Sinclair ZX Spectrum +2 128K


The ZX Spectrum +2 was Amstrad's first Spectrum, coming shortly after their purchase of the Spectrum range and "Sinclair" brand in 1986.

Launched in early 1987, the machine featured an all-new grey case featuring a proper spring-loaded keyboard (at last), dual joystick ports, and a built-in cassette recorder dubbed the "Datacorder".  The Amstrad influence was obvious with its design similar to the CPC464, but was in most respects it was identical to the Sinclair ZX Spectrum 128.

The main menu screen lacked the Spectrum 128's "Tape Test" option, and the ROM was altered to account for a new 1986 Amstrad copyright message. These changes resulted in minor incompatibility problems with software that accessed ROM routines at certain addresses (this had been sorted to a certain degree by the time the Issue 2 boards were fitted in machines.

The new keyboard did not include the BASIC keyword markings that were found on earlier Spectrums, except for the keywords LOAD, CODE and RUN which were useful for loading software. This was not a major issue however, as the +2 boasted a menu system, almost identical to the ZX Spectrum 128, where one could switch between 48k BASIC programming with the keywords, and 128K BASIC programming in which all words (keywords and otherwise) must be typed out in full (although the keywords are still stored internally as one character each). Despite these changes, the layout remained identical to that of the 128.

Built in Taiwan, it became the first "Sinclair" product to be built outside the UK.  However, with Amstrads greater emphasis on quality control, it was far more reliable than any of the earlier generation Spectrums.
Production costs were lower and the retail price dropped to £149 soon after launch.

Technical Details

  • CPU: Z80A
  • Speed: 3.5MHz
  • ROM: 32KB (16K for 128K mode, 16K for 48K mode)
  • RAM: 128KB (8 x 16K banks)
  • Full-stroke 58 key Keyboard (same layout as Spectrum+)
  • Colour Graphics
  • Sound - 3 Channels, 7 Octaves
  • RGB, Keypad, RS232/Midi
  • Local price at launch: £200

Problems and Issues

Immediate problems often experienced with a Spectrum +2 which hasn't been used for a long time are poor program loading performance due to a dirty read head (easily cured with a little isopropyl alcohol on a cotton bud (or better, a chamois stick), or worn drive belt issues.

Over time without use, the main drive belt develops a pronounced and very visible (when you turn the pulley) kink where it has been parked stationary over the drive motor pulley wheel.  This causes poor loading performance as the belt "skips" each time the kink passes over the capstan..  Replacing the belt is a simple task and I always have them available in the Sinclair Sales pages of the Sinclair section of the website.  Instructions are available in the Repairs & Servicing section, including advice on how to check the azimuth (head alignment) and how to cure other cassette loading issues.

The small belt rarely gives problems even now, and doesn't seem to perish.  They are available in the Sinclair Sales section if you need a replacement though.

Another common problem is the main drive wheel catching on a small spring loaded pad positioned just below it.  This causes the speed to fluctuate and again adversely affect loading performance.  Again, more details on how to put this right in the Servicing & Repairs section.

My own experiences of trickier problems with the Spectrum +2 loading performance mostly tend to centre around tape deck issues with early models with Issue one main boards.  Loading was very much a hit and miss affair with some +2's and no amount of head cleaning, belt changing or tweaking the azimuth really helped.

Machines with Issue 2 boards (and onwards) do seem to be much better and I have swapped over cassette mechanisms from otherwise dead Issue 2's to Issue 1's with a lot of success.  This doesn't seem to make sense on the surface of it as the new mechanism is now in an Issue 1 board and so should be poor!  There was obviously some "invisible" upgrade made to the cassette decks (and their controllers) as well as the upgrade to the Issue 2 motherboard.

Another problem which I seem to be seeing more and more frequently is a complete failure to load a program and no visible loading border or loading sound.
This can be a faulty play head, but most often the problem is temporarily cured by sharply banging the spectrum sharply again the table it is sat on.  The loading border then appears and the usual sound can be heard.  Pressing down with your thumb on the case at the bottom left of the cassette flap during loading also often temporarily cures the problem, but standing there with your thumb aching for 10 minutes is hardy a cure...

The problem is due to a "dry" solder joint in one or more of the soldering lines on the cassette drive controller board - usually the joints for the red and white wires which connect the read and write heads. 

The standard of soldering on assembled items like the cassette on these machines was variable to say the least, and at the point where the leads from the play head meet the cassette drive controller board, the solder is a dull grey colour, rather than the shiny metallic colour it should be.  Poor solder technique using insufficient flux often causes this "oxidation" effect and over time, the joint breaks down and electrical contact is lost.

To do this, add a little flux with a fine children's paint brush to the grey solder blob, then hold the tip of a soldering iron against the poor joint.  The solder will melt fairly quickly - take away the iron and the solder will set again, usually repairing the joint.  Remake the joints for both the red and white leads as shown in the photograph.