with the control panel installed, but I don’t want to risk damaging the work I’ve made on the control panel, nor my door frames or walls.  Snipping the wires and installing the insulated quick connects was a breeze.  I guess I’ll find out whether my crimping was good whenever I get the machine back up and running.

So that’s five tasks down, twelve to go.  Looks like I’m still going to be plenty busy!

The next day, I figured I’d better get things more permanently affixed within the cabinet.  On my lunch break, I stopped up at Ax-Man, looking for a molex connector with .156 pin spacing, preferably for 2 or 3 pin connectors.  I was prepared to kludge something together (like hack up a 4+ pin connector, to serve as my connector for the composite sync.  Luckily I found a 2 pin connector.

While I was at Ax Man, I decided I would look for a power switch for the frankenputer.  I knew I wasn’t going to be reaching into the cabinet every time I wanted to turn the thing on, and the power switch that was affixed to the case was quite lame, so I found an apt alternative:

Oddly, you can get the SAME exact switch at allelectronics.com and I’ve noticed that they routinely has many of the same components as Ax-Man. Unfortunately, it’s not illuminated, but at least it won’t look out of place on an arcade machine, I just need to find an out of the way spot to mount it, so it isn’t accidentally hit during the heat of battle.

Additionally I found some replacement “locks” for my coin doors (I don’t have any keys for the existing ones).  The locks are opened with a hex wrench rather than a key.  I’ll probably eventually get the locks re-keyed, but at 95 cents a piece, the hex wrench ones will work for now.

Again, you can get almost the exact same locks from allelectronics.com.  Weird.

When I got home, I first wired up the new power button.  I didn’t know how much slack I was going to need, as I didn’t know where I was going to mount it, so I gave myself about 4 feet of wire, twisted it up, to keep things neat, and crimped some connectors on the wire.  I soldered the molex that goes to the PC motherboard and used heat shrink to hide the solder splice.  Plugged it in, and it worked like a charm.

Since I already had the blue and yellow stations wired up to the JAMMA interface, I needed to get the I-PAC2 hooked up.  Since the Red and Green controls were already wired all I had to do was get the front start buttons wired up, and I could begin plugging wires into the I-PAC2.

This was relatively quick, or at least a lot less time than it took me to do the rest of the control panel.

After that, I wanted to get the Comp Sync wire mounted in the Molex connector I picked up.  Using a razor blade, I chopped off the keying that was present on the connector, so it would fit on the monitor side. I soldered the wire into the metal connector and slid the connector into the molex housing.  Luckily, it worked without a hitch.

(the top connector is the connector I made, and the two empty spots on the connector below it are where the old horizontal and vertical wires were going into).

Now I needed to program the I-PAC, since it was by default setup to run as Player 1 and Player 2.  Unfortunately, WinPAC, which comes bundled with the cards, cannot program both cards while they are attached to the system, and since the J-PAC was running my video, I needed to use my laptop to program the I-PAC.  WinPAC was surprisingly easy to use though, and I had things programmed in less than 5 minutes.

I powered the machine on, and like clockwork, everything was operating as planned.

In a further attempt to wrap things up (for phase 1), I tried to fit the case for the Frankenputer into the bottom cavity of the cabinet.  I had measured this out, and it was just going to fit.  Unfortunately, I neglected to account for the small retaining piece on the back that keeps the PC cards in the PCI/ACG slots.  Nor did I account for the cables coming off the back of the motherboard.  I don’t have another case available, so either I’m going to have to find another case on the cheap, or figure out a way to mount all of the components into the machine without it being a complete clusterfuck.

But that’s for next time.

I first hooked up the Arcade VGA to get the drivers installed, and with the DVI to VGA converter, I was able to see the picture on the VGA Monitor while I did this.  After the drivers were installed, I shut down the PC and hooked the Arcade VGA card up to the J-PAC (JAMMA to USB converter).  I turned on the arcade monitor and fired the PC back up, and we got a whole load of back and forth shifting of video.  We played with the vertical sync and horizontal sync a bit, but I was concerned because there was an amber LED lit up on the J-PAC board, and that it possibly meant that there was an issue.  So we opted instead to plug the card into the VGA monitor I had, and go through control panel “testing” (aka. play a bunch of video games).

Unfortunately, something with the Arcade VGA drivers seemed to be messed up, as any game we’d fire up would be choppy and the audio would be unintelligible.  The previous video card I had in it (an ATI AGP card) worked fine, so I rolled back the system with a system restore, pulled the Arcade VGA card, re-installed the crappier AGP card, and fired the machine up.  No problems.  Not caring to deal with the video sync or video card issues, I tabled those sub-projects and we got our game on.

Fast forward to yesterday. After letting this sit around for a week and a half (and working on some other (neglected) projects), I got back to work on the Gauntlet machine.  I had exchanged a few e-mails with Andy @ Ultimarc, to see if he knew why the drivers on the Arcade VGA had fucked up the performance of my machine, and he wasn’t sure.  He did confirm that the amber LED on the J-PAC board was normal though, so I figured I’d give things another shot. But first I was going to see if I could resolve the sync issues.

I powered on the monitor and the PC, and quickly found out that by playing with the Vertical Sync, Horizontal Sync and Horizontal Position, I could get a somewhat acceptable image, however the image wasn’t staying still.  For example, if dialog box popped up on the black desktop, the screen would start flipping left and right and/or up and down.  I thought this might be because I was still using the AGP video card, so I tossed the Arcade VGA card in.

Unfortunately, this did not solve anything.  I was still having sync issues.  I even tried testing the continuity on the composite video sync wire (coming off of the JAMMA), but that was good. I wasn’t getting system slow down this time (from the video card), but since I couldn’t test it in game, I couldn’t be sure.  I decided to sleep on it and tackle it later.

Today, I got home and looked hard and long at the machine.  I started tracing wires.  Off of the JAMMA board I had a red wire for red signal, blue wire for blue signal, green wire for green signal, a black video ground wire and a brown composite sync wire. There were two unused wires from the video harness, a white wire and a purple wire, for vertical sync and horizontal sync.

I followed Red to red.  Blue went to blue. Green went to green. Black went to black. Brown went to another Molex connector that was not connected to the monitor chassis, even indirectly (this connector was daisy chained off of the Molex connector I was using, so it had RGB and ground going to it as well).  The purple and white wires were hooked up to the Molex I had connected.  It was an A-Ha! moment. So I needed to hook up the Brown composite sync wire to the monitor chassis, but where?  There were four pins on the chassis that would make sense: Positive Horizontal Sync, Positive Vertical Sync, Negative Horizontal Sync or Negative Vertical Sync.

The brown wire was wired to the Positive Veritcal Sync slot on the Molex connector, so I figured that might be it.  I fired up the machine and the sync was still bad (was synced vertically, but flipped horizontally depending on the image displayed).  Stumped, I looked up at the monitor control diagram inside of the monitor and noticed the following passage “WHEN USING COMPOSITE SYNC USE HORIZONTAL SYNC INPUTS”.  OK, but which one?

Since the brown was already wired up to the positive vertical sync, I popped the pin out and moved it to the positive horizontal sync, housed in the same connector.  Fired it up, and the image was stable!  …Except not centered correctly.  The image looked like the offset filter in photoshop.  The edges of the image were in the center of the screen and the center of the image was on the outside corners of the screen. I tried playing with the horizontal and vertical position knobs, but no go.  I couldn’t get the picture to move over that much.  So I turned to the internet.

In a stroke of serendipity, Google led me back to Mr. Bob Roberts (from whom I had purchased a good deal of my replacement parts), in a page where he breaks down monitor sync issues. After a quick scan, it seemed obvious that I needed to hook up the composite scan wire on the JAMMA board to the negative horizontal sync pin on the monitor chassis (for those that are wondering and for the aid of Googlers out there like myself, the Gauntlet cabinet has a Wells Gardner 19k4914 monitor in it).

I hooked the wire up to the pin using a hook-up wire, fired up the PC, and it looked VERY promising.  Shit wasn’t moving around like before.  Once XP was fully booted, I started fiddling with knobs and BINGO!  It works!  Of course I had to fire up a game:

What’s next? Well hook up the rest of the controls to the I-PAC2, and start fastening shit down. Then it will be time to work on the permanent control panel!

This has been the most complex circuit I’ve worked with (as far as number of components and number of external controls (Pots & switches), so using Visio, I began to map out how I was going to lay this thing out on a PCB I picked up from Rat Shack (Part No. 276-150, the same board I used for the Triggered Drum Light).

I probably spent two or three hours mapping out how I wanted to lay things out, trying to leave room in case I wanted to piggy back some LEDs onto the board (there are standoffs built into the faceplate of the enclosure I bought for this project, and they are perfectly spaced to fit the pre-drilled holes on the PCB.  The board would then sit right in front of the old 7-segment LED window.  Putting a few LEDs on the board, I could then make a logo up and have a little backlit nameplate, or at least that’s the idea).  Overall I was happy with how things turned out:

(Click Image for Larger View)

As you can see, there is a lot going on here. It wasn’t going to be a quick project to solder up, nevermind the fact that I was manufacturing my own cables to run from the faceplate to the control boxes (that I bought from Ax-Man).

I ended up starting with the cables.  They were going to be about a foot and a half long, with a 1/4″ phono jack on one end (to go into the faceplate), and an RCA jack on the other end (to go into the controller).  I started soldering them up, but quickly realized that the shrink tubing I had bought (on-line) was too narrow to fit the two wires I had soldered up.  Luckily I knew a place locally that sold shrink tubing in longer lengths, and while it probably took me an hour of cutting wire, stripping wire, soldering screwing on connectors and shrinking the tubing, they ended up turning out pretty nice.

I was able to hook them up to my breadboard, and surprisingly they all worked like a charm.

Within the next few days I figured I’d get started in transferring the parts to the PCB.  This was not an easy process.  With over 100 solder points, I was going to be at this for a while.

Over two nights and probably four hours of soldering, I finally finished with the soldering.  I hooked everything up, including the new switches I had bought a day or two prior, and it wasn’t working.  ICs were there, so I hadn’t made that mistake twice.  Everything looked solid, no components were loose.  The only thing left I could think of was the soldering.

I started to check each solder point, making sure nothing was loose, and nothing was shorting.  quickly I found two spots where the leg from a component or a wire was too far over, and probably shorting.  These were kind of long as it is, so I took out the clippers and nipped off the long shorting ends.  Fired it up, and now I was getting sound!

Unfortunately, some of the controls were not working, particularly the two components that were hooked up to the switches, the killswitch and one of the tone knobs.  I had already put in enough time that night, so I called it a day, knowing that was the next thing on my plate.

Today I picked up the PCB, after letting it sit for a day, and started troubleshooting. Using the continuity setting on my multimeter, I went looking for shorts on or near the switches, and couldn’t find any.  Switches seemed to be working fine, so I couldn’t figure out what was the problem.  So I snipped the wires on the switches and using some jumpers, hooked up a switch I had been working with on this project prior to finding these new switches.  And….

“Great”, now the switch that was working, no longer was working. It didn’t take long to figure out why.  One of the wires had wiggled loose from it’s solder point.  Soldered it back up, and the switch worked perfectly.  Resoldered the snipped wires back to the existing switch, and went to work on the other switch (to activate the killswitch).

This one was acting as if it was always on, regardless of what position I had it in.  In the off position, the signal should be routed to the Out Jack, and in the on position, the signal should be routed to the killswitch and then to the Out Jack.  A simple SPDT switch setup.  Hooking up my multimeter, the continuity was acting just like I thought it should, but it wasn’t working right when I had everything running.  I snipped the wires on this one, and used three jumpers, and started playing around how they were hooked up to the switch.

It turns out that this switch is just funky (or at least I’m not familiar with).  Basically the switch will send signal to one out pin, or both out pins, depending on the position.  So I would need the “always on” pin to be to the killswitch, and the “part on” pin to be wired directly to the out.  This was as easy as flipping the outputs around.

I soldered everything back up.  Powered it on, and bingo, everything is working!  Short of the case and running power, this project is getting really close to done.  I enlarged a few of the existing holes on the faceplate and started mounting up the jacks, switches and pots.

It’s looking really sharp.  Next step is working on the lighting that I’m going to install in this box!

P.S. – I’ve also decided on a name for this project. From here on out, it will be the T1 Corruption, in homage to it’s humble beginnings as a T1 Test Unit.

I had a Athlon 64 XP CPU and motherboard that I salvaged from a PC repair I made a few years back, but I was pretty sure that either the motherboard or the CPU was bad. I cobbled a machine together and got to installing Windows XP.  Everything installed fine, but I was running into some really bizarre issues.  Task manager was showing CPU Usage of 100%, but under Processes, System Idle Process was sitting anywhere between 95 and 100 usage, which would suggest that something just wasn’t right (for the record, these were the same symptoms I was seeing a few years back).

I searched the net for a solution, and the best I could come up with was a guy who suggested it might be a bad motherboard, but I just couldn’t confirm it, since I didn’t have a spare socket 939 motherboard.  Luckily, my buddy Andrew had one laying around that he was willing to sell me for a little more than I wanted to pay (but ultimately cheaper than I was going to find locally, and short of going the Ebay route).  I swapped out the motherboard, and bingo, the Task manager was now showing an accurate CPU usage (and the other idiosyncrasies I was seeing went away).

Now that I had Windows up an running, I tossed the Arcade VGA card into the machine, but nothing was coming out on the small CRT monitor I had.  I tried the projector I had.  No go.  I tried the S-Video out to a TV.  No go.  I pulled the card out, and realized that there was a PAL sticker on the bottom of it.  I then realized I was an idiot, and was trying to send PAL signal to an NTSC TV (which of course makes sense, since the company is based out of the UK).  Still didn’t explain why the VGA port wasn’t working (I didn’t have a DVI device to test the DVI port on).

I then checked the Ultimarc website, and little did I realize, but I wouldn’t be able to use the VGA port with any device other than the arcade monitor.  Crap.  Luckily they sold DVI to VGA adapters and I still needed to order an I-PAC2 from them anyhow, so I ordered it up.

While I was waiting for the rest of my arcade parts to arrive, I decided it was time to get to work cutting wire, stripping wire and crimping on connectors for my temporary control panel.  I had no idea it was going to be such a time intensive task.  After probably 3 hours of cutting, stripping, crimping and labeling (to make it easier to track wires), I had the temp control panel wired up.

Unfortunately, I still had to solder a couple dozen of these wires up to the JAMMA harness I had picked up from Bob Roberts. This took me at least another hour or so.  Basically, over a four day period I spent over 4 hours working on the control panel wiring, but the time I spent now was going to be time saved when I transfer this to it’s permanent control panel.

Now that I had things partially wired up, I couldn’t help testing it out, even if it was just running to my laptop.  Boy was it sweet.

I then went to town and started drilling.  I quickly found out why waferboard was a poor choice.  First off, even with a good bit and a steady hand, the hole saw wandered quite a bit while drilling through the irregular waferboard. After the first hole, I used a center punch to cut down on drifting, but even that was only partially effective.  Secondly, drilling waferboard is messier than normal particle board, as the backside tends to “explode” as the hole saw pops through the backside of the board.

I went through it though, and tossed buttons onto one station and realized I had a few issues.  Because of the wandering bit, some of the buttons were way too close to each other.  This was going to be basically unusable as a template.  Additionally, the seventh button (where the pinky would go) was way too high, leaving the feel of the layout uncomfortable.  So I decided I was going to have to redo it, but this time I’d do it right.

I was sick over the weekend, but by Monday had felt good enough to get back to work. I went and picked up a 4′x4′ sheet of 1/4″ hardboard.  This was still going to be messy, but the consistency would be much more desirable.  Plus, with a 4×4 sheet, I had enough board to make 4 templates, if the next one didn’t satisfy me.

I laid out the updated control panel configuration, and started drilling. And drilling.  And drilling.  I seriously was drilling 1 1/8″ holes for what seemed like 2 hours.  I think I need a better hole saw bit.

I also fastened a board near the bottom edge of the mock up, to add support, as well as to serve as an indicator on how close to the front of the cabinet the controls were coming (in case I felt like moving things around a bit).  I then began to attach the controls I had recently received from Bob Roberts to the mock panel.  This took much longer than I thought, partially because I put half of the buttons on backwards (so the switch connectors would face each other, making hook up a difficult proposition).

All in all, I’m happy with the layout, however after a bit of research, I might adjust the button layout on the red and green stations from an arced layout to an isosceles triangle formation, and move the blue and yellow stations down just a hair (to give a little more elbow space).  I’m supposed to receive the rest of my parts (wiring and I-PAC) any day now, so I’ll hook stuff up and invite some folks over to test this layout out.  In the meantime, here’s a picture of the CPO I received from Arcade Overlays:

Looks fantastic!

Luckily, there is a site called The Gauntlet Hackers, which has a bunch of information on the arcade machines, including the pinouts for all of the boards in the machine (mainboard, power supply and audio amp).  I printed them out and started snipping.  Using some leftover parts and a kludge of audio adapters, I was able to get a small MP3 player running through the speakers.

Not too bad eh?  Both speakers work, so I’ll have full stereo sound (I made sure to label which is left and which is right, to save me hassle down the road).  However I still have to figure out how I want to permanently wire the audio up on this, but my first impulse is just to use RCA jacks and run a standard 3.5mm to dual RCA cable from the PC to the cabinet.  I guess we’ll see what I decide on.

Since I had the overlay and was waiting on controls, I figured I’d start working on a mockup of the control panel.  My plan is to install the controls into a mockup, making sure everything fits right and is ergonomically pleasing.  Assuming that works, I could then use the mockup as a template for the real thing.

I went to the local hardware store and bought a 2×4 sheet of 3/4″ plywood (sanded, the nice stuff) and a 2×4 sheet of 1/4″ particle board.  Together with a set of new jigsaw blades, I spent less than $15.

I took the overlay, laid it flat on the particle board, and traced the outline.  I then brought it outside and used the jigsaw to cut the outline.  It turned out pretty well aside from the one part I had to freehand (since I forgot to lay a straight edge down for it, but I figure it’s a template and not a big deal.

Now I’m just waiting on my order from Bob Roberts, and once they arrive my next task will be drilling out holes for the controls.