Tuesday, 1 August 2017

Kung Fu Master pcb repair

I've had this Kung Fu Master pcb sat in the to do pile for a year or two now as the custom KN6034201 chip on the top pcb was bad and I wasn't able to find a replacement.

It caused the background graphics to just show up as solid colour blocks.


A big thank you goes out to Caius for not only making a modern day replacement for the custom but also for selling me one of his spare pcbs.

 The KN6034201 reproduction pcb populated and ready to be fitted.



All graphics are now 100%, the only issue remaining is that there is no sound. This turned out to be a bad 6803 cpu, I confirmed this by swapping it into my other Kung Fu Master board. The 6803 cpu seems to be pretty rare at least when it comes to video games, they were more common in pinball machines so I didn't have any on hand.

I ended up ordering some from utsource (who I don't recommend) but they were very cheap so I ordered a few, this was lucky as the first one I tried was DOA. The next one I tried worked and the sound was fully restored.

Repair complete, thanks again Caius for making this repair possible.

Friday, 28 July 2017

Donkey Kong pcb repair

This two stack Donkey Kong pcb had all sounds apart from the music at the how high screen and the background music. A bit of basic research suggested it was most likely caused by bad EPROMS 3H or 3F, the 8035 or LM324 amp. All these were fine so using the schematics I made a note of what else was connected to EPROM 3H (the one that contains the music) the hp comparator flagged all outputs as being bad on the first TTL I checked a 74LS374 @ 4E.


Piggybacking a good 374 brough the music back, I fitted a socket and the new 374 repair complete.

Tuesday, 4 July 2017

Robocop pcb repair

This Robocop pcb booted up but the majority of the graphics were missing with only some text and the life bar showing up. The first thing I did was try the rom board on a known good main board and it worked perfectly so I knew the issue was with the main board.


 

After ruling out all the ram and ttl I started to look at the surface mount customs, pins can often lift on these. I reflowed them all but it made no difference. It went back in the pile after that for a few months until I pulled it out again today and decided to check the only two things I hadn't already which were the two prom chips. Of course it could have been a bad custom but I like to rule everything else out before I even think about replacing those.

Comparing the outputs to the ones on a good board I found that pins 13 and 14 of 7112E prom @E17 on the bad board were stuck low.

I had two scrap data east main boards from a trade many years ago. They were scrap in every sense of the word, the previous owner took a drill to the cpus trying to remove them and cut away parts of the board trying to remove certain chips, but luckily the proms were still there.


I fitted a socket and the 'new' prom.




This restored all the missing graphics, repair complete.

Thursday, 22 June 2017

Thunder Cross II pcb repair

I bought this board originally just to use for parts as I always need konami custom chips. However when it arrived I could see it was in really good condition apart from the missing 051550 custom near the edge connector and a couple of damaged traces. So I decided to see if I could get it going.

I took a 051550 from a Main Event parts board soldered it in and fixed the three damaged traces.


Powering on the board I was greeted with a screen of garbage, the cpu wasn't running. I checked the two main program maskroms with the eprom burner and both returned unknown code. I then burnt two 27C010 eproms with the code from mame which you can see in the picture above and the board would now boot.

It would get as far as the ram / rom check screen and reported 3 failures. Rams 1H, 17H and 18H. 17H and 18H had already been removed from the board and socketed so these were easy to check and both were fine. The soldering on the sockets looked good too, after ruling out the ram at 1H and all other surrounding logic I was left with the custom chips.

On closer inspection two pins on the 051960 custom were bridged, I removed the bridge and reflowed the chip but this made no difference. I decided to replace this custom as it was the most likely candidate from what I had to go on. Luckily I already had a spare desoldered in my parts box marked 'ok'. I used the usual chip quik technique to remove the original chip and soldered on the replacement.


Time for the moment of truth.


All rams now pass, repair complete.

Sunday, 18 June 2017

Mr Do! pcb repair

This Mr Do! pcb wouldn't boot and just displayed a screen of garbage.


This was caused by a bad 74LS138 @A3. Once I replaced the 138 nothing changed, I checked reset pin with my logic probe and it wasn't kicking in. After about 15 seconds it finally kicked in and the game booted.

The capacitor that causes the delay is the cap @ C14 which was a 22uf, interestingly the schematics suggest a 47uf cap which would make the delay even longer.


I checked another Mr Do! pcb and that had a 0.47uf disc cap at C14, so I replaced the 22uf cap with a 0.47uf electrolytic as that's all I had on hand but it worked perfectly and reset kicks in right away now.

This just left one remaining issue, the player and enemy sprites were missing (I forgot to photograph this). I finally tracked it down to a bad output on 74LS32 @ E3.


Repair complete.

Sunday, 21 May 2017

TMNT pcb sound repair

This Teenage Mutant Ninja Turtles (TMNT) pcb was working but without sound, a common cause of no sound on a TMNT is a missing or bad 640khz resonator. The first one checked fine on the scope but the second one labeled X2 was dead. I replaced it and could see it working on the scope but this made no difference.


The Z80 and 6116 ram @ F16 both looked healthy and piggybacking the ram made no difference. I then decided to make use of the very nice high quality schematics available at jammarcade.net. The next most likely cause was the 74LS374 @ F21. This was confirmed bad with the chip tester and a new 374 brought all sound back except for the theme song.


The theme song is stored on a maskrom @ D5, I desoldered it at read it in my programmer as a 27C400 and it read as blank. I took another D5 maskrom off a scrap board and the theme tune returned.

Turtle power!

Wednesday, 17 May 2017

1084ST Commodore monitor repair

These Commodore monitors are great for the test bench (when they work) as they accept RGB video and have lots of nice adjust pots. This ST version of the 1084 seems to be the most obscure version with little information online and one of the only versions that doesn't have schematics available (typical).

When I got it, it was completely dead the led didn't even light up. With no schematics available I just poked around looking for any obvious problems and hoped for the best but found nothing. After testing all the capacitors, diodes, bridge rectifier and anything else I could test I couldn't find any issues. I even tried a replacement HR Diemen flyback but that made no difference, still dead.

I took a break from looking at it for a while but when I came back to it I spotted an 8 pin ic labeled TDA4605, a google for the data sheet reviled that it's control circuit ic for switch mode power supplies. I am sure I could have tested it in circuit with the scope but as I was running out of parts to check and a new one was only £2 including postage I decided just to order a new one and try it.

I fired it up and.....

 

Composite works too. :)