Womble Posted January 21, 2011 Share Posted January 21, 2011 (edited) Finally got round to writing this one up, actually while the carpet is going down in the new toyroom :) Bought a faulty TMNT from Furballx a month or two back, it was actually the faulty one he was asking about 3 years ago! http://www.aussiearcade.com/showthread.php/10890-TMNT-pcb-problem-%28Graphics%29?highlight=TMNT I have a real soft spot for this game, it's the one I probably played the most in an original cabinet as the sports centre by my secondary school had this, its also the first board I ever tried to fix, I failed and that board is still in my scrap drawer. Anyway - it was quite an easy fix, the fault was as he described, solid bands through the graphics and a bit of a flickery mess.. http://i40.photobucket.com/albums/e230/Womble76/TMNT/DSCN7120.jpg http://i40.photobucket.com/albums/e230/Womble76/TMNT/DSCN7118.jpg http://i40.photobucket.com/albums/e230/Womble76/TMNT/DSCN7125.jpg http://i40.photobucket.com/albums/e230/Womble76/TMNT/DSCN7117.jpg ..it had also developed a new fault on the road south, the theme tune sample was missing and was replaced with a high pitched tone. The clattering in the movie I think is the auto focus on my camera having a small fit. http://www.youtube.com/watch?v=wHMAkIBRDgg Bad gfx on these board is often due to a failed mask ROM and when the board was put into test mode it would report that the mask roms at K4 and K6 were bad.. http://i40.photobucket.com/albums/e230/Womble76/TMNT/DSCN7113.jpg .. this would explain the fault and as there was no track damage or lifted pins on the custom ICs I took it's word for it. I didn't have any 40 pin sockets to fix that initially so I left it alone and focused on the sound problem. The theme tune sample is held in a single mask ROM at D4... http://i40.photobucket.com/albums/e230/Womble76/TMNT/DSCN7133.jpg ....which is the big chip in the above shot, the fault could be due to a failed MASK at this location, (the test routines didn't even attempt to test this one) but a quick check with the scope showed the address bus was doing absolutely nothing. This ROM is somewhat unique on the board in that it is addressed by a chain of 4-Bit Binary counters spread over 3 74LS393 chips at B7, C7 and D7. When working correctly these allow the board to scroll through the full address space on that chip with a single activation signal, the CPU has no involvment in playing the sample as the counters basically empty its contents directly into the sound system for decoding and DAC'ing. Each of the LS393 chips contain a pair of 4-Bit Binary counters whose outputs connect to the address pins on the mask ROM, the clock of one chip is provided by the output of the counter upstream. The master clock is provided to the first chip (C7) in the chain by the board, one of its outputs is the clock for the second counter on the same chip and thats ones of its outputs is the clock input for the 1st counter on the second chip, and so it goes on. The final chip in the chain only has one of its counters used so silence on the second ones pins is normal. As the second counter on the 1st chip was dead, chips 2 and 3 got go clock so would never do anything anyway. Replacing the 1st chip in the chain at C7... http://i40.photobucket.com/albums/e230/Womble76/TMNT/DSCN7138.jpg ...lit up the clock input for the counters on chip 2, the 1st counter fired up happily but again the second one on the chip was just as dead. Out of curioisity I fired up the board after replacing the 1st LS393 chip, instead of the high pitched whine I got the first guitar chord of the theme tune played over and over again for the duration of the full tune. Progress, but the address bus was still constricted, hence the repeated small segment. Replacing the second chip in the chain fixed the problem as all active counters now got their clock signals and had active outputs that provide the address bus for the ROM with the theme sample. The third chip in the chain is probably just as dead on its second counter, but it seemed pointless to replace a chip that's only half used and working on that half, with a chip that has a working second counter that would never get used. These were all the much dreaded Fujitsu TTLs which are renowned for failing these days, thankfully this board only has 7 Fujitsu TTLs on it, well 5 now, so it should give very little trouble in the coming years, some TMNT boards are 90% Fujitsu chips and are contsantly falling over as a result. When the boards were being made Konami would clearly buy the chips from which ever supplier was cheapest and when this board rolled into the flow solder machine Fujitsu were presumably undercut by other IC makers. When my sockets arrived I went on to the gfx issues, I desoldered the first of the two mask ROMs... http://i40.photobucket.com/albums/e230/Womble76/TMNT/DSCN7119.jpg .. (the rightermost pair of the 4 above) that the onboard diagnostics flagged as bad, these can be read as bog standard 27c400 eproms and my reader dumped the first chip and confirmed it was a perfectly healthy TMNT mask. The second mask rom had a number of out-of spec address bus pins and the dump was not recognised as anything useful. So I fitted a couple of machined-pin 40 pin sockets and put the healthy mask ROM back in place. I didn't have any 27C400s spare but I do have a scrapped TMNT board and the K6 mask rom from that board checked out fine, installing it in the new socket totally fixed the board. :) Edited February 1, 2011 by Womble Link to comment Share on other sites More sharing options...
kane Posted January 21, 2011 Share Posted January 21, 2011 Another good Womblog, now with video!! Link to comment Share on other sites More sharing options...
Frank_fjs Posted January 21, 2011 Share Posted January 21, 2011 Great game and great log. Link to comment Share on other sites More sharing options...
truxpin Posted January 21, 2011 Share Posted January 21, 2011 Great write up as usual mask rom and 74ls393 failure seems to inflict TMNT boards worse than any other board i have ever come across for some reason seen way too many with these problems. Link to comment Share on other sites More sharing options...
Hewitson Posted January 23, 2011 Share Posted January 23, 2011 The masks always go bad on these. Strangely, you don't often hear of problems with the Simpsons boards. Link to comment Share on other sites More sharing options...
mastercello Posted January 26, 2011 Share Posted January 26, 2011 Excellent read as always...i hope i can master these area one day like you. :) Link to comment Share on other sites More sharing options...
dave_wellington Posted January 26, 2011 Share Posted January 26, 2011 Awesome repair log Womble! Wish my brain could comprehend all of the bits, buses, addresses and mask parts, but i'm sure that'll come with time :D Link to comment Share on other sites More sharing options...
Davefjedi Posted January 31, 2011 Share Posted January 31, 2011 Good read cheers mate Link to comment Share on other sites More sharing options...
Administrators Arcade King Posted January 31, 2011 Administrators Share Posted January 31, 2011 Great write up as usual mask rom and 74ls393 failure seems to inflict TMNT boards worse than any other board i have ever come across for some reason seen way too many with these problems. Konami loaded their boards around this era with F (Fujitsu) IC's and 99.99% of the time they are always the culprits. Great log. Link to comment Share on other sites More sharing options...
Hewitson Posted February 1, 2011 Share Posted February 1, 2011 Fujitsu might be "Australia's Favourite Air", but they're certainly not Australia's favourite IC's. Link to comment Share on other sites More sharing options...
Womble Posted February 1, 2011 Author Share Posted February 1, 2011 Oh I dunno, without them there would be far fewer faulty boards to play with :) I wonder when they started to die actually, at the 21 year mark they have done pretty well, far longer than anyone expected these boards to be out of landfill. If they started dying mid 90s then thats another matter. Link to comment Share on other sites More sharing options...
TheGuru Posted April 10, 2011 Share Posted April 10, 2011 (edited) I read somewhere here that replacing all of the F chips is not really an option. I would like to correct that opinion :cool: I once fixed a Turtles board by checking the logic chips with a logic probe and replacing each one as I found it but the end result was that I replaced all of the Fujitsu chips (24) to get a totally dead board working. Edited April 10, 2011 by TheGuru Link to comment Share on other sites More sharing options...
cal2 Posted April 11, 2011 Share Posted April 11, 2011 Often it is easier to replace chips with new known working than loosing time to check them. At least you know they are working and you can take time on other parts of the pcb. This doesn't apply of course to all pcb since older pcb can have a LOT of them. Link to comment Share on other sites More sharing options...
Administrators Arcade King Posted April 11, 2011 Administrators Share Posted April 11, 2011 I read somewhere here that replacing all of the F chips is not really an option. Actually if you want a reliable PCB its the ONLY option. Link to comment Share on other sites More sharing options...
Womble Posted April 11, 2011 Author Share Posted April 11, 2011 True, some are all fujitsu so it's a major job! Link to comment Share on other sites More sharing options...
Recommended Posts
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 accountSign in
Already have an account? Sign in here.
Sign In Now