ZX Spectrum +2 Video Ghosting Fix

Another vintage computer to take a look at, this is a 128K +2 which was made by Amstrad. They made a few mistakes on this model including installing transistors upside down (issue 3 only), mixing the audio with the composite video and a ghosting problem with certain colours again with composite.

You can avoid the ghosting / poor image problem simply by using a RGB scart cable but composite cables are far easier and faster to make. :)

How to separate the audio from the composite video has been covered many times and is covered in this very handy pdf ZX Spectrum 128K video fixes and video cables.pdf so I won't go over it again.

The ghosting fix on the other hand is quite a recent discovery by JoulesperCoulomb on youtube although I found out about it from GadgetUK164's video again on youtube.

 

Basically all you need is a 15uH inductor and two ceramic caps (20pf and 68pf). Connect the inductor to pin 8 of the video encoder join the two caps in parallel and connect one end to the inductor and the other to pin 17, no more ghosting.

Commodore 64 (Breadbin) repair

A rather sorry looking C64 early breadbin model, I've had this for a while and finally decided to have a look at it. First thing I had to do was replace the missing ram at U21 with a socket and 4164. After powering on there was garbage on screen, so still a bad ram or two to find.

This is where my previously made test cartridges came in handy (dead test and diagnostic), first I tried the diagnostic cartridge but the text on screen was too messed up to show anything useful. Next I used the dead test cart and the screen flashed five times, the manual states five flashes is a bad ram @ u10. So I removed u10 and replaced it with a socket and new 4164 ram, this fixed the garbage screen and the familiar blue screen was now visible but it isn't quite sorted yet.

There was a scrolling rainbow colour effect across the whole screen, doing a quick google for 'color' ram shows that the 2114 @ u6 is the colour ram. I removed and replaced this again with a socket and new ram and the colour problem is fixed. I thought this was the end of the repair but after letting the diagnostic cartridge run fully (it was getting stuck at color ram before) it showed that char rom was bad. This rom has the same pinout as a TMS2532, luckily I had one and the GQ-3X was willing to program it.

 

The eagle eyed will notice who ever looked at this previously marked all the working chips with a yellow dot. The character rom didn't have this mark, so who ever it was knew it was bad but obviously didn't have a TMS2532. :)