;

Welcome to my Arcade Cabinet Design website.

The arcade cabinet that you see on the left is a drawing that I made ,during my student period in Delft. I created this cabinet with all the rubish that I could find without spending much money. It was using a Pentium 233 running DOS-MAME. The joysticks came from an old pac-man machine and buttons from a local guy dealing in pinball machines. All buttons and the 2 joysticks are soldered to a ps2 keyboard with copper wires for each key. I bought a Sony Trinitron TV with video-s input which was connected to an very uncommon VGA to TV device that was in some cases able to bring a descent color image. As a coin door I use a coin proof device from a shower and the coins that I inserted would fall on the ground next to me for infinite credits. 2 small speakers were glued to the cabinet and 3 hot bulbs were illuminating the perspex logo that I printed on my Epson Stylus color printer. Oh boy was I happy with the results. We have played hours and hours in a row and I though that I had build the perfect machine.  However, I was wrong...

That was 22 years ago.


The pictures above are from taken in Delft in 1997. The wood that was used was lying around in the house. On the right the connections to an old keyboard are visible. Finding out the right keys to allow multiple keys presses at the same time, was quite tricky. In the middle, the Arcade was moved outside during the D.S.C. intro weekend.


Recently I read an article about new methods that have been developed over the years to fully enjoy 240p RBG experience of original arcade games. After reading many articles and watching many videos, I realized that I had to modify several parts of my current arcade cabinet. The most important part was the emulation software that is nowadays capable of converting a 31kHz VGA signal into a 15kHz RGB signal that exactly match the original game resolutions and framerates. Getting rid of the interlaced 480i videos modes and seeing the true 240p video modes with perfect scanlines has changed the way I look at Arcade Cabinets these days.

The following sections are created to help anyone who is trying to create the perfect Arcade cabinet.

Designing your Cabinet dimensions

Probably the most complicated task is to start the actual building of your cabinet. Especially because in a later section you will read that a big CRT monitor needs to be installed that should be able to fit within the available space. Since I am currently designing a new Cabinet that should hold a large Sony 27'' PVM monitor, I really need to use 3D modelling software to make sure that my monitor is actually not sticking out at the back and also is tilted enough to ensure the perfect viewing angle. Luckily I have been able to experiment with several angle using my old cabinet, to find out that it is very important to put your monitor under a very steep angle. In fact the preferred angle may be so steep that your monitor will tumble over, if it would not be fixed at the top. Also very important is the fact that the cabinet needs to have the side panels to extend way beyond the monitors front bevel to prevent incoming light and light reflections. (Another reason why you want you monitor tilted a lot). I will soon visit (or may already have visited) the National Game Museum in Zoetermeer, to find out if other cabinet sizes and shapes would be preferred. Also the height of the joysticks and the location of the buttons should be carefully positions, especially because I want to have convenient space for two people playing at the same time, standing next to each other.


At this moment, I have created a simple 3D sketch of my monitors dimensions, such that I can see how deep my cabinet should be. In my final design, I will include a 6mm glass plate as well to protect the Monitor screen and to greatly improve overall looks. When using the appropriate 3D modelling software, you can determine and fix a rotational axis, such that you can easily see that effect on different tilt settings.

On the right, I have quickly rendered an intermediate result of the design phase in progress. This is my first attempt to get a feeling of how I want the design to be.
  • It should be possible to open and close the front panel for easy access.
  • I want a heavy glassplate in front of the screen that should be firmly fixed. But it should also be possible to take it out front the front of the cabinet.
  • The panel width should be sufficient for two people playing next to each other.
  • The amount of buttons should be at least 4 per person, including a player 1 and player 2 button. The credits need to be enabled via the coin door, but a coin button must be present for setting up the machine software.
  • With my current system I only have digital inputs and no analog inputs, meaning that I cannot play games like: Outrun, Space Harrier, Arkanoid, Marble Madness and many others. For this I will soon test a track ball and a spinner. How well I can play a race-game without a steering wheel is something I have to find out.
  • The screen should be lowered with respect to the edges. In the picture on the right, still is still not the case for the bottom part of the monitor. This also holds for the glassplate.
  • The top of the cabinet should be higher than me (1.83m), so here the highest part is 1.95m. This way the illuminated logo will be at the perfect position, slightly tilted.
  • During the creation of my first cabinet, I realized that adding wheels is a useful addition to the comfort. To prevent side movement. to front 2 wheels need to be fixed, the other 2 should be able to rotate. However I am still thinking about a simple mechanism to be able to lift the wheel such that the cabinet is standing directly on the ground.
  • It must be possible to quickly swap the PC in case of issues, like a failing HDD.
  • It should have a main switch, an removable extension power cord and some spare power sockets at the back.
  • The front of the cabinet needs to be inclined such that you can freely put your feet on the ground while playing.
sk

The simplest question would of course be, what kind of cabinet you like to design. Although the question sound easy, it took me a great deal of time to find out what shape and layout I actually preferred. Below I have a picture of some of the most common cabinet and their 3D models, that I found on the internet. Using this picture, it is easy to decide what you do NOT want. If you read on, you will see that I prefer the monitor layout as used in Galaga and Frogger, where it is mounted under a large angle. When you compare Frogger and Pac-Man, you see that the ceiling construction is also different. I prefer the extra gap that is created in the Frogger version.
SK

After creating several attempts to create a nice model in modern 3D modelling software that would fit my enormously large PVM 2730QM, I came to the results as shown in the image below.  The top left 3 images are from SolidWorks 3D modelling software. I created 3D models of the buttons, joysticks, trackball and most important, the Sony PVM 2730QM. The monitor can be rotated around a fixation point this makes it very easy to see how a certain monitor angle looks like and whether the side panels are large enough. The example shown has a rim of 1cm.
After the main shape was roughly completed I used photoshop to draw some example arrows and striped onto the side panel as shown in the 3 bottom left pictures. The rendered black 3D model was used as a basis.  The three colors patterns are just example of a nice retro look. The 6 top centre images are modification of the front panel patterns,also drawn in photoshop. From the 6  bottom centre images the left one has become my favorite pattern. The 3 on the right have the same front panel pattern, but here I changed the arrows on the side panel, to see which i would prefer. The bottom right 3 images have the final selected pattern with 3 different colors sets.
Currently I still have to decide between the green and the purple version.
The large image on the right is a final render where the final pattern has been drawn accurately on the front and side panels. Those have been applied to the 3D model in order to get real results with correct shadings, colors and reflections. Note that the text is invalidly mirrored and also the joystick panel has no color yet.
As a marquee, I used a neon template in photoshop to create 'ARCADE'. Whether I will create REAL neon letter of print them with backlight is still to be decided. Neon may be very cool but creating a customer neon design with high quality will cost a lot of money.
SK

SK

In search of the perfect scanlines and true 240P images.

After some reading on the internet and testing several different methods, I realized that it is very simple and also very cheap to create the perfect Arcade Simulator. Forget about expensive up-scaler , down-scalers and dedicated arcade video cards. What you need is the following:

  1. An old Pentium 4 PC with AGP slot or better a Duo Core PC with PCI-E slots
  2. A old CRT RGB SCART TV or better an CRT RGB Monitor (BNC or SCART)
  3. An old compatible AGP ATI Radeon card or better an PCI-E ATI Radeon card.
  4. A simple VGAtoRGB cable or better a VGAtoRGB converter.
  5. An old PS2 keyboard or better an USB keyboard convertor
  6. An old set of PC speakers or better an old amplifier with car speakers.
  7. A CRT protection circuit.
  8. GroovyMame Emulator for Windows XP
  9. CRT EMU driver for Windows XP
1.
The power of the PC is depending on the games you are going to play. The following games from section 1 will be running at 100% on a pentium 4 PC with an ATI Radeon 7200 and ATI Radeon 9200. The games from section 2 show consistent slowdown on any pentium 4, even though the games are not in 3D. Also with other pentium 4 motherboards I have seen slowdowns even for many games in Section 1. So either get yourself a fast motherboard with Pentium 4, or get a duo core instead. Also not that some games CANNOT run smoothly at all. For example, MSLUG 2, does not run smoothly when many sprites are on the screen. This issue seems not directly CPU related but is an artefact of the emulator and specific game itself. Game will run smoothly most of the time, so most people will be extremely happy with the performance. Memory is not so important as 512MB or 4GB will give you the same results in most (all?) cases.

2.
Forget any flatpanel display you might own, since the games will look bad on them. Whatever tricks you apply, a CRT will always give you much better video quality for games with original resolutions lower then 640x480. Since all games in game section 1 have resolutions below 640x480, you will need you video card to be able to provide you with these low resolutions. Some examples are:
Metal Slug: 320x224p 59.18Hz
Green Beret: 240x224p 60.60Hz
Commando: 256x224p 60Hz
The ONLY way to get your CRT display these modes correctly, is by making sure that your video card can natively support them. This is possible thanks to the CRT EMU driver.
The best monitor you can get would probably be a professional video monitor or called a PVM by Sony. Sony PVM 's and BVM's are absolutely fantastic, but may be hard to find, especially if you want a large size. Arcade monitors will do too of-course but require extra HW to be connected to your 15kHz vga card. Finally, an old Sony Trinitron TV with scart input will do great too, since the RGB and Sync signals are directly accepted as well. The S-logo on the left is taken from Metal Slug 2, on a Sony PVM 2054QM taken with an iphone 6 camera.
ms2
3.
Most ATI cards are compatible with the CRT EMU DRIVER, that is needed to set the VGA vertical output frequency to 15kHz for the CRT instead of the default 31kHz (or higher). An ATI Radeon 9200 for AGP motherboard or an ATI Radeon HD4670 for an PCI-E motherboard are excellent choices and these cards are almost freely available on the internet.
Note that the ARCADE VGA card is also an ATI card where this software is already available on the card itself. We can do it better and much cheaper.


4.
VGAtoRGB sound easy, but may be the most complicated topic for many. If you own a RBG monitor you have either SCART input or BNC input.
For VGAtoSCART you can buy the UMSA. This will convert 15pin VGA to 21 pin SCART and adds 2 audio inputs aswell. Besides the high price, it also does NOT protect your monitor if 31kHz is being transmitted.
For VGAtoBNC you can buy cables with 4 and 5 BCN outputs. The cable with 4 BNC outputs will most like combine the vertical and the horizontal sync together, which be not be a good thing for your VGA output. I would opt for an VGA to 5BNC cable, because then you know exactly what you get: R, G, B, H & V. R=Red, G=Green, B=Blue, V=Vertical Sync 60Hz, H=Horizontal Sync (15kHz/31kHz etc).
For aSony PVM you have only 1 BNC for your input Sync. Also for the SCART connector, just 1 pin is used for the SYNC. This SYNC input is called Composite Sync and expectds the H and V signals mixed together in a special way. This must be done by LOGIC ports and NOT by connecting the together as you sometimes see.

sk

How to prevent Monitor damage due to 31kHz high frequent VGA signal

I have spent a very long time on this matter but I have found the PERFECT solution.
First of all the question if it really can destroy your monitor when you apply a vertical frequency above 15kHz. I am no expert on this matter but a friend of mine explained me why monitors DO get damaged and why this signal should NEVER be applied. Without going into too much details, a 31kHz VGA 640x480(p) signal has a pixel clock that puts a higher frequency dark/light variation on the R,G,B pins. Although the monitor tries to follow this higher frequency with the RGB amplifiers, it does not hurt since the applied voltage is relatively low. The problem lies in the HIGH-voltage part. The monitor receives the sync signals and tries to lock on this frequencies to create a horizontal saw shaped pattern that is creates the scanlines and a vertical saw shaped pattern that create the frames (60Hz). Increasing the horizontal frequency will force the saw pattern to become more dense and the power electronics will follow this saw pattern. Higher frequencies mean shorter rise times and the electronics are not designed for this.
Some monitors will create a high pitches noice when a non 15kHz signal is applied and the screen is jumping around trying to get a lock on the applied frequency. Some monitors will show a blank screen after few seconds of trying. Both monitors will have a shorter life expectation when exposed to these high frequency signals.
I figured that when I leave out the horizontal sync signal towards the monitor when the frequency is above 15kHz, the monitor does NOT suffer. Since I already created my own CSYNC PCB, a small modification would be very easy. On the internet I read an article about using a PIC for counting the frequency, but since I have no experience with PIC programming, I decided to use a small ARDUINO for this. The simplest arduino I could find on Alibaba costs around 2 euro and it suits the job perfectly. When the H-sync is applied to the arduino interrupt pin, the interrupt service routine is execute every sync pulse, so 31000 times a second. A created a small program that counts the incoming horizontal pulse count and only feeds the H-sync to the monitor if the frequency is below 15.5kHz and above 14.5kHz. A created a PCM design using Fritzing using 2 sided prints. It has 3 leds, a red led, a orange led and a green led. When I turn on the PC, the signal is undefined, so the led turns orange and the hsync is disabled. Then when windows starts to boot, the signal becomes 31kHz and the red led turns on. After 20 seconds when my windows XP is booted, the 15kHz signal becomes active and the green led turns on. The h-sync signal is passed to the monitor and the images appears for the first time. The great thing about this is that also any annoying windows popups or startup screens that may we present are not shown, so it looks like we have a real arcade boot sequence, instead of an ugly windows upstart screen.

Great Horizontal Arcade Games


Section 2: (100% Duo Core)

dmnfrnt - Demon Front;
elvactr - Elevator Action Returns

Section 3: (<30% Duo Core)

Zero Gunner 2

SK
Section 1: (100% Pentium 4)

720, aburner2, aerofgt, airduel, alpham2, aof1-3, arkanoid, armwar, asterix, avsp, bagman, bionicc, bjtwin, blazstar, blktiger, block, bloodbro, bmaster, bouldash, breakers, bubblem, bublbob2, bublbobl, buggychl, cabal, cbdash, ccastles, cninja, cobracom, congo, contra, cyberlip, ddragon1-3, digdug, dino, elevator, elvactr, fatfury1-3, ffight, gaiden, gaunt2, gauntlet, gberet, ghouls, gng, goldnaxe, gradius1-3, gtmr, hharry, ikari, indytemp, joust, joust2, jrpacman, kof94-98, kungfum, lastblad, lastbld2, ldrun, maglord, marble, mercs, mk3, mpatrol, mslug1-5, mslugx, mspacman, msword, mtwins, nam1975, ncommand, neodrift, ninjak, nspirit, opwolf, outrun, pacland, pacman, pang3, paperboy, pengo, pitfall2, popeye, rambo3, rampage, rastan, rastsaga, roadrunn, rthun2, rthunder, rtype, rtype2, rygar, salamand, samsho1-4, sbrkout, sengoku1-2, sf2, sfa1-3, sharrier, shinobi, shocktr2, shocktro, silkworm, solomon, sonicwi1-3, spacedx, spinmast, ssf2, ssi, ssriders, timeplt, tmnt, tophuntr, tp84, trojan, twincobr, wb3, wbml, wboy, willow, xsleena, zaxxon


Great Vertical Arcade Games

Section1: (100% Pentium 4)

1941, 1942, 1943, bombjack, commando, ddonpach, dkong, dkong3,
dkongjr, donpachi, frogger, galaga, galaga88, gaplus, gunsmoke, guwange & gyruss.

Section2: (<30% Duo Core)

Ikaruga
SK

HOW TO REPAIR YOUR SONY PVM 2730QM


How to damage your Sony PVM:

I bought this huge SONY PVM 2730QM monitor from internet some time ago. It was working fine using a UMSA scart converter and a PC vga output.
Since i did not own a UMSA yet, I connected a cable I found from 15 pin vga to 9pin DIN. In the PVM manual, this 9pin DIN fits in the 'DIGITAL RGB' input and I thought that
the signals from SCART RGB and DIGITAL RGB are the same. Currently i think they are not. Maybe also the H and V sync signals are different, because when i connected the cable,
i saw a bright white line at the top of the screen and i heard a ticking sound that seems to be the sync process. Anyway, I spent few minutes changing and checking signals when
I heard a big loud 'BANG' coming from the inside of the PVM Monitor.
I though i would just take the PVM apart and look for the blown component in order to replace it, but none of the components seemed damaged.
This is the point where I decided to repair the unit.

SK

How to discharge you Sony PVM:

After removing the casing, the first obstacle was obviously the large amount of electrical charge being present everywhere. After 3 days of hesitation and watching discharge video's on youtube,
I decided to connect a big wire to the chasis and bolt it firmly to a metal BBQ fork I find in the house. The good think about the BBQ fork is that it has to metal pins that I can use to evenly
lift the rubber cap on the Cathode without touching anything. At this moment I have removed the cap 10 times, but I have never heard a discharge sound. Apparently this monitor has a method to
slowly remove charge over time.
Before i discharged the first time, i used a voltage meter to verify if the tip of the BBQ fork is indeed conducting easily to the chasis everywhere (BEEP in OHM's mode).
Also for extra protection i put a transparent plastic HOSE around the BBQ fork to improve isolation and to prevent me touching (or getting close) to the bolt where I connected the wire to the chasis.
When the rubber CAP is removed, I always touch the small spring that is located at the cathode to force discharge.
Note that the inside the rubber cap, we can see that the conduction wires are located on to sides of the centre pin that falls on top of the cathode pin. So to discharge correctly, you really have to
rotate the BBQ fork 360 degrees (or at least 180 degrees) to touch 1 of the wires.
I was lucky that while doing that the rubber cap flipped off so that i could see that it is not all conducting metal inside the rubber cap.

SK


Disassembling the Sony PVM:

I started with removing the backpanel (with the SCART interface) and then pulling all the connectors from the power supply board. I needed some force to pull the power cable out of this socket
and I had to desolder and remove 1 black wire from the back panel to the power supply board in order to remove it and put it aside. All connectors have this 'click' system that requires a small
screw driver to lift the plastic click plate upwards. After some connectors, I found out that just pulling it quickly also does the trick quite well without damaging the connectors.
I removed the following PCB boards: BackPanel (Qb), PowerSupply Board (Fa), Board (D), Board (B)
Then i removed the neck board as i realized that I may damage the tube when i accidentily hit it. The I removed board D2, D3.
Finally I removed Board C itself, but that requires to also remove Board Fb (with the flyback) and the large RED unit including rubber CAP. (They are all connected and cannot be (easily)
disconnected I think.
The Audio board (K) I left.
I also wanted to have the power switch (N) and 1 front panel to verify the led indicators (Board Fb) since I have been unable to get the leds to turn on at this moment.


SK

Measuring the Power Supply:

Since my Monitor was totally dead, the power must be broken. I removed the metal casing to see inside the power supply. As soon as I applied 220V to the power supply, smoke is shown.
After some trouble shooting the smoke is coming from a large grey resistor R671, next to the  THP601. This resistor is here to test the board and it getting very hot for a very short time. If you
leave the power connected for 2 seconds, you will notice that the smoke goes away. this is due to the THP that reduces the current when the component gets hot. This resistor is here, because normally
the degauss coil (DGC) is present. Here the resistor R671 takes over the degauss coil. The coil normally also get power for a very short time thanks to the THP.
The power supply is supposed to deliver +-18volt +-8volt and 140volt. In my case, the 140Volt was gone. When followeing the diagrams in the manual apparently there
was a small fuse blown. This fuse is an IC-fuse. I replaced it with 3 0.5Amp fuses soldered together to get 1.5 amps. Voltage returned nicely at 140V.
So it connected it to the D- board, that requires 140V using 1 cable that I had to entangle from the wire bundles, still left in the monitor chasis.
When turning power on, it immediately would my my 1.5A fuses again. Damn! Note that when measuring the 140V input pints, no shortcut is present. The shortcut only appears
after you turn the board on. Following the schematics did not bring me much further into finding out where the shortcut could be,so ...


Fixing the shortcut in D-board:

Without much knowledge of fixing electronics, I started measuring ALL diodes (0,6V / 0V) & ALL transistors (2x 0.6V / 1.2V / 0V). I used the list from the manual
and tested 1 by 1. Sometimes values are different for other types, but in general values are between 0.4 and 0.7 volt for the junction voltage.
I realized that some diodes are in parallel with very low resistor, from which I concluded that diode is bad, but after removing them they seemed OK again.
Also transistors in pairs sometimes show 1.2V in both direction, but that is OK in those cases.
Also I checked ALL capacitors, by selecting OHMs and verifying if the value increases and decreases. For some capacitors, I check with 100kohms other use 1kohms.
After doing this checking, I found 3 broken parts:
D1516 diode V30N
D1513 diode V11N
Q1507 Thyristor (transistor?) SG-264A
After ordering these expensive (as they come with shipping costs) parts, I noticed that the thyristor was not broken at all. So before you order parts, first remove them from the DCB board and recheck them separately.
The V30N diode is most likely causing the shortcut.
After replacing these parts, and connecting ONLY this D-board to the power supply (using only 1 cable), the power was OK and the shortcut was removed. YES!!
If I remember well could hear the high voltage pitching sound again. Maybe it was after i connected everything together. Anyway, high power was back
and when looking at the screen, all i could see was blackness. When turning the knobs in the back i could scroll the screen down to see a green line with some part combines.
Anyway no nice image, no vertical lines, not good. Also i still did not have any leds being illuminated at the front panel, so i could not even select the proper AV input.
I decided to look at the B-board that is supposed to deliver the power for these leds, and then i saw that the schematics show some test pins T1, T2 .. T19 etc.
When I was walking up and down from the house to the garage, with test voltage levels in my head, i saw suddenly SMOKE coming out of the Monitor, that was on for 30 minutes.
Turned power off quickly, but I could just detect that the smoke was coming from a large capacitor. Maybe it was not smoke but water vapour? It came out of a very tiny crack inside
the top of on of the ELNA capacitors.
Before continuing, I decided to replace ALL electrolytic capacitors on the following boards: Fa, Fc, D, D2, D3 and C. These boards are dealing with power. The audio board i skipped
as well as the input board (with the SCART) and the input boards, led boards etc. The B-board contains also a lot of electrolitic capacitors, but max voltage is 12 volts I think.
Since i did not have any sync signal, this B-board may be the main suspect for now. I figured, how hard can it be to go through the schematics to see see what is doing what, how and why.
Of course I have nice schematics for the layouts, voltage meter for diagnostics, Google for help and manual with Test Points and voltage levels as well. HOW DIFFICULT can it be?
Let's go !


Fixing voltage on B-board for the led illumination:


To be continued in March 2018 ...

Click to watch Sony PVM 2054QM vertically mounted. Setup is still in old Arcade cabinet.