Saturday 22 October 2016

Space Invaders PCB Repair

I've had this Space Invaders board a while along with another one but kept putting off looking at it. Mainly because my 8080 fluke pod has never worked right and even when I thought I'd fixed it and had it passing self test it would still fail to do anything outside of self test.

Anyway lets see how far we can get without the fluke.



This is basically the screen you get when the reset circuit isn't working on the power supply but as I am using a jamma adapter I've tied the POR (power on reset) to ground via a resistor and it still didn't make it past this screen even if I pulled the reset pin high.

I already knew the cpu and eproms were good from when I had a working board so I moved onto the logic as even if any rams are bad it should still boot. The HP comparator flagged a 7486 @ A4 saying output pin 8 was bad. I checked it with the logic probe and it was stuck high, shorting it to the input pin next to it and the board booted.

I replaced the 7486 with a 74LS86 which worked fine and it booted every time. The problem now is the screen is a complete mess so we have at least one bad ram. I tried the two different test roms but couldn't work out what either one was reporting as bad.

So I did the old trick of removing all the roms and booting to get lines on the screen then shorting pin 7 of each ram to ground until things look better.

I ended up replacing two rams and using some turn pin sockets off of a scrap bootleg board.

It's alive!

There was just two problems that remained the missile shot sound was very high pitched and the launcher exploding sound when you die was missing completely.

This very helpful page suggested a bad 4006 but it was actually the 4030 chip next to it @ P5.

The motherboard is a L000 revision that I couldn't find anything online about so I had to figure out how to strap it for 2716s myself (Intel type not TMS). It's basically the same how you strap other revisions but S1 (A9) is not connected by default and there's a few more pads on this one.



Repair complete.

Tuesday 4 October 2016

ZX Spectrum +2 Issue 3

Another revisit, I've covered the ghosting mod already and mentioned the mod to separate the audio from the video. I never snapped a picture before though and doing the mod again today I couldn't remember where capacitor c31 was.

So here's a blurry picture for my future reference or anyone else that prefers to see a picture of the mod rather than schematics with ms paint edits. :)


Transistor TR4 on the right of the RF box needs to be turned around on issue 3 motherboards too which this was.

Hyper Olympic / Track and Field pcb revisited

This is the Hyper Olympic board I've mentioned in previous posts, I've already replaced most of the sockets and almost all the eproms as all but four were missing when I got it. It ran the test rom but with colour problems so I knew it was very close to working. The reason I kept putting this one off is I had no way of testing all the custom chips.


This is where I left off and sure enough still the same.

First to fix the colour issue.


This was caused by a bad fujitsu ttl on the bottom pcb (74LS157). You'll also notice I've setup my TV properly since last time as it's 4:3 and the AV1 doesn't stay on the screen now. :)

The test rom always flagged the NVRAM as bad, I just assumed this was due to the missing battery. I removed the battery when I first got the board as it had started to leak. I also didn't think a problem with the NVRAM would stop the game from running but might as well sort it out now.

I fitted a battery just to see if the NVRAM test would pass but it still failed hmm. On closer inspection the tiny blob of battery acid had taken out a trace and worked it's way into pin 24 of the NVRAM (6116).

It's not often you see the markings scraped off ram ics.


 A tiny bit of kynar wire and some solder mask to cover it up.

or should I say solder mast.

I tried the board without the replacement 6116 in it's socket just to see if I still had a screen of zeroes and sure enough it wasn't just the trace. Very strange that konami would make it so the game wouldn't run if the nvram is bad since it's only used for scores.




Repair complete.

Sunday 2 October 2016

Phoenix pcb pinout (button 2 / shield)

I've noticed people mention Phoenix boards that don't have button 2 / shield. While this is possible I think it's unlikely and more likely the Jamma adapter isn't wired correctly.


This board which I assume is a german boot (spiele) uses the exact pinout found every where online apart from the shield buttons. This board has P1 button 2 on pin H and P2 button 2 on pin 7.

Anyway I thought I'd make a note of this for anyone doing a google that has a Phoenix board with non working shield.

Legendary Wings (Aresu no Tsubasa) bootleg repair

At first I thought the colour green was missing but on closer inspection all colours were present, the background is obviously a mess though.



First thing I checked was some rams on the top board (2114s near the end of the board), shorting two data pins together made the background change colour so I was in the right area. The one closest to the edge was a different manufacturer to the other two and also improved the picture the most when having it's data pins shorted.

I removed it and fitted a socket / new 2114.




This fixed the colour / graphics issue, now on to the sound.

The sound was working but was noisy and distorted. I checked and replaced some caps these two were the worst.


10uf cap so not far off but look at that esr!

 After replacing the bad caps the sound was back to normal, repair complete.