Jump to content

Atari STE Troubleshooting Guide

From RetroTechCollection
Revision as of 20:30, 27 March 2026 by Josh (talk | contribs) (Created page with "<templatestyles src="Template:StyledTable/styles.css" /> The '''Atari STE Troubleshooting Guide''' covers systematic fault diagnosis for the '''Atari 520STE''' and '''Atari 1040STE'''. Both models share the same motherboard, custom chipset, and PSU design. The only hardware difference is factory RAM: 520STE ships with 512 KB (2× 256 KB SIMMs), the 1040STE with 1 MB (4× 256 KB SIMMs). All troubleshooting procedures apply to both models. For routine preventive...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

The Atari STE Troubleshooting Guide covers systematic fault diagnosis for the Atari 520STE and Atari 1040STE. Both models share the same motherboard, custom chipset, and PSU design. The only hardware difference is factory RAM: 520STE ships with 512 KB (2× 256 KB SIMMs), the 1040STE with 1 MB (4× 256 KB SIMMs). All troubleshooting procedures apply to both models.

For routine preventive maintenance, see Atari STE General Maintenance. For capacitor replacement details, see Atari STE Capacitor Replacement Guide.

Safety

  • Disconnect mains power before opening the case. Wait at least 5 minutes for PSU capacitors to discharge.
  • Never operate the PSU unloaded. If bench-testing, connect dummy loads: 2.2–4.7 Ω 10 W on +5 V, 22 Ω 10 W on +12 V.
  • Use an ESD wrist strap when handling the motherboard.
  • PSU primary-side capacitors carry up to 400 V. Exercise extreme caution.

PSU Voltage Rails

The STE internal PSU provides three DC rails. Measure at the motherboard power connector pins with a multimeter (PSU connected to the motherboard and powered on).

Rail Nominal Acceptable Range Measured At
+5 V 5.00 V 4.75–5.25 V Motherboard power connector +5 V pins; also measurable at pin 7 of any DE-9 joystick port, or across bulk filter cap C103 (4700 µF)
+12 V 12.00 V 11.40–12.60 V Motherboard power connector +12 V pin; also at pin 7 of the 14-pin floppy DIN connector
−12 V −12.00 V −10.80 to −13.20 V Motherboard power connector −12 V pin; also at the RS-232 line driver IC (MC1489 pin 14)

PSU ripple measurement: Use an oscilloscope set to AC coupling on the +5 V rail. Acceptable ripple is under 200 mV peak-to-peak after a proper PSU recap. Stock aged PSUs commonly exhibit 500 mV to 1 V+ of ripple, which causes random crashes, video noise, and audio hum. Ripple above 300 mV warrants immediate capacitor replacement.[1]

PSU Troubleshooting

Symptom Possible Cause Recommended Action
PSU completely dead (no output) Blown mains fuse; failed switching transistor (BU508/2SC2979); open primary winding Check mains fuse. Test switching transistor for short between collector and emitter. Check transformer continuity.
+5 V rail low (< 4.7 V) Degraded output capacitors (C201–C205 depending on PSU model); heavy current draw from fault Recap PSU. If rail remains low after recap, check for short circuits on the motherboard +5 V plane.
+5 V rail high (> 5.5 V) Failed voltage feedback loop; damaged opto-coupler; incorrect trim pot setting Check opto-coupler. Adjust trim pot. If uncontrollable, replace the PSU.
Video dims during floppy access PSU peak current demand insufficient; degraded capacitors Recap PSU with low-ESR capacitors. On Mitsumi SR98, modify R201/R202 bias resistors per exxos fix (R202→1K, R201→1.2K).[2]
Audible ticking from PSU Normal switching noise under load (some PSU models, e.g. ASP34); can also indicate marginal regulation If regulation is within spec, ticking is normal for some designs. If accompanied by instability, recap.
Burning smell from PSU Component failure (capacitor, transistor, rectifier) Disconnect immediately. Inspect for burnt components. Replace failed parts. Consider full recap.

Boot Sequence Faults

The STE boot sequence proceeds as follows: PSU powers on → CPU reset released → TOS ROM executes from $FC0000 → RAM test → hardware initialization → GEM desktop loads. Problems at each stage have different symptoms.

Symptom Possible Cause Recommended Action
Completely dead (no video, no drive activity, power LED off) No PSU output; blown fuse; disconnected power connector Verify mains power reaching PSU. Check PSU output voltages. Check power connector is seated on motherboard.
Power LED on but completely dead (no video, no drive seek) CPU not running; /HALT or /RESET held low; failed GST MCU; failed CPU; no clock Check reset circuit: pin 4 of 7406N should show clean reset pulse (~200 ms low then high). Check 8 MHz clock at CPU pin 15. If clock is present but CPU does not run, suspect CPU or GST MCU.
Power LED on, floppy seeks, no video GST Shifter failure; video output circuit fault; monitor/cable issue Try different monitor. Check 32 MHz video oscillator for output. Check GST Shifter chip (C300588/C301712). Verify Shifter is receiving 16 MHz from oscillator.
White/colored screen, no boot (no GEM) TOS ROM failure; ROM socket contact oxidation; wrong ROM jumper configuration (W102/W103/W104) Re-seat ROM chips. Clean ROM socket contacts. Verify jumper configuration matches installed ROM type. Try known-good TOS ROMs.
Boot to desktop with corrupted display ("garbage" or missing elements) RAM fault; GST Shifter fault Run RAM diagnostic (e.g. from cartridge). Try reseating or replacing SIMMs. If RAM tests pass, suspect GST Shifter.
Bombs on boot (row of bomb icons) Software exception in TOS; hardware fault triggering bus error or address error 2 bombs = bus error (often bad RAM or address line fault). 3 bombs = address error (CPU or bus issue). 4 bombs = illegal instruction (ROM corruption). See "Bomb Code Reference" below.
Boot to desktop, then random crashes Marginal PSU; intermittent RAM fault; overheating CPU or custom IC Check PSU voltages under load. Run extended RAM test. Check for thermal issues (touch test ICs — none should be too hot to hold finger on for 3 seconds).

Bomb Code Reference

Bombs Exception Common Cause
2 Bus Error Bad RAM; faulty address decoding (GST MCU); accessing non-existent hardware
3 Address Error Word/long access to odd address; CPU or bus fault
4 Illegal Instruction Corrupted TOS ROM; software incompatibility
5 Zero Divide Software bug (rare hardware cause)
6 CHK Exception Software bug
7 TRAPV Software bug
8 Privilege Violation Software attempting supervisor instruction in user mode; TOS bug (known issue with 68030 compatibility code in TOS 1.06/1.62)

Video Troubleshooting

RGB Output

Symptom Possible Cause Recommended Action
No video output at all GST Shifter failure; no 32 MHz clock; failed video output transistors Check 32 MHz oscillator output. Check GST Shifter supply voltages (+5 V at pin 64). Check video output buffer transistors near the monitor connector.
Missing color channel (e.g. no red) Failed video output transistor for that channel; broken trace; monitor cable fault Test with known-good cable. Trace signal from Shifter output pin through buffer transistor to DIN-13 connector. Check for cold solder joints on monitor connector.
Wavy/unstable picture PSU ripple; ground loop with monitor; failing filter capacitors Check +5 V rail ripple. Recap PSU if ripple exceeds 300 mV. Try different PSU. Check ground continuity.
Picture rolls vertically Wrong V-sync frequency (50/60 Hz mismatch with monitor); sync circuit fault Verify sync register setting at $FF820A (bit 1: 0=60 Hz, 1=50 Hz). Check monitor accepts the frequency. If sync is correct, suspect GLUE portion of GST MCU.
Jailbar pattern (vertical stripes) RAM noise coupling into video; marginal RAM timing Re-seat SIMMs. Try different SIMM brand. Check bulk decoupling cap C103 (4700 µF 16 V).
Only 512 colors instead of 4096 Software writing only 3 bits per channel (ST software, not STE-aware) Normal for ST software. STE-specific software or programming is needed to access the extra palette bit.

Monochrome Output

Symptom Possible Cause Recommended Action
No mono video Mono detect pin 4 not grounded (SM124 cable issue); GST Shifter fault Check cable grounds pin 4 of DIN-13 to signal ground. Test with known-good SM124 cable. Check mono video output on DIN-13 pin 11.
Mono display with flicker or shimmer SM124/SM144 monitor fault; failing GST Shifter Try different monochrome monitor. If problem persists, suspect Shifter.

DMA Sound Troubleshooting

DMA sound issues are among the most common faults on STE machines and involve the GST Shifter, 74F374 DAC latches, LMC1992 mixer, and associated passive components.

Symptom Possible Cause Recommended Action
No DMA sound at all (YM2149 PSG works) SCLK oscillator (8 MHz) not running; GST Shifter DMA section fault; LMC1992 failure; disconnected RCA output wiring Check 8 MHz SCLK oscillator output. Test with DMA sound test program (fenarinarsa's ste_dmatest). Check LMC1992 is receiving power. Verify RCA jack wiring continuity.
DMA sound crackles/pops 74F374 DAC latch noise (ground bounce from fast switching) Replace both 74F374 ICs with 74LS374 equivalents. Alternatively, add short ground wires between DAC IC ground pins and nearby ground points. If using 74F374, lift VCC pin and reconnect via 40 Ω resistor.[3]
High-pitched whistle on DMA audio Degraded coupling capacitors C202, C204, C205 (4.7 µF 35 V NP); PSU noise; missing RF shielding Replace C202, C204, C205 with fresh non-polarized electrolytics or film caps. Recap PSU. Ensure RF metal shielding is fitted over the motherboard.[4]
YM sound too loud relative to DMA LMC1992 mixer wiring error (AUDIO2 path tied to ground) Apply P.Putnik's LMC mixer fix: add 10 kΩ + 47 kΩ resistor network to restore the −12 dB YM attenuation path on LMC1992 pins 4/5/25/26.[5]
DMA sound distorted or clipped RCA output level too high for connected amplifier Add external voltage divider on RCA outputs: 1 kΩ in series followed by 150 Ω to ground (approximately 6× attenuation). This can be fitted inside an RCA plug.
Microwire volume/tone controls have no effect LMC1992 not receiving Microwire data; LMC1992 failure Verify Microwire clock and data signals at $FF8922/$FF8924 with oscilloscope. If signals are present but controls non-functional, replace LMC1992.
One stereo channel missing Failed DAC latch (74F374) for that channel; broken trace; LMC1992 channel fault Signal-trace from GST Shifter through DAC latch to LMC1992 input on the affected channel. Replace the DAC latch IC. Check coupling capacitors.

Enhanced Joystick Port Problems

Symptom Possible Cause Recommended Action
Enhanced port completely dead GST MCU failure (port logic is integrated); bent/broken pin in connector Inspect 15-pin connector for bent pins. If physically intact, likely GST MCU fault — no user-serviceable repair (BGA/QFP rework required).
Paddle input erratic Dirty connector contacts; noisy analog input; paddle controller fault Clean connector contacts with DeoxIT. Test with known-good controller. Check for noise on +5 V supply to the port.
Standard joystick works in enhanced port but no analog Normal — standard joysticks do not provide analog signals Use STE-specific analog paddles or controllers.

RAM (SIMM) Fault Diagnosis

RAM faults produce a wide range of symptoms depending on which bit(s) or address line(s) are affected.

Symptom Possible Cause Recommended Action
Machine reports less RAM than installed Failed SIMM; poor SIMM contact; SIMM speed too slow Re-seat SIMMs. Clean SIMM contacts with eraser + IPA. Try known-good SIMMs. Ensure SIMMs are 80 ns or faster.
Random crashes during operation Intermittent RAM bit failure; marginal SIMM contact Run extended RAM diagnostic from cartridge (e.g. Atari diagnostic cartridge). Swap SIMMs between sockets to isolate the faulty module.
Consistent crash at specific operations Stuck bit in RAM RAM test will identify the address and bit. Replace the faulty SIMM.
2-bomb error on boot Bus error during RAM test — likely a completely dead SIMM or address line fault Remove all SIMMs and install one pair at a time to identify the faulty module. Check SIMM socket solder joints.
Garbled display Video RAM region corrupted; bit fault in first 32 KB If only video is affected, the fault is in the RAM region used for the screen buffer (default at top of RAM). Move screen base address to test. Replace SIMMs if fault follows.

520STE note: With only 2 of 4 SIMM sockets populated, ensure the SIMMs are in the correct pair of sockets (typically the two closest to the edge of the board). Consult the motherboard silkscreen for socket numbering.

1040STE note: All four sockets populated. When diagnosing, try removing one pair at a time (operating with 512 KB) to isolate the faulty bank.

Custom Chip Failures

Chip Part Number Failure Symptoms Notes
GST MCU C302183-002 No boot; no clock outputs; address decoding failures; DMA faults; enhanced joystick ports dead 144-pin QFP; not user-replaceable without SMD rework station. Common point of failure if machine has suffered power surge.
GST Shifter C300588 / C301712-001 No video; corrupted video; no DMA sound; palette faults Socketed on many boards (DIP-64); replaceable. Two known versions — some boards have compatibility issues with certain Shifter revisions.
BLiTTER C302014 BLiTTER operations fail (software falls back to CPU); some games crash PLCC-68; can be removed to test without BLiTTER (TOS falls back to CPU blitting). On later combined MCU boards, integrated and not separately replaceable.
MC68000 CPU MC68000FN8 / MC68000P8 No boot; illegal instruction bombs; random crashes PLCC-68 (socketed) or DIP-64. Relatively available as replacement parts. Test by substitution.
MC68901 MFP MC68901 No interrupts; no serial I/O; timer-dependent functions fail; keyboard unresponsive DIP-48; available as replacement. If keyboard works but RS-232 does not (or vice versa), check the specific MFP I/O port rather than replacing.
DMA Controller C025913-38 No floppy access; no ACSI hard disk access DIP-40; relatively rare failure.
WD1772 FDC WD1772-02-02 No floppy disk access; read/write errors; disk not recognized DIP-28; test with known-good disk first. Available as replacement.
YM2149F YM2149F No PSG sound; printer port non-functional; control line failures DIP-40; dual-function IC (sound + I/O). If printer is dead but sound works (or vice versa), check associated traces before replacing.
MC6850 ACIA (keyboard) MC6850 Keyboard completely dead; mouse non-functional DIP-24; test by substitution (two ACIAs on board — swap keyboard and MIDI ACIAs to isolate).
MC6850 ACIA (MIDI) MC6850 MIDI In/Out non-functional DIP-24; same IC as keyboard ACIA — swap to confirm.
LMC1992 LMC1992 No audio output; volume/tone controls non-functional DIP-28; specific to STE. Source from STE donor boards or specialist Atari parts suppliers.

Floppy Drive Troubleshooting

Symptom Possible Cause Recommended Action
Drive does not spin disks Drive motor failure; drive belt (if belt-driven) broken/stretched; +12 V rail absent Check +12 V at drive power connector. Test drive motor separately. Replace belt if applicable.
Drive seeks but cannot read Dirty heads; head misalignment; WD1772 fault; bad disk Clean heads. Try multiple known-good disks. Test WD1772 by substitution.
Reads only single-sided disks Drive head 1 (lower) failed; head select signal fault Check drive head select signal. Try known-good drive mechanism.
Write errors / cannot format Write protect detect stuck; write head fault; WD1772 write circuit fault Check write protect switch on disk. Clean write protect sensor. Try formatting a fresh disk.
Drive LED stays on permanently WD1772 hung; DMA controller fault; GST MCU fault Reset machine. If persists, check WD1772 INTRQ and DRQ signals. Check DMA controller.

MIDI Troubleshooting

Symptom Possible Cause Recommended Action
MIDI Out dead MC6850 ACIA (MIDI) failure; output driver transistor/buffer fault; DIN connector solder joint Check solder joints on MIDI DIN connectors. Verify ACIA output at the IC pin. Test with loopback (connect OUT to IN and echo test).
MIDI In dead MC6850 ACIA failure; opto-coupler (6N138 or equivalent) failure Check opto-coupler. Test ACIA by swapping with keyboard ACIA.
Intermittent MIDI data Cable fault; marginal ACIA; noisy +5 V supply Try different MIDI cable. Check +5 V rail stability. Re-seat ACIA IC.
MIDI timing unstable 500 kHz ACIA clock drift (unlikely with crystal); software sequencer timing issue Verify 500 kHz clock at ACIA clock pin. Usually a software issue rather than hardware.

Diagnostic Tools

The following tools are useful for STE troubleshooting:

  • Atari Diagnostic Cartridge — Tests RAM, ROM, chip timing, BLiTTER, display, keyboard, mouse, joystick, floppy, DMA, MIDI, serial, and parallel ports. Requires a cartridge ROM image burned to a 128 KB EPROM.
  • fenarinarsa's STE DMA Sound Test — Specifically designed to diagnose DMA audio crackling and popping issues. Available at https://github.com/fenarinarsa/ste_dmatest/releases/
  • Multimeter — Essential for voltage checks, continuity testing, and capacitor ESR measurement (with appropriate meter).
  • Oscilloscope — Required for PSU ripple measurement, clock signal verification, and signal tracing.
  • Logic probe — Quick check for clock signals, reset lines, and chip select activity.
  • Loopback plugs — For testing serial (RS-232) and MIDI ports in self-test mode.

See Also

References

  1. The LaST Upgrade — Atari PSU Repair, exxos—link(accessed 2026-03-27)
  2. SR98 peak current update, exxos—link(accessed 2026-03-27)
  3. STE Sound DAC Crackle & Pop Fix, exxos Forum—link(accessed 2026-03-27)
  4. STE DAC Audio Noise Fix, exxos — The LaST Upgrade—link(accessed 2026-03-27)
  5. Atari STE and Mega STE sound fixes, AtariForumWiki—link(accessed 2026-03-27)