Jump to content
Due to a large amount of spamers, accounts will now have to be approved by the Admins so please be patient. ×
IGNORED

Pacman PCB Repair 20/05/2019


Recommended Posts

  • Administrators

Since I got the bug I decided to have a crack at a POS Pacman PCB that's been sitting neglected and dumped in the cupboard of "someday I'll get to it" for probably the last 10 years or more.

The condition of the PCB was pretty dirty and sad with a lot of IC's with corroded legs so there wasn't any incentive to fix it. I know I tried to fire this thing up in the past and it was dead as a door knob.

IMAG0096.jpg

Where to start?

First I did the single rom hack as pictured above. I prefer working on the board like this as the sockets are notoriously bad and I don't want to deal with them yet. This hack also connects the rom directly to the address buss and bypasses several IC's that handle the chip enable.

I tested all 2114 rams in the other board and replaced a missing one.

 

Fault: Board Dead

When dealing with a dead board I start by checking the clock at pin 6 on the cpu which is bad.

Untitled-1.jpg

This is the clock circuit. It basically takes the 18 mhz signal and breaks it down into 6mhz and 3mhz for the cpu.

 

Fix: Not an easy fix unfortunately. The 74LS368 @ 8B is bad. Nothing coming out of pin 7.

IMAG0104.jpg

We now have a 6 mhz signal however still no 3mhz cpu clock we need to dig further down.

Untitled-3.jpg

We get our 3mhz signal from a 74LS74 at 8C Pin 5

IMAG0097.jpg IMAG0102.jpg

I took a pic of 8C so you can see how shit it looks. I replaced the 74ls107 @ 8A just to be on the safe side and a 74ls245 @ 4H which is a common failing IC

We now have our 3mhz clock on pin 6 of the cpu.

 

Fault: Board still dead. Address lines are active but data buss is not moving, watchdog doing nothing.

Fix: Untitled-3.jpg

Basically if your horizontal and vertical counters aren't working the cpu wont start.

IMAG0116.jpg

74LS161 @ 2S, 2R and 3R outputs were all stuck. Replaced them as above with TI 74LS161's.

IMAG0105.jpg

Just to confirm. These chips were "SA" branded, never seen them before.

IMAG0107.jpg IMAG0108.jpg

Fired her up and game now works however no character sprites and colour issues. Static ghosts have no eye's and Clyde's colour is wrong.

 

Fault: Game plays blind.

Fix: Untitled-4.jpg

From the last repair where all the characters were playing in the bottom half of the screen I knew the issue was going to be in the Attack Ram circuit and I had a pretty good idea which chips were at fault.

IMAG0115.jpg

SA branded 74LS161's @ 1E and 2E outputs were both stuck (top left).....anybody see a pattern? 2A is also important and I'll come back to that.

IMAG0109.jpg

We now have our characters back! but with colour faults....

 

Fault: Colour issues. Ghosts have no eye's and Pacman Red.

Fix:

IMAG0098.jpg IMAG0101.jpg

2A had been damaged so I removed it before starting this repair. Replaced damaged D2115 ram with a D2125 I have a heap of these and they finally came in handy. The D2125's run soooooo much cooler also.

IMAG0111.jpg

Fault: Pacman back to normal and Ghosts have eyes however Clyde is wrong colour (looks the same as Pinky) so is the power pellets which are orange (which Clyde should be) and the text and dots seem to be an "off white"

Fix: I started by probing around the Attack Ram circuit but couldn't find any faults so the issue has to be in the final stages of the colour circuit.

Untitled-6.jpg IMAG0113.jpg

and here's our culprit R33 a 220 ohm resistor part of the blue RGB signal. The leg was still in tact so looks like somebody lifted it for trouble shooting purposes.

Its worth pointing out that the Attack Ram does not effect the colour of the static ghost images in the attract mode and considering both static and moving sprites were effected the bad resistor makes sense.

IMAG0121.jpg IMAG0112.jpg

Replaced the resistor, game graphics and colour now 100%

 

Fault: Sound.

Nothing serious here. While I was working on the board I was getting noisy feedback though the amp and the music sounded kind of high pitched. The amp was also getting hot sue to oscillation.

IMAG0120.jpg IMAG0117.jpg

Fix: Replaced the following components.

Broken .01uf ceramic cap @ C1

2200uf cap @ c29

330uf cap @ c51

10uf cap @c10

0.47 Tantalum cap @C49 (this was missing)

Missing 1N914B @ D1 put in a 1N4148

IMAG0122.jpg

Board now 100% and running with 2532 roms.

Things to do.

Replace program rom sockets.

Give board a good clean with compressed air, I know its dirty as hell!

To future proof the board I'll probably change the remaining 2 SA branded 74ls161's @ 3s and the one for the watchdog @ 9C

 

Damn this took forever to write up! but the repair was fun.

 

Tip. I'm was using one of those RGB/CGA to VGA boards so I could work on this using a LCD screen however both Pacman boards sync wouldn't lock in....annoying as hell.

IMAG0119.jpg

Added a 220ohm resistor in series to the sync wire which worked great.

 

UPDATE: 30/03/2019

Just an update from soak testing the board today, no surprises a few faults popped up.

[ATTACH=CONFIG]148774[/ATTACH] [ATTACH=CONFIG]148775[/ATTACH]

Put some new sockets in. The board was just too flaky, just breathing on it made it reset...or worse.

[ATTACH=CONFIG]148776[/ATTACH]

Heat shrink on the caps to keep @channelmaniac from having a meltdown.

[ATTACH=CONFIG]148769[/ATTACH]

2114 ram fault developed at 4L

 

Fault: Blank screen. Can coin game up and play blind with all sounds working.

Fix: Poking around the characters rom/Attack ram section the roms were active however all the attack rams outputs were sitting high.

[ATTACH=CONFIG]148771[/ATTACH]

I found Pin 5 and 6 on the prom at 4A stuck low. Pulling the prom out gives you a white screen and you can just make out moving blocks in the background.

[ATTACH=CONFIG]148770[/ATTACH]

Traced it back to a bad 74LS157 @ 5A...another bad SA chip

 

Last fault was a doozy.

Fault:Game randomly resets and does weird shit. Sticking it in test where it shows you the dip switch settings, all the settings were going crazy and changing by them self.

Normally I would have suspected the 74LS367 buffers but the problem was going across several controls and the dip switch settings are across 2 IC. It is possible for a one on the way out to effect the entire data buss but I wasn't convinced. Eventually it got worse check out the vid below

 

Fix: Swapped out the sync buss controller (little rider board that sits above cpu) from the other board and the problem vanished. Suspect the 74LS244 but will confirm when I get time to have a further look at it.

 

6/619

Ongoing issues with this poor old neglected board.

 

In my last log I wrote about a fault with the dip switch settings going nuts in test. Unfortunately this only got worse as the real culprit failed.

 

Fault: Game just goes nuts. Will generally boot when cold but is very picky about the ram, CPU and even sync buss controller used. Watchdog isn't barking it all so the program is just crashing.

Sometimes it boots coining itself, sometimes boots into the game however the right hand side of the screen starts filling with random letters and characters. When Pacman reaches the right hand side of screen he will continuously move down the edge of screen from top to bottom until it returns to attract mode. Ghosts will follow the same pattern if they hit the edge of the screen (see pics.) Further more to that the sound was funny.

[ATTACH=CONFIG]149065[/ATTACH] [ATTACH=CONFIG]149066[/ATTACH]

[ATTACH=CONFIG]149067[/ATTACH]

Eventually it would just boot into Bad V Ram-0

 

Fix:Annoying fault to say the least. After going over the data buss with a fine tooth comb I couldn't find any issue so I turned to the ram address buss in particular the chip select circuit for the main scratch pad rams.

[ATTACH=CONFIG]149068[/ATTACH] [ATTACH=CONFIG]149069[/ATTACH]

Long story short I found pins 5 and 6 shorted on a 74ls02 @ 7L you can see the odd signal on the scope however removing the IC the short remained.

Traced it back to a bad 74ls138 @ 7J pins 5 and 6 shorted. Looking at the schematics it now makes sense as to why so many things were effected and going haywire, it's amazing I could get the board booting at all. I was able to get the board booting by just touching my probe on pin 1 of the 74ls02 which gave me my first clue.

[ATTACH=CONFIG]149070[/ATTACH] [ATTACH=CONFIG]149071[/ATTACH]

Things are looking much better on those two pins now.

 

Fault. Board ran for half the day but once again another problem popped up.

[ATTACH=CONFIG]149072[/ATTACH]

Fix. 74ls273 @4D just decided to pack it in.

 

After all that I'm still getting an intermittent graphics fault...that's tomorrows problem.

 

I didn't think this board was going to be easy to fix and was expecting more problems to pop up and I'm sure this isn't the end of them but at least the program is running 100% now without any crashes. I have no doubt this pcb hasn't been fired up in over 20 years.

Edited by Arcade King
Link to comment
Share on other sites

Seeing as you knocked that one over so quickly, it only took me 5 minutes to read your whole post:lol, I have a Tube Panic board here. Last time I powered it up, 20 years ago it worked but in 20 years a lot can go wrong aye.

There is also a Defender board here that I know does not work......Just saying seeing as you are on a roll. There may be a Jungle King board and a few others. Can't have you loosing your touch.

Link to comment
Share on other sites

Or you could keep all the fun for yourself and have a go at repairing them :lol

 

I thought some of those titles you may be interested in for yourself.

 

I don't want them back, you repair then, they are yours mate.

Link to comment
Share on other sites

  • Administrators

Just an update from soak testing the board today, no surprises a few faults popped up.

IMAG0123.jpg IMAG0124.jpg

Put some new sockets in, the board was just too flaky, just breathing on it made it reset...or worse.

IMAG0127.jpg

Heat shrink on the caps to keep @channelmaniac from having a meltdown.

IMAG0125.jpg

2114 ram fault developed at 4L

 

Fault: Blank screen. Can coin game up and play blind with all sounds working.

Fix: Poking around the characters rom/Attack ram section the roms were active however all the attack rams outputs were sitting high.

Untitled-2.jpg

I found Pin 5 and 6 on the prom at 4A stuck low. Pulling the prom out gives you a white screen and you can just make out moving blocks in the background.

IMAG0128.jpg

Traced it back to a bad 74LS157 @ 5A...another bad SA chip

 

Last fault was a doozy.

Fault:Game randomly resets and does weird shit. Sticking it in test where it shows you the dip switch settings, all the settings were going crazy and changing by them self.

Normally I would have suspected the 74LS367 buffers but the problem was going across several controls and the dip switch settings are across 2 IC. It is possible for a one on the way out to effect the entire data buss but I wasn't convinced. Eventually it got worse check out the vid below

 

Fix: Swapped out the sync buss controller (little rider board that sits above cpu) from the other board and the problem vanished. Suspect the 74LS244 but will confirm when I get time to have a further look at it.

Edited by Arcade King
Link to comment
Share on other sites

Re the supergun / video encoder.

 

Happy to upgrade you to one of my supergun's. There's a VGA connector directly on the supergun so you can run a VGA cable directly between it and the video encoder. Also has RGB scart, s-video and composite video output should you desire, not to mention properly attenuated and buffered video output. I don't have any assembled right now but since you can solder happy to buzz one your way in kit form. :)

 

IMG20190111221633.jpg

Link to comment
Share on other sites

  • Administrators
Re the supergun / video encoder.

 

Happy to upgrade you to one of my supergun's. There's a VGA connector directly on the supergun so you can run a VGA cable directly between it and the video encoder. Also has RGB scart, s-video and composite video output should you desire, not to mention properly attenuated and buffered video output. I don't have any assembled right now but since you can solder happy to buzz one your way in kit form. :)

 

https://www.aussiearcade.com/attachment.php?attachmentid=148773

 

Hook me up mate and I'll fix you up.

I bought this one from Germany a few years back, the appeal was the scart connector. Usually I'd use a Commodore 1084 but I'm too lazy to lug it out just for a few repairs.

Link to comment
Share on other sites

  • Administrators
Do you have any info or links to that little tester your using?

 

Just cheap shit from ebay mate

https://www.ebay.com.au/sch/i.html?_from=R40&_trksid=m570.l1313&_nkw=ic+tester&_sacat=0

 

It only does gates,counters etc I need to find one that does buffers,octal transceivers and what not. Having said that its a neat little thing to have.

Link to comment
Share on other sites

I use this one...

 

https://m.aliexpress.com/item/32716880687.html?pid=808_0000_0101&spm=a2g0n.search-amp.list.32716880687&aff_trace_key=492ebf65a35c4dbe934e8b3ee3102409-1556801582801-06759-KnSGCh6&aff_platform=msite&m_page_id=6993amp-4-bMx2a3T1T3AtlohR-K9g1559360666131

 

 

I have the USB powered version which seems to no longer be available.

@Arcade King I'm happy to test any chip numbers that you want to know if it works with them.

Link to comment
Share on other sites

  • Administrators

Thanks mate all this discussion I ended up buying one of these to try.

https://www.ebay.com.au/itm/Digital-IC-Tester-Detect-IC-Tester-Meter-Maintenance-Transistor-Tester-MOS-PNP/181861993009?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2060353.m2749.l2649

 

Probably double the price but I noticed it had 74HC245 on the screen which is one of the types of IC lacking from the other one, I've also got an old school logic comparitor but hardly if ever use it.

I'll be honest I still rely far more on my trusty Elenco LP560 logic probe and scope than one of those chip testers. Problem I have with the chip testers is there's no stress test so you can get chips that say their ok but on the board they start playing up in a few minutes. Still good to verify if an IC is completely dead though.

I'll let you know what I think of this tester when I get it.

 

Just a heads up though if anybody is looking for one of those Elenco LP560's I bought 2 from Amazon 2 weeks ago for $15 usd ended up costing $80AU for 2 with postage. Cost over $60 just for one on ebay. I've never found a better probe than these.

Link to comment
Share on other sites

I usually use my tester to confirm the chip I remove is faulty or if I have a random chip floating around that I can't remember if it was good or not.

If in doubt, for the cost of chip it's better to just replace it.

 

The Elenco's are great, I bought mine 4 years ago and never had an issue with it though the next one I buy will have audible tones so I dont have to shift my gaze from the chip to the LEDs.

Link to comment
Share on other sites

I usually use my tester to confirm the chip I remove is faulty or if I have a random chip floating around that I can't remember if it was good or not.

If in doubt, for the cost of chip it's better to just replace it.

 

The Elenco's are great, I bought mine 4 years ago and never had an issue with it though the next one I buy will have audible tones so I dont have to shift my gaze from the chip to the LEDs.

 

Don't wait. Get it NOW. You'll hear a distinct difference with racing gates, weak outputs, and more. It'll greatly expand your troubleshooting.

Link to comment
Share on other sites

  • Administrators

Ongoing issues with this poor old neglected board.

 

In my last log I wrote about a fault with the dip switch settings going nuts in test. Unfortunately this only got worse as the real culprit failed.

 

Fault: Game just goes nuts. Will generally boot when cold but is very picky about the ram, CPU and even sync buss controller used. Watchdog isn't barking it all so the program is just crashing.

Sometimes it boots coining itself, sometimes boots into the game however the right hand side of the screen starts filling with random letters and characters. When Pacman reaches the right hand side of screen he will continuously move down the edge of screen from top to bottom until it returns to attract mode. Ghosts will follow the same pattern if they hit the edge of the screen (see pics.) Further more to that the sound was funny.

IMAG0144.jpg IMAG0147.jpg

IMAG0142.jpg

Eventually it would just boot into Bad V Ram-0

 

Fix:Annoying fault to say the least. After going over the data buss with a fine tooth comb I couldn't find any issue so I turned to the ram address buss in particular the chip select circuit for the main scratch pad rams.

Untitled-2.jpg IMAG0148.jpg

Long story short I found pins 5 and 6 shorted on a 74ls02 @ 7L you can see the odd signal on the scope however removing the IC the short remained.

Traced it back to a bad 74ls138 @ 7J pins 5 and 6 shorted. Looking at the schematics it now makes sense as to why so many things were effected and going haywire, it's amazing I could get the board booting at all. I was able to get the board booting by just touching my probe on pin 1 of the 74ls02 which gave me my first clue.

IMAG0150.jpg IMAG0151.jpg

Things are looking much better on those two pins now.

 

Fault. Board ran for half the day but once again another problem popped up.

IMAG0155.jpg

Fix. 74ls273 @4D just decided to pack it in.

 

After all that I'm still getting an intermittent graphics fault...that's tomorrows problem.

 

I didn't think this board was going to be easy to fix and was expecting more problems to pop up and I'm sure this isn't the end of them but at least the program is running 100% now without any crashes. I have no doubt this pcb hasn't been fired up in over 20 years.

Link to comment
Share on other sites

  • Administrators

Here's my final log on this board (hopefully).

IMAG0145.jpg

Forgot to put this up last log.

While probing around I found a bunch of floating pins on the 4 74ls161 counters. R59 was completely missing, it's supposed to be a 1k resistor so I put one in.

 

As mentioned in the previous log I was getting intermittent graphics glitches both separate issues.

 

Fault: Ghosts started flickering changing colour then finally horizontal coloured bars across the play/maze area of the screen

IMAG0159.jpg

As you can see only the maze area is affected.

Fix: I've read that a failed attack ram(s) 2A-2D can cause this issue however I couldn't find any obvious issue with the data or address lines.

Looking at the schematics I suspected either 3A or 3D as the culprit. Since 3A 74ls158 looked the worse with rusty legs I just swapped it out and was right. Game now 100%...for now.

I've also desoldered the last 3 attack rams HM2510-1 (3B-3D) and replaced them with MCM2125's I had floating around which run soooo much cooler.

IMAG0160.jpg IMAG0153.jpg

Picture on the left shows how bad the rust on the TI 74ls158 is along with the crappy looking HM2510 rams. Picture on the right I meant to put up earlier which shows more crappy IC's and you'll notice legs have been snipped on the two IC's top right. Somebody had a good crack at repairing this board in the past and did the old snip the leg trick to find the culprit. For reliability sake I've replaced all IC's I could find that Mr Snippy has been at which was around 5 or 6.

 

Last fault.

IMAG0164.jpg

The picture says it all. It started off intermittent and eventually sprites were affected which weird little sparkly blocks at the top of their heads.

Fix. Pretty easy fix, usually when I see faults like this I head straight for the 74LS86's

Untitled-3.jpg

Culprit was the 74ls86 @ 2H the outputs were weak. I could easily replicate the fault by rubbing my finger along the side of the rom pins 8-6

For future proofing sake I also replaced the 74ls86 @ 4F both were rusty TI (Texas Instruments) branded IC's. TI seem to be prone to rusty legs.

A lot of the IC's I've replaced on this board were the SA brand or rusty TI's. I think I've replaced almost if not more than half the IC's on this board and pretty much encountered every fault possible repairing it. There's only 2 TI chips left 74ls08's I'll probably replace these however the board ran 8 hours straight without a missing a single beat so I'm pretty happy with the effort.

IMAG0165.jpg

If this was a anything other than a original board I wouldn't have wasted my time.

Edited by Arcade King
  • Like 1
Link to comment
Share on other sites

  • 2 years later...
Quote

Replaced damaged D2115 ram with a D2125 I have a heap of these and they finally came in handy.

@Arcade King Sorry to bump an old thread! I'm looking to buy a couple D2125A-2 or similar RAMs to finish off an Exerion PCB repair if you still have any spare to sell, let me know.

I did order a couple from OS months ago but they must be on a slow boat somewhere... Thanks if you can help, great repair log btw! Regards, John

Link to comment
Share on other sites

It's sad to see 'high tech' components affected by such a fundamental issue as rust. This from my Taito SI repair...

taitoza.jpg.e0067e321c1d0493ac0d56ba0b5cae9b.jpg

On a brighter note, while reading this log I saw your collective recommendation for the Elenco LP560 logic probe - I've been meaning to replace my El-cheapo logic probe with something better and took all your advice, finding them still available on Amazon and it has arrived already. Can't wait to try it out on my next repair, probably the Atari Centipede PCB.

Thanks for the tip. Regards, John

 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...