No announcement yet.

Mirco Challenge PCB Repair


Share any logs or helpful information you may have to repairs of arcade PCBs, Monitors etc

  • Filter
  • Time
  • Show
Clear All
new posts

  • Mirco Challenge PCB Repair

    Hi all,

    Firstly, thanks go to Kaizen a 'resident expert' on the Mirco Challenge for posting a heap of useful information including connector pinouts and repair logs of numerous PCBs all fixed successfully without the aid of any circuit diagrams - quite a feat. I'd like to add to the pool of resources here with some information I've compiled, including a component layout from my PCB. On it I've highlighted all the components which are known to have failed on various units, often causing specific symptoms and mostly documented in Kaizen's logs.

    This is the earliest coin-op video game I've collected so far. It also predates all of the consoles and obviously, all the home computers as well.

    'Challenge' is a Pong clone for 1,2 or 4 players and was made by Mirco Games in Phoenix, Arizona in approx. 1974. At the height of their popularity Mirco set up an Australian subsidiary and produced (or at least assembled) the same machine locally in Artarmon, Sydney - the only obvious difference being the screening on the glass.

    Click image for larger version

Name:	challenge.jpg
Views:	73
Size:	82.5 KB
ID:	2187002
    Now referred to as the 'Bronze Age of Videogames', the early to mid '70s machines were mainly Black and White units, some with colour screen overlays and using discrete logic components, prior to Microprocessors becoming commercially viable.

    The market was initially dominated by 'Ball and Paddle' games led by Atari's 'Pong' and many copies by other companies, some with minor variations, such as this.

    It intrigues me that Syzygy's (relatively) sophisticated space shooter 'Computer Space' released as early as 1971 did not capture the public imagination whereas Atari's (same people, new name) 1972 'Pong' was a huge success. I guess we all had to learn to walk before we could run...

    Click image for larger version

Name:	challenge1.jpg
Views:	46
Size:	73.5 KB
ID:	2187003
    As purchased, this particular unit had apparently been sitting uncovered for decades, in the garage of a house with water views. As far as I can tell the white haze on the table is probably from the salty air. Seen here with a few test cleaning spots to find the best method for removing the haze.

    Click image for larger version

Name:	challenge2.jpg
Views:	47
Size:	89.9 KB
ID:	2187004
    When powered on, it still shows the 'Attract Screen' with the ball bouncing around and the score displays are present but there are reportedly issues displaying the paddles and some of the controls are frozen after all this time being unused.

    The coinmech is jammed, full of assorted denominations (presumably it is set for 20c coins) and as yet I don't have a key to open the door and clear it. I'm going to have a go at picking the lock rather than immediately resorting to drilling it out.

    The lock isn't a high security model so picking it turns out to be easier and less messy than drilling. The machine obviously hasn't been opened in ages and needs a really good clean inside...

    Click image for larger version

Name:	challenge3.jpg
Views:	47
Size:	103.8 KB
ID:	2187005
    There's a popular legend that Atari engineers were called to repair their prototype Pong machine which had stopped working while being tested on location. Upon investigation, the engineer who attended found the machine had been so popular that the coin box had filled and coins had banked up, jamming the mechanism.

    Well, my faint hopes that a similar situation might exist here were not realised but there was a good pocketful of 20c coins in the coin tray as well as a few ring-ins jammed in the mech. Finding any coins in the machine is always good and the same coins can be used over and over, to play the games in this collection.

    Even better, there were enough 20c coins in the tray (just over 60) to estimate its last use. Coin dates ranged from 1966 (the start of Decimal Currency) to 1979 but nothing from the Eighties at all so the machine has probably been 'retired' for 40 years.

    And while on the subject, I'll even hazard a guess why. There was some sticky tape residue, visible in the first photo on this page which was never cleaned off, just where you would place an 'Out of Order' sign. I suspect it developed a fault, was taken out of service and never used again. By 1979 or '80 it was already obsolete. Home systems were becoming common, offering paddle games and much more while new coin-op machines such as Space Invaders and Galaxian took the amusement industry by storm.

    Having said that, the coin meter reads 17800 - assuming it started near zero when first used that translates to about $3500 which I imagine would have paid the machine off a few times over during its working life...

    Click image for larger version

Name:	challenge4.jpg
Views:	47
Size:	87.0 KB
ID:	2187006
    So, coining it up confirms there is an issue with the logic. 3 of the 4 paddles display OK though only one moves freely (Potentiometer problems most likely) but the left hand paddle is replaced by a solid wall (similar to the single player game but without the 'goal' area) making the left hand players invincible! I doubt a faulty control could cause that though one pot is very loose and might have a broken wire so I will check that first.

    Click image for larger version

Name:	challenge5.jpg
Views:	46
Size:	77.4 KB
ID:	2187007
    There is also no sound at all. That problem could be as simple as a faulty speaker or wiring but could equally be caused by a fault on the game board or in the audio amplifier.


    I've replaced all four potentiometers and fitted new knobs, the nearest in size and style to the originals which I could find so far. I would have liked to retain the original knobs but the rim around the top edge was chipped on a couple of them, making them feel a bit jagged. Two were also missing their aluminium insert, the other two inserts were very corroded.

    Click image for larger version

Name:	challenge6.jpg
Views:	46
Size:	79.7 KB
ID:	2187008
    Having replaced the pots, all four controls now operate smoothly. The one which was loose (top left) did have a broken wire, this one corresponds to the second (inner left) paddle. There is still an issue with the outermost left paddle, with the new pot it now moves up and down but there is no space between it and the top boundary, only below so it appears more like a sliding door. In Player(s) versus machine mode the 'goal' area is still completely closed.

    There's one more issue which I've noticed, when the first coin is inserted only the outer paddles should be displayed but both right hand paddles are present as well as the outer left one. Once a second coin is inserted the inner left paddle correctly appears.

    So it's time to take the game board out, make up a test cable and troubleshoot the logic. I haven't been able to source a circuit diagram as yet so it looks like I'll have to do it the hard way, tracing from some known point such as the video output or control inputs and making a 'mudmap' as I go along. The other possibility is to probe each chip in turn using the relevant datasheets and look for irregular behaviour, missing or incorrect levels on inputs or outputs etc.


    Making a bench test cable for the Mirco Challenge PCB is not as straightforward as I imagined - instead of an edge connector there are four PCB mounted multi-pin connectors with an unusual footprint, different to any Molex style connector I have seen. The Molex socket-pins which I have are also a larger diameter than the pins of the PCB plugs, making them a loose fit.

    Edit: they appear to be "AMP Commercial Mate-N-Lok" type connectors but I haven't found a readily available local supply of them. For testing purposes I don't need the locking function - or the expense of specially ordering these from overseas.

    Fortunately I do have some other multi-pin plug and socket sets which I use to make up test leads, usually where there is no need to match up with an existing connector. Although the connector body doesn't match the Mirco's components, at least the socket-pins are a good fit for the PCB mounted plug-pins. These connectors are available from Jaycar stores in Australia or online. I'll need 23 socket-pins so 2 packs of the 12 way connector set (see photo, right) will be sufficient.

    Click image for larger version

Name:	challenge7.jpg
Views:	45
Size:	84.5 KB
ID:	2187009
    I could just use heatshrink tubing on each pin but then for testing would need to connect each wire to the PCB individually which is not ideal. Instead I think I can make up some new socket bodies with the 3D printer, to use with these pins. The tolerances are pretty tight as the pin spacing (pitch) of the PCB plugs is only 5mm. I'll have to make the pin receptacles narrower than the originals, still providing a groove for the locking barbs which retain the pins in place.

    Click image for larger version

Name:	challenge8.jpg
Views:	44
Size:	102.2 KB
ID:	2187010
    My 3D printer is set up for ABS filament and the narrow walls between the pins will test its capability. I had planned to print the sockets in colour-coded material but find that the 'natural' ABS filament without additional pigments flows a little better for very fine detail, with my 3D printer at least. These might print a little better with PLA filament and would be great using a laser type 3D printer.

    connectors.123dx 3D drawing
    6way.stl 3D object
    10way.stl 3D object
    Click image for larger version

Name:	challenge9.jpg
Views:	44
Size:	139.8 KB
ID:	2187011
    The finished products can be seen here, connected to the Challenge PCB. Although there were a few imperfections in the 3D printed housings, in practice they work perfectly, retaining the pins and keeping them separated.

    Click image for larger version

Name:	challengea.jpg
Views:	46
Size:	83.4 KB
ID:	2187012
    The test panel evolved into this full-function wired controller. All 4 Potentiometers are connected as well as the game select switch and start button. An illuminated Arcade button serves as the coin input - its Single Pole Double Throw microswitch is similar to the switch on a coin mechanism. The 'coin door illumination' supply from connector P4 is wired to its internal LED. The Green LED is connected to the 'Free Game' signal also on P4 while the Yellow LED indicates the 'coin counter' signal from P3.

    Click image for larger version

Name:	challengeb.jpg
Views:	46
Size:	84.7 KB
ID:	2187013
    Connector P2 comes from the 5V power supply and also goes to the Audio / Video monitor. This Dick Smith Electronics late-model CRT TV (actually made by Konka) doesn't lend itself to RGB modification but can display Composite or Component video at 50 or 60Hz framerate - in this case Monochrome Composite video, 60Hz.

    Once the game is 'coined up' and started the audio blares from the TV speaker. At least that part of the game board works perfectly though there must be a fault within the audio amplifier or speaker wiring, in the Mirco cabinet.


    The Mirco Challenge PCB has made its way onto my test bench so it's time to troubleshoot the two known issues. Firstly, the left most paddle positioning is working however the space above the paddle is displaying as a solid wall. Additionally when only one coin is inserted both right hand paddles are displayed - only the outermost paddles should be present in this situation.

    Click image for larger version

Name:	challengec.jpg
Views:	46
Size:	65.2 KB
ID:	2187014
    I still haven't been able to source a circuit diagram for this PCB so we'll have to work it out as we go. Referring to documentation for the original PONG and other clones may help in a general sense however differences in the exact detail between systems can cause confusion so it may be best not to overload with irrelevant information.

    A good start is to compile a PCB layout showing the component numbers and their locations. As there is no CPU the circuit functions are determined entirely by interconnection of the TTL logic components. The ICs are laid out on the PCB in a grid fashion. There is also a crystal at the 'top' end (row 1) while all 4 connectors for power, controls and outputs are at the 'bottom' end of the PCB. Based on that I'd assume repetitive functions such as sync generation occur near the top while interactive functions such as paddle positioning are handled closer to the inputs at the bottom end.

    mirco.pdf Mirco Challenge PCB layout PDF document
    mirco.ods Mirco Challenge PCB layout spreadsheet

    Browsing some documentation for the original Atari Pong PCB, there were two 555 timer ICs which determined the 'vertical' delay from top of screen to the two paddles based on the position of their respective control potentiometers while another two '555's controlled the serve timing and the duration of the game 'score' sound.

    The Mirco PCB has two 555 timers near the top end of the PCB which wouldn't be sufficient to control the position of 4 independant paddles anyway. Instead I'm assuming these are for the serve timing and sound duration as seen in the Atari example.

    Right next to the P1 connector however there is a LM339A Quad Voltage Comparator which can be used to create the required 4 timing signals. After 'coining' the pcb up to display all four paddles as per the above picture, checking the LM339A's outputs with an oscilloscope confirms this as our starting point.

    Each output of the LM339A is a rectangular wave whose mark / space ratio varies with movement of the corresponding player control. The transition point which varies will determine the vertical position, being the top of each paddle.

    At this stage all four signals look correct so tracing from this point forward we can compare the working 'Player 2' circuit to the faulty 'Player 1' section and find the point where it goes wrong for Player 1, mapping the circuit out as we go. It's not necessary to check every pin of every IC along the way, if both signals are correct at the output of each stage we can move on to the next.

    Click image for larger version

Name:	challenged.jpg
Views:	46
Size:	39.1 KB
ID:	2187015

    <iframe width='500' height='300' frameborder='0' scrolling='no' src=''></iframe>

    From the LM339A at location A10, both signals pass through two stages of 7402 NOR gates at B10 and C10 arriving at the 74109 J /K flip flop at D10. The inverted /Q outputs enable a pair of 74161 4 bit counters at G10 and E10. These counters would be clocked by a horizontal frequency signal, effectively counting the lines for which each paddle is active. The output data will also be used to determine the bounce angle of the ball, from horizontal for a collision at the centre of the paddle to sharply angled at either edge.

    For the Player 2 circuit the /Q output also goes directly to a 7420 Dual 4 input NAND gate at location F10 whose output at pin 6 I've labelled 'Paddle 2'. It's a pulse which would correspond with the 'active' duration of the Player 2 paddle, shifting in time with movement of the Player 2 control input.

    Assuming the other half of the 7402 at location F10 is for Player 1, I check the other output at pin 8 and it's all wrong. Checking its inputs, the signal arriving at pin 12 is noisy and low level, causing the incorrect output. Tracing back from there, it turns out instead of being directly connected from the 74109 at position D10, its non-inverted Q output travels through an additional 7402 NOR gate at location F8 which is faulty - with noisy non-TTL level signals on multiple pins.

    So why the additional gate? I think it's used for the Player Vs Machine mode where the left hand paddle is replaced by a wall with a 'goal' area. In Player Vs Player mode the signal should pass through, albeit inverted which is why the non-inverting Q output from the 74109 is used instead of the inverted one.

    Replacing the faulty 7402 at location F8 with a new 74LS02, also installing a socket - solves that problem. Surprisingly it has also solved the secondary problem where both right hand paddles (3 and 4) were appearing with a single coin input. It turns out that was caused by another gate in the same IC which should disable the paddle for Player 3 when only 1 coin has been inserted - its input on pin 12 going High, then Low again once a second coin is received.

    That was a bonus, fixing both faults with a single IC replaced! If ever these two faults are observed on a Challenge PCB the 7402 at location F8 is almost certainly responsible - it is likely the only component which these two unrelated functions have in common.

    After a bit of time to test out the repaired PCB I'll reinstall it in the machine and look into the lack of sound which is not a fault of the game board. It must be a speaker or wiring problem or a fault in the audio amplifier which is part of the monitor in this machine.

    The only feature which I've been unable to test so far is the 'free game' awarded in Player Vs. Machine mode if the player(s) on the right hand side can score highly enough by deflecting the ball obliquely through the 'goal' area without missing the ball enough times to lose the game. So far, I've not had the reflexes or the patience to achieve that...

    Regards, John

    more from John's Retro Workshop -

  • #2
    It is hard, but the free game is so worth it. I'm yet to then beat the free game...


    • #3
      Hi Eugy, that would definitely be something, to score enough 'goals' to win a free game then win again!

      It would be interesting to see what happens exactly but there's a clue in the game manual, link below. (I think that upload was also courtesy of Kaizen )

      I would need a lot of practice to even unlock the free game, not sure the concentration will hold up.

      As a technician rather than a gamer as such I tend to look at the machines a bit differently, mid game I'd probably notice some minor adjustment required to the monitor or something and that would be 'Game Over'.....

      All the best, Regards, John
      more from John's Retro Workshop -


      • #4
        Thanks for the link to the manual John (and thanks Kaizen)! Very interesting reading!


        • #5
          Hi Eugy, here's some photos of the original knob, pot and game select switch artwork you asked for - and for anyone else who's looking for them.

          The top glass on my machine says it's made in USA so I'm not sure if the locally assembled version was identical or not.

          Hope this helps, Regards John
          Attached Files
          more from John's Retro Workshop -


          • #6
            Wow!! Thank you so much for all your help, John! Thank you for the detailed photos and measurements, its a massive help!!


            Users Viewing Topic: 0 members and 1 (guests)