AtariFreak's Arcade Adventures
Personal History of Arcade Restorations and Repair – NW Oregon


Happened upon a bunch of non-working Commodore 64’s and had to buy them.   So, this is the tribute to how I fixed the first….

It had the ‘Blank screen of death’.  This can be caused by several things, U4 (Kernal ROM), U17 (PLA), U7 (6510 MPU), U3 (ROM), U8 (7406 IC), U19 (VIC II), U9-U12 (4164 RAM), U21-U24 (4164 RAM), VR2, CR4, VR1.

Turned on the computer and let it warm up for 5 minutes and felt the top of all the chips -> if your lucky one or more of the 4164 RAM chips are super hot indicating that its bad.  In this case all the chips felt fine and so it made sense to test the next highest probability being the PLA chip.  These are known to cause the Blank screen and run hot so they fail often.  Unfortunately it was soldered to the board on this one, but on one of the other C64’s I bought the chips were socketed.  So I decided to unsolder the suspected PLA and solder on a socket, make life easier.

When I put the socket in and tried the other PLA chip the system worked!   So I knew what the problem is…now to figure out the cheapest/best solution.

The PLA was originally a MOS chip based on the 82s100 which is impossible to find.  So whilest surfing the net I discovered Ray Carlsen’s post regarding using a more common and re-writable 27C512.  I happen to have a bunch laying around from pulls from bad arcade boards…so here was the attempt.

First erase the 27C512.

Then program the chip with the modified PLA data (this chips pins are in a slightly different configuration so the software must match the hardware…)

Next up – wire the chip to a socket modifying a few of the pins around to more match the original PLA.  And plug it into the socket.  Details are here if you can read German:

Not the prettiest wire job, but hey I didn’t even know if this would work and its hidden under the keyboard anyhow…

Flipped the power and got this screen:

So, after a bit more research it turns out that the chip that I used was rated at 120ns, seems fast enough but it has been suggested that the replacement needs to be quite fast under 70ns to pull of the trick.  I dug around a bit more and found some 45ns’s lying around and decided to give it another go.

Thanks Ray!  Sweet, the new chip runs cooler than the original and I can replace them indefinately for free!!


The joystick grommets have arrived!

Pretty easy fix, just flip the control panel down, and unscrew the base of the assy

Throw away the cracked grommet

Insedentally, the left grommet was in backwards causing a binding issue and probably reduced its life.

Now if I could just locate a better speaker, need to do some goodwilling!   Starting to run out of things to do!!  Oh No!!

This game should come with a warning label ‘Warning, Extremely Addictive’.  I never played it very much growing up cause the games were fast and the quarters went quick.  Heck, I just played 10 games in under a half hour.  Got to be in my top 5 arcade games of all time.


The schematics show the output on the Soundboard to be matched for 8 ohms.  Its funny that Williams would put a 6″ 4 Ohm 3 watt speaker in the cabinet.  Since the ohm value of the speaker is less, the amp needs to run twice as hot – I wonder if this is the reason that the AMP chip cooked…hmm. This wouldn’t be a huge deal, but there is very little volume control its either loud or louder. 

 Think I will start the hunt for a good 8ohm slightly higher say 10 watt speaker, I really don’t want to wire up a big resistor in line or cut the cab for an additional speaker (just kidding of course haha). 

Placed an order for 2 Joystick Grommets, the ones on this machine are wore out.  That will be a nice improvement.  Also need to do some final clean up of the wiring and connectors, etc. 

More later!


I mentioned in my last post that there is a constant background noise, a hissing or crackling sound.  It’s substantially quieter now that the soundboard has been rebuilt, however, it is still there and quite annoying.  Here’s how that one got solved:

First start with the easiest to check, the speaker….

Here is a pic of the original 4 ohm 3 watt power house:

Unsoldered that one and wired up a test but known good speaker in its place:

Same results…crackling, hissing…

So this time I decided to check the power source, make sure that there isn’t any AC ripple or interference causing the issue from the linear power supply (this part would be easier if I had a O-Scope).  So I disconnected the power lines and wired up my test power supply directly to the board:

Exactly the same….hissing, crackling…

Next up, test the audio amplifier on the soundboard.  For this one, I had to get out my old-school Panasonic Boombox (what can I say….) and use its amplifier or line-input to check the audio before it gets amplified on the soundboard. 

Next I plugged in one end of the RCA cable to the line-in on the boombox.  Then, stripped the other end of the cable (since the soundboard doesn’t have a line out I had to get creative).

Used some aligator clips, one connected to ground.  The pic above shows the aligators on the wires which I nested in the wire loom just to keep from touching each other.   The positive connection I used a nail clipped into the aligator clip so I could touch pin 1 of the TDA 2002 audio amplifier chip on the soundboard (pin 1 is input for the amp chip)

Its a little hard to see what I’m doing here, but when I touched the nail to PIN 1 the sound came through my boombox PERFECTLY CLEAR.  No static.  That narrowed it down to basically the audio amp chip, or 3 caps or a bad ground…I had already replaced 2 of the caps and the 3rd is a ceramic and what are the odds there. 

Digging around the bone-yard I found a dead Williams Arch Rivals sound board that had 2 of the TDA 2002 chips, I was able to salvage 1 of them and replaced it on the Robotron and now the sound is PERFECT.  Back to gaming!!!


Got a few more hours to work on the game on Monday night.  Finished capping the CPU Board and installing the Lithium Battery modification

Also got a chance to Cap kit the G07 monitor.   The tube needed some scrubbing from the 20 years of crud that built up on the face – made the picture look out of focus.  Looks great!  

There seems to be a problem in the audio….a constant back ground fuzz or crackling that occurs.  Not sure what the problem is yet, hopefully just a speaker gone bad…something to trouble shoot for the next update….till then.


Yes, the parts have arrived.  Luckily they did as my blood-sugar was getting low.  Another couple of moon pies later, thanks BOB!!  Which I happened to have eaten almost instantaneously after the box was opened, long before I grabbed the camera…

Parts installed on the Power Supply circled in yellow, also a power transistor that isn’t in the picture.

…And new parts for the SoundBoard…no surprises here, all capacitors..

Once thats done, its time to test the DC voltages coming off the power supply with nothing plugged in except the coin lights.

Perfect just shy of 5V for the CPU Board.

14V for the soundboard and CPU – a little higher than the 12V it requires but that is typical and should drop with an audio load anyway..

Next up, Lithium Battery Mod the CPU Board, change 3 filter caps on the CPU Board, Cap kit the G07 monitor, make a back door latch, make a log book.


Yep, put in the order from BOB Roberts again…should be here soon!  I sure hope he included some more of those moon-pies from Louisiana mmmm.

Got the Coin mechs that I ordered from TNT Amusements, great deal thanks guys!  These williams games are awesome.  If the game is shut off, it won’t accept quarters, it simply ejects them.  Too cool.

Also pulled the monitor chassis out in prep for the cap kit to arrive.  Wow, these things are always dusty and gross.  Nice G07-CBO monitor though, looks great with very little screen burn!

Removed the switcher and rewired the CPU board back to the original power supply which I started to cap with some extra caps I had laying around, looks much better inside…will finish when parts arrive won’t fire it up till its been rebuilt and tested to correct voltages, stablility etc.  Man, I can’t WAIT to play this game!!

This arcade is in great overall condition….no mice nests, very little wear and tear – a pleasure to work on!


Start simple…

Major cleanup of outside and inside of cabinet.  This baby is going all original.  The switcher is outta here…. The old school power supply is a simple thing of beauty and should work like a champ when rebuilt, no more worrying about losing high scores and user settings among other annoyances.

Coin mechanisms are on order.  I got the coin lights to fire 6.3V A/C (started rebuilding the powersupply already).

Also started working on the key mechanisms….Didn’t do this to Joust, damn…but this one will only require 1 Key to open all the locks, coin door, bottom door and back door too.  That screw in the side of the cabinet system just isn’t working for me.  I’ll post updates as I fix them.


Project number next…..Robotron!

Very similar to Joust, Robotron is built on the same platform as Joust using essentially the same board sets but with different roms and control panel. 

As arrived, here’s what she looks like….

To Do:

Fix sound issue, rebuild powersupply, joysticks, monitor, boards, etc.  Fix coin mechs and lights.  General clean up.


Yes you read it correctly, Joust finally works. Yahoo!

So after endless hours fighting the monitor, I finally took a break and came back with a fresh approach.  Turns out the monitor was working great, it just wasn’t being fed the V sync signal. 

The schematics show the V-Sync and H-Sync signal coming from chip A4 on the CPU PBA (A TTL 7486N chip).  So, I took my radio shack TTL tester and probed output of the H-Sync (pin #11) and got a series of beeps, typical…then, I probed pin #8 the V-Sync output and nothing but the low light and no beeps (no signal)..  Ah HA! Next, I checked the input signal for the V-Sync on the same chip (pin #10) and again got the beeping.  So, that part of the chip was being fed the signal, it was just stuck in the low state.

So I unsoldered the old chip, soldered in a socket and plugged in a new 74SL86 (see pic below).

Bingo Bango Bongo, the image FINALLY locks into place