Sinclair ZX Spectrum+ Troubleshooting Guide

This guide gives fault diagnosis for the Sinclair ZX Spectrum+. Because the Spectrum+ is electrically a 48K Spectrum in a different case, everything here applies equally to the late rubber-key 48K machine; the Spectrum+-specific material is the keyboard assembly and the reset switch.

It draws on the official ZX Spectrum Service Manual — including its "Repair Data" field-service section — and on the repair knowledge built up by the Spectrum community over four decades.

⚠️ Before you start

• There are no mains voltages inside the computer. The Spectrum+ runs from an external 9 V supply; the highest voltage on the board is +12 V. The mains transformer is inside the external PSU brick.
• Never plug or unplug anything into the expansion edge connector with the machine switched on. This is not a general caution — it is reported as the primary cause of 80–90 % of Spectrum faults. The keyway locating peg on the interface plug drops out with use, allowing the connector to move sideways and short the bus lines together. Because most of the ICs hang off that bus, the damage can be extensive.
• If the keyway peg is missing from an interface, make a replacement: cut the hinge section from an old cassette box, shape a peg from it and glue it in place.
• Hairline solder splatter can measure several ohms and produce very misleading symptoms — inspect visually before chasing anything complicated.

Start here: the physical check

Sinclair's own Repair Data section opens by noting that around half of machines returned as faulty had no real fault — the problems were loose or badly fitted parts. Before any measurement:

1. Check the heatsink first on any reported fault. Is it loose? Are the screws over-tightened?
2. Look for dry joints, cracks in the print, cracked boards, badly fitted ICs, and dirty or tarnished contacts, pins, holders and edge connectors.
3. Check for damaged keyboard tails.
4. Inspect any add-ons and interfaces for broken solder caused by the user wiggling them, and note whether an add-on is itself loose or faulty.
5. Look for hairline cracks and breaks in the board, especially radiating from the heatsink screws.

Knowing what add-ons the machine was used with is a diagnostic clue in itself: the manual specifically says to suspect a blown TR4 and/or destroyed RAM if a Kempston interface was used.

Supply checks

Basic checks, in order of priority (Service Manual)

Function	Test point	Expected
Voltage regulator input	+ve side of C50	+9 V DC ± 2.0 V. Below +7 V the regulator may not operate correctly
Voltage regulator output	+ve side of C34	+5 V DC ± 0.25 V, no discernible ripple
On-board +5 V	IC6 pin 9	+5 V DC ± 0.25 V, no discernible ripple
On-board −5 V	IC6 pin 1	−5 V DC
On-board +12 V	IC6 pin 8	+12 V DC
+12 VA (to IC14 only)	Left end of C52	+12 V DC
Clock	IC1 pin 32, TR3 base, IC2 pin 6	14 MHz square wave at +5 V amplitude
Address and data lines	Right side of R17–R23 and R1–R8	3.5 V amplitude

Source: Service Manual. If these pass, check IC1 pin 14, IC2 pin 11 and IC5 pin 28 for +5 V, and ground at IC1 pin 40, IC2 pin 29 and IC5 pin 14. If all is well and the ULA is socketed, replace IC1.

The community multimeter-only sequence

Ian Worsley's field guide gives a diagnostic path needing only a multimeter on the 20 V DC and 20 kΩ ranges, with the negative lead on the modulator case:

1. Power up briefly and measure the 7805 output (right-hand leg, tab away from you). 4.85–5.2 V is normal.
   • Higher than that — change the regulator, and hope not too many chips have been taken out.
   • Lower, but not below 4.3 V — there is a short somewhere. Feel the chips: the ULA runs quite warm, everything else should run cold.
   • Missing altogether — replace the regulator, but do not run the machine yet.
2. Switch to 20 kΩ. Measure pin 1 of IC6: it should read above 12 kΩ. If lower, nick the track to the right of C4 (above IC6).
3. Measure pin 8 of IC6: if below 700 Ω, nick the track to the left of C44 (100 µF 16 V, bottom right of IC6).
4. Leave the tracks cut, power up, and measure on 20 V DC to the right of each cut: the first should read −4.5 to −5.2 V, the second +11.5 to +12.3 V.

This isolates the DC-DC converter from the RAM it feeds, which is the crux of most Spectrum faults.

The DC-DC converter — the classic Spectrum fault

The 4116 lower RAM needs −5 V and +12 V, generated on-board by an inverter around TR4 (ZTX650) and TR5 (ZTX213). When it fails, the lower RAM has no supply and the machine shows a white or patterned screen.

If −5 V or +12 V is missing or low:

• Replace TR4 (ZTX650) and TR5 (ZTX213). Do not substitute other transistor types here — they will not work. (A ZTX651 is an acceptable TR4 substitute, being the same part with a higher voltage rating.)
• Check R79 (2k2, Issue 6), R62 (15 Ω, Issue 6) or R55 (56 Ω, Issue 4 and earlier) for discolouration. If a 56 Ω resistor is badly discoloured, change D15 (BA157).
• Check the zener and signal diodes: on Issue 6/6A these are D19 (BZY88 5.1 V), D18 and D11 (1N4148); on Issue 3 and 2, D16 (BZY88 5.1 V), D11 and D12.
• There is roughly a 20 % chance the coil has shorted turns. With the converter working you should be able to hear the coil buzzing.

The Service Manual's Repair Data adds an important warning: TR4, TR5 and the coil often fail together. If you replace only TR4 it can fail again immediately because the others are still faulty — check all three.

No coil buzz and the 7805 open circuit — before replacing the regulator, check the heatsink for damage or bad fitting, remedy any physical defect in the print or damage around the screws, and check TR4/TR5 and the coil.

Continual failure of power supply transistors — verify all modifications are present and the edge connector is sound and not loose or damaged; check the peripherals are properly fitted and establish whether the owner is hot-plugging add-ons. Check the plug, socket and cable, and check the mains for interference from other equipment. If all that is satisfied, confirm 0 V is correct on earth and monitor it while running to catch an intermittent, then work systematically through the capacitors and resistors to earth.

Reading the screen

The failure pattern on screen is the fastest route to the fault:

Faulty screen designs

What you see	Likely cause and action
Alternate thick black and white vertical bars, with or without flashing	Check the CPU (IC2) is seated correctly and any cut legs resoldered; check the ROM (IC5). Test or change TR6 and TR3 (both ZTX313). Then re-measure legs 2–14 of IC6–IC13
Bright, shimmering, fine herring-bone pattern filling the screen	Check for −5 V on pin 1 of IC6 — it may have vanished. Change TR4
Black screen with one or more thin vertical dashed lines	Almost certainly one or more 4116 lower RAM chips. Also check the ULA for straddling, bent or damaged legs
Blank black screen, nothing else	Likely one of the upper RAM chips IC15–IC22
Flashing blocks, or a mixture of the above	Check both legs of any upper-memory chip that was cut are resoldered; check the ROM is the correct type for the fitted links and correctly seated
Character break-up when the machine warms up	Nearly always the ULA (IC1)
Black dashes or dots appear when warm, then the machine locks up	Typical of the lower RAM (IC6–IC13), or the ULA. Do a resistance check on pin 1 while the machine is still warm

The Service Manual adds that various faults which appear only when warm should start with a thorough heatsink check; if the heatsink is sound, the ULA is the cause, and if the ULA has been replaced before it may be worth adding extra heatsinking to it.

Finding a faulty RAM chip

If the machine boots but a RAM fault is suspected, the Spectrum can name the failed chip itself.

Type:

PRINT PEEK 23732 + PEEK 23733 * 256

A healthy machine returns 65535 (48K) or 32767 (16K). Any lower value n means the first faulty location is n + 1. A result below 32767 puts the fault in the lower 16 K (IC6–IC13); above 32767, in the upper 32 K (IC15–IC22).

Then, using that address:

POKE n+1,85 : PRINT PEEK n+1
POKE n+1,170 : PRINT PEEK n+1

85 and 170 are 01010101 and 10101010, so between them they exercise every bit. If either value reads back wrong, look it up:

RAM fault location table

Reads back after POKE 85	Reads back after POKE 170	Error size	Bit	Lower RAM (<32767)	Upper RAM (>32767)
84	171	1	0	IC6	IC15
87	168	2	1	IC7	IC16
81	174	4	2	IC8	IC17
93	162	8	3	IC9	IC18
69	186	16	4	IC10	IC19
117	138	32	5	IC11	IC20
21	234	64	6	IC12	IC21
213	42	128	7	IC13	IC22

Values from the Service Manual. If more than one location is faulty, the first must be repaired before the test can proceed past it.

If the value read back is not in the table at all, suspect bent or straddling legs on socketed upper RAM, or faults in IC23 (74LS32), IC24 (74LS00), IC25 or IC26 (74LS157).

Modern alternative. Purpose-built diagnostic ROMs make this much faster: Brendan Alford's ZX Diagnostics test ROM will run from the ROM socket and identify bad bits even when the display RAM itself is faulty, reporting the failing data lines and IC numbers.

Checking for shorted or leaky RAM

• After the supply rails are proven, feel each lower RAM chip. None should be even warm — any that is should be removed.
• The manual's method is to switch on and check whether any RAM gets hot; a very hot chip is a sure sign it is short circuit. For open-circuit failures, bypass each RAM in turn with a known-good device — and note that after switching off you should wait about ten seconds before removing the test chip, or it will damage itself.

Keyboard faults

The matrix is eight blocks of five keys horizontally and five blocks of eight vertically. A dead block of five points to the KB2 circuitry or the 8-way membrane tail; a dead block of eight points to KB1 or the 5-way tail.

• Entire keyboard dead — check the 5 V supply is reaching it before condemning the ULA. The commonest cause is a faulty 10 kΩ resistor in the R65–R69 group, identified by which keyboard column has failed.
• Keys fail when the machine is warm, with a known-good ULA — replace the membrane.
• Most keyboard faults are simply membrane failure: the pads under the keys wear out, or the ribbon cracks with age.

Spectrum+ specific keyboard fixes

Two faults belong to the Spectrum+ keyboard assembly in particular:

• Some keys not working, especially on the bottom row. Both ribbon cables are clamped to the keyboard case. Remove the clamps and carefully clean between the layers of each ribbon — around the clamped area — using a cotton bud and WD-40. When re-clamping, add extra packing between the ribbon and the clamp. This often avoids a membrane replacement entirely.
• A character printing repeatedly. This is caused by the keyboard membrane backing plate being clamped too tightly. Remove the keyboard top to expose the plate and its nine screws, then slacken by a quarter turn the screw or screws nearest the offending key.

When handling the tails, insert them with a slight side-to-side movement while easing the ribbon gently down into the connector. If the ends are crinkled, creased or cracked, you can trim back up to about an inch — but strip the insulated part an extra quarter inch or the keyboard will stop working entirely.

Loading faults

Symptom	Action
Loading bars appear but no program name	Almost always the tape deck. Clean the head and adjust the azimuth — with a tape playing, turn the azimuth screw for the sharpest possible note. If the note sounds at all muffled the Spectrum will not load from it
Program name shows as ????????	Volume set too high — reduce it
Intermittent loading	Check the tension of the EAR/MIC socket contacts; they can be re-tensioned by pushing the spring contact down and inward rather than replacing the socket
No loading bars at all, tape deck known good	Change the ULA (IC1) first, then check the MIC socket, then D9 (1N4148) behind the MIC socket

Sources: Service Manual Repair Data and the community repair guide. The manual notes that a recorder can play music perfectly and still be misaligned for the Spectrum, and that it is better to set volume for maximum treble than to use the midpoint.

Sound faults

Loads correctly but no sound:

1. Check the ULA (IC1) first — faulty or missing sound most often identifies the ULA, which is the likeliest suspect in any case.
2. Enter 10 BEEP 1,1 : GOTO 10 and RUN it, then disconnect the keyboard and hold a replacement speaker across the existing one. If you hear a tone, replace the speaker.
3. On Issue 3 and later, change TR7 (ZTX450, ZTX650/1 also acceptable) and D9 (1N4148), found to the left of the speaker.

Video and colour faults

• No colour. On Issue 2, adjust TC2, VR1 and VR2 first. On other issues, change the LM1889 (IC14) — though it is a very reliable part. Some later boards were fitted with an SN-series device in the IC14 position, and this was always the cause of no colour; it should be replaced with an LM1889, which the board still has provision to accept despite the different pin count.
• No video input at IC14 pin 12 — check the waveform at IC14 pin 13; if correct, check TR1 and TR2. Check the waveforms at IC1 pins 15, 16 and 17; if correct, check VR1 and VR2 on Issue 1 and 2 boards. Then change IC1. If +5 V and video input are both correct, change the modulator.
• Dark screen — check tuning, then change the modulator.
• Dim Sinclair logo — check +12 V, check C65 for high impedance, check R52, and check +5 V on IC14 pin 3.
• Corrupt paper after a ULA change — check the ULA socket, check the power socket is not intermittent, and check IC1 pin 14 for +5 V with no discernible ripple.
• No 3 MHz at IC2 pin 6 — if the amplitude is low check R25; check TR3; check IC1 pin 32 for 3 MHz and IC1 pin 39 for 14 MHz; check crystal X1; check IC1 pin 13 for +3 V.
• Wavy lines across the screen — press down on each electrolytic capacitor in turn and see whether the lines disappear; if so, that capacitor is the culprit.

Power supply faults

The external PSU is a common failure point, and it is repairable:

• The usual fault is the wire snapping or going intermittent in the moulded plug or where the lead leaves the case.
• To fix the plug end, fit a replacement 2.1 mm plug with negative to the centre (the traced wire), then fill the plastic cover with silicone rubber before screwing it up — this prevents internal shorts and takes strain off the solder joints.
• At the PSU end, split the case, cut off the faulty section, tie a double knot about three inches down the cable, slide protective sleeving on, and resolder. The knot inside the case stops the joints being strained.
• A bare wire shorting the output blows the thermal fuse and/or takes out the 1N4007 rectifiers. The thermal fusible link can be replaced by spot-soldering a 160 mA slow-blow 20 mm fuse across the old device. A faulty transformer, however, means the unit is scrap.

Intermittent crashes or resets, especially when the machine is moved, are usually the power plug or the power socket. Replace the socket — do not try to tighten it.

Intermittent and warm-only faults

On a comparatively young machine, intermittent faults are most likely poor contacts or soldering from manufacture. Check all holders and pins for tarnish and dry joints; replace faulty holders by cutting them out and fitting new ones. Look for parts that were poorly soldered or not soldered at all. Finally check the board for hairline cracks and breaks, especially radiating from the heatsink screws.

If the power supply is fine but the machine still misbehaves

Work through single then multiple RAM failure as above. If the RAM is sound, or the fault is too severe to test RAM, check that +5 V is reaching all the ICs in the order IC1, IC2, IC24, IC23, IC3, IC4, IC26, IC25 and IC5, and replace whatever open-circuit component is cutting the supply.

Common fault catalogue

• Machine dead after using an interface — edge-connector damage from hot-plugging; check TR4 first, then the RAM.
• White or patterned screen — DC-DC converter (TR4/TR5/coil) or lower RAM.
• Herring-bone pattern — −5 V missing; change TR4.
• Thin vertical dashes on black — lower RAM, or ULA legs.
• Blank black screen — upper RAM.
• Character break-up when warm — ULA.
• Won't load, bars but no filename — tape azimuth.
• No sound — ULA, then speaker, then TR7/D9.
• No colour — LM1889, or an SN-series substitute in IC14.
• Dead keyboard — 5 V supply, then the R65–R69 10 kΩ resistors, then the membrane.
• Bottom-row keys dead (Spectrum+) — clean and re-pack the clamped ribbons.
• A key repeats endlessly (Spectrum+) — slacken the backing-plate screws nearest that key.
• Random resets when moved — power plug or socket.

Related pages

• Sinclair ZX Spectrum+
• Sinclair ZX Spectrum+ General Maintenance
• Sinclair ZX Spectrum+ Capacitor Replacement Guide
• Sinclair ZX Spectrum Troubleshooting Guide — the wider Spectrum series guide
• Sinclair ZX Spectrum Keyboard Membrane Repair
• Sinclair ZX Spectrum +2 Troubleshooting Guide