It’s well known that the Investronica 128 (otherwise known as the Spanish Toastrack) and the UK 128 machines have different RGB pinouts, meaning that a standard UK SCART lead won’t work on such a machine.
Personally, however, I prefer composite outputs from my Sinclair machines – RGB looking too ‘perfect’ for me. Part of the nostalgia for me is the imperfect video output, and while I’m not such a masochist as to insist on RF connection, composite in my mind is a good compromise between quality and the aforementioned nostalgia.
It’s quite easy to get composite output from a standard UK 128 – a composite signal is available on pin 1, and using pin 2 as ground, a simple 8-way DIN to phono socket adapter allows the use of the same composite cable setup as 48K machines use.
On the Spanish 128, the composite signal is on pin 3 instead – but to my bemusement, when I wired up an adapter to use this pin, I found that this was in black and white.
A quick look at the schematics for the analogue circuitry of the UK and Spanish 128’s reveals the culprit. The signal on the Spanish machine is actually derived from pin 7 of the TEA2000 RGB to PAL encoder (LUMO, Luminance Output) instead of pin 6 (COVO, Composite Video Output) on the 128 machines.
On the UK machines, the signal derived from COVO passes through a variable coil L5 (this filters the chrominance signal, although it’s not perfect), and then goes to diodes D34/D33/D31 and then TR14 for buffering and amplification before going to pin 1 via a 75 ohm resistor.
On the Spanish machine, the luminance signal from LUMO is simply buffered by TR13 and then channeled through D34 etc. in the same manner as the UK models.
So to fix this, and obtain proper composite output on your Investronica 128, the following modifications need to be made:
- Remove TR13 (this is not present on the UK 128’s), or alternatively disconnect its base (middle leg)
- Connect the right (anode end) side of D30 to the anode of D34, or the emitter pad of TR13
- Lift the right side of C126 and connect it to the north side of R130
- Install LC filter comprising of a 82pF capacitor and a 15uH inductor between pins 8 and 17 of the TEA200 as per this article. (If this isn’t done, you’ll have horrible smeary colour on your composite output, because there is no L5 coil to act as a chrominance filter.)
See below for photographs of the modifications, which are a bit rough and ready but work perfectly.
As a footnote to this article, one of the known UK 128 prototypes is based on a Version 2 PCB which is more usually the PCB revision used for Investronica machines, and has a hand wired modification in the area of TR13. It’s assumed that this was the Sinclair engineers prototyping the modifications described above in a different manner.
Professional Code Monkey, part-time ZX Spectrum tickler.