The Issue 5 Spectrum motherboard has always been shrouded in mystery. For a long time nobody really knew for certain if it even existed and the only mentions of it appeared in the Spectrum+ Service Manual along with schematics and a board layout diagram. According to the information in the manual there were only 1,000 units made.
Earlier on this year, a photo of an issue 5 motherboard surfaced proving that the boards had actually been produced and presumably distributed to the public. Unfortunately no-one has been able to find another trace of the issue 5 motherboard….
There are several changes that were made between the design of the board and when the board was actually populated with components.
The biggest give-away that this board is an Issue 5 is the extra marking for a capacitor C47 next to the electrolytic capacitor C79. Although the board says “Issue 5” on it, from a distance this could easily be confused for a different issue. However, no other issue board has these two capacitors next to each other.
Both of these capacitors are for the -5V line for the lower RAM.
C79 is a 1uF capacitor which sits on what will become the -5V line, but is isolated from that line by resistor R79. On the other side of R79 the zener diode D19 creates the -5V line – the extra capacitor on the board, C47, sits in parallel with the zener diode. This would create extra filtering for the -5V line, which presumably Sinclair felt they needed at the time. However, C47 doesn’t appear on the schematic, and nor does it appear on the Issue 4B or the Issue 6A – yet it does appear on the Issue 4A. It appears that Sinclair were transitioning with Issue 5 by moving the filter capacitor for -5V line – originally it was connected to the RAM chips and isolated from the coil (when it was C47) and then after the Issue 5 it was connected to the coil side of the isolating resistor.
This strongly suggests that the Issue 4B was designed after the Issue 5.
Here we see C105 added (a 180pF ceramic capacitor). This was obviously not planned as it doesn’t appear on the Issue 5 schematic, but appears on the Issue 6A schematic. It appears Sinclair realised that C47 and C79 were both useful, but they only needed one of them – so C47 was removed (presumably being less important) and C105 was added in its place. It connects from ground to pin 15 (the /CAS line) of the upper RAM chip next to it.
Another component Sinclair didn’t realise they needed was an 8K2 pull-up resistor on the MREQDL line on the ZX8401 chip. Apparently they were, however, pessimistic on one thing because they put space for C102 and C103, which according to the schematic are both capacitors connected to ground from the DRASLL and MREQDL lines respectively. These capacitors weren’t fitted, so the new 8K2 resistor (R89) was added using one of the
solder points designed for C103.
For some reason, when the layout of the board was done, the designer gave space for an electrolytic capacitor. The ceramic capacitor was ample for
the 22nF value that this capacitor has.
The most obvious change is the wire toward the right side of the board. This is a board correction!
This is the real reason the Issue 5 was never produced in numbers. The yellow wire is a correction for the board.
The highlighted area above shows an incorrect connection had to be broken.
The area below is the connection from the base of the transistor TR8 – unfortunately, instead of making a connection to the rest of the display circuit on the board, it connects to the D6 line – this is a data line for the CPU! What this means is that the ULA would create a display and completely disrupt the data that the CPU needs. Also, of course, when the CPU did anything at all it would destroy the video signal. Connecting this up probably
wouldn’t damage anything, but there is no way at all this would make a working Spectrum.
This is what the circuit should look like. The broken track from the base of TR8 should connect to C70 and R75 – which is what the wire does. The two components C70 and R75 are connected together here, which is why the track should go to the upper of the two contacts, but the wire connects to the bottom (it could just have easily connected to the top contact).
Below are the board layout and the schematic from the Spectrum+ Service Manual.