Equites

Atari Asteroids repair log

Jan 282016

I was recently sent an Asteroids PCB which I had previously bought back from the dead, the PCB had developed distorted vectors but still running.

Here is what the TEST screen looked like, the actual diamond shape was perfect but text was oversized and distorted;

IMG_20160128_242023439

The actual GAME mode looked worse;

IMG_20160128_241959634

Check out the giant asteroid!

I knew the problem was in the Vector State Machine (VSM) area, specifically the area which is responsible for handling object vectors, so I looked there first.

It was not long until I found a decoder (LS42) at 7E which seemed to have some pins stuck LO;

IMG_20160128_242035901

I desoldered the decoder;

IMG_20160128_242531687

I tested the decoder in my IC tester;

IMG_20160128_242615090

I then socketed and replaced with a fresh decoder;

IMG_20160128_243027782

Asteroids now fully working;

IMG_20160128_243034333

IMG_20160128_243046130

Space Invaders (Taito) 3-Layer repair log #1

PCB Repair Logs Comments Off

Dec 102015

I was recently given a Taito 3-Layer Space Invaders boardset by a friend to look over. The problem was that it was not running, completely dead with a clicking sound coming from speaker (sounds like watchdog barking to me).

I powered up the PCB stack on the bench and checked the status of the 8080 CPU. I could see right away from probing pin 19 (SYNC) that the clock circuit was not running.

On the ROM board, I could see that the crystal was intact, so I checked the 74S04 (IC1) which I was sure was drives the clock next to the crystal. I found that most of the pins were floating. Replaced it and I now have a clock (note to myself, 74HC04 will not work), watchdog no longer barking but still no display.

ace_si_1

On further probing, found a bad 74161 counter at location IC6 with pins 11 to 15 stuck HI. Replaced this and I now have a display, but it appears to be doubled up;

ace_si_3

Eventually located the problem to IC28 which is a 74LS83, pins 2, 6 & 9 were stuck LO. Replacing this restored the display;

ace_si_2

Note, I am using the composite input on my Plasma TV, as the PCB is composite only and not RGB.

Space Invaders live to invade another day!

Rainbow Islands repair log #8

PCB Repair Logs 1 Response »

Dec 042015

Another Rainbow Islands PCB to repair.

This, like the previous repair (#7) showed a solid white screen, however, if left alone for about 30 seconds I would get the following displayed;

I found a bad object/sprite 2018 (6116) RAM at location IC50, so this was replaced with a fresh 2018 RAM.

The PCB was now working 100%, and put on the shelf. That was not the end of it though.

The next day, the PCB was dead again, just showing another solid white screen. Probing around PCB at the logic, eventually I came back to the object RAM and found that another had failed at location IC51. This was then replaced, and the PCB sprang back to life but now with issues. Colours were missing or wrong on the display.

You can just see from my blurry picture that the RAINBOW PROJECT screen is all wrong, no blue background and the colours are missing. Only the first two letters on each row have a colour (green and grey);

Here, interestingly, the 1UP and 2UP are white, should be green (1UP) and blue (2UP). Also, the HIGH SCORE and TAITO logo should be red. The Rainbow Islands logo looks intact, but the colours are solid and not rotating like they should;

The MAP screen has not escaped either, most of the colours are gone;

The actual game has suffered too. It’s hard to tell, but everything is there and perfectly formed, just missing colour so either appears corrupted or simply missing;

Not even the TEST CARD escaped;

I have already ruled out the TC0070 RGB SIP module as being the cause as the colours are there, just not in the right places.

While I was probing around the board, things got worse, now the sprites have broken up;

Eventually, the PCB stopped running again and I got this message on the display;

Another failed object/sprite 2018 (6116) RAM, this time at location IC50.

So this was replaced and I now got this display;

At first glance I thought it was still the same message, OBJECT RAM ERROR, but then noticed this time it was COLOR RAM ERROR.

So, I replaced the bad colour 2018 (6116) RAM at location IC70 and the PCB was running again, with the sprites now intact again.

At this stage, I was becoming a bit frustrated with the failing RAM, so I bit the bullet and just replaced them all so I can concentrate on the colour problem;

It was a good move because now at least the PCB was running stable.

I was starting to get a bit fed up looking for the colour problem, and Caius very kindly offered me some help. Infact, he was a massive help and told me exactly where to look for the problem.

With help from Caius, eventually I found that pin 15 of the 74LS157 at IC36 was floating (using my logic probe). This is the IC’s enable line, and it should have been tied to VCC (through the orange disc 104 capacitor).

It was then I notice the burned/damaged trace going to the orange disc 104 capacitor;

Very odd place to find a damaged trace, certainly not very accessible. I had no idea how it was damaged until Caius mentioned it has happened to him also. Like Caius said, I must have accidently fried the trace by short-circuiting it during my in-circuit testing.

I do recall the PC PSU for my Supergun shutting down due to a short when I was probing the PCB.

So, using kynar wire I patched the damaged trace from underneath;

The game’s display was now 100% restored;

Massive thanks to Caius, whom without I would not have solved this and just given up.

Rainbow Islands repair log #7

PCB Repair Logs 2 Responses »

Dec 032015

I was fortunate enough to obtain an original Taito Rainbow Islands PCB recently, which of course was not working.

I tested the PCB in my Supergun and was presented a solid white screen;

Checking the 68K CPU, the clock was fine, and RESET line was held high, there was activity around the CPU which indicated it was trying to execute code.

The work RAM looked good, so I checked the two 62256 SRAMs that the PC080SN custom uses for all backgrounds/tiles. The outputs looked bad on both. I also noticed that when I touch the pins on the SRAMs the screen display changed.

I removed the SRAMs (IC3 & IC4) and confirmed they were bad in a RAM tester.

I installed sockets for easy maintenance and replaced with fresh SRAMs;

The PCB was now running, but had some issues with background/tiling;

Speaking to my friend Porchy about this fault, he stated that it appeared that Bit 4 (counting from 0) was stuck LO, and that the PC080SN custom was responsible for it (RAM lies at address 0xC00000 – 0xC0FFFF).

I was hopeful it was not a fault within the custom, and eventually after probing found that IC12 (74LS373 Octal Latch) Pin 9 was stuck LO (should have been pulsing LO/HI).

I removed the IC, installed socket with fresh 74LS373 and was greeted with a fully restored display;

I did notice however that sometimes the display would lose all the BLUE. So my attention turned to the TC0070 SIP custom that handles RGB. I found a cracked pin (highlighted with Red arrow), which I resoldered to resolve it;

It was now running perfectly, except for the sound which appeared to be absent totally.

I checked the Z80 CPU which handles sound, and it was definitely running code. So I checked the TL074 op-amp at location IC67 and found it appeared dead (no output, but probing input I could hear the sound!). So I removed it, installed socket with fresh TL074 to fully restore the sound;

Many thanks for Porchy for his help with this repair.

Another Rainbow Islands Log to come very shortly, solved with help from my new friend Caius.

Newer Entries