Atari 1050 Disk Drive: Checking and Adjusting Speed, Cleaning and Common Repairs
The Atari 1050 replaced the 810 in mid-1983 and became the most common Atari drive. It is robust, but forty-year-old drives drift. The usual problems are rotation speed creeping away from specification, a dirty head, a tired belt, or a failing sensor. A drive that reads its own disks perfectly while rejecting disks from elsewhere is very often running at the wrong speed. This guide shows how to measure the speed, how the speed is set, and how Atari's service manual diagnoses the rest.
For general care (power supplies, drive select switches, safe handling) see Caring for Atari 8-bit Hardware. This guide goes deeper into the drive itself.
What you'll learn: the 1050's speed specification, a BASIC program that measures RPM, where the speed is adjusted, safe head cleaning, the service manual's fault checklist, and which upgrades change the picture.
What you need: a working Atari with BASIC, the 1050 and a formatted disk you don't mind using as a test disk. To adjust or clean: a screwdriver to open the case, a small non-metallic trimmer tool, cotton swabs and 91% isopropyl alcohol.
1. How the 1050 spins a disk
The 1050 is controlled by its own 6507 processor, with a 6532 RIOT for I/O and a Western Digital 2793 floppy controller (late 1985 units use a 2797, which needs ROM revision L). Singapore-built drives use a Tandon mechanism, while the rarer Hong Kong units use World Storage Technology mechanisms.
The drive motor is a small DC motor that turns the disk through a belt and flywheel. The motor has a built-in tachometer, and a feedback circuit on the main board keeps its speed constant. That circuit includes a trimmer potentiometer, VR2, which sets the speed.
The specification is 288 RPM ± 3% (279.36–296.64 RPM). That's one revolution every 208.3 milliseconds, the figure Atari's field service manual gives for setting VR2. (Atari drives run at 288 RPM, not the 300 RPM common elsewhere.)
2. Measuring the speed
The trick is simple. Ask the drive for the same sector over and over. After each read, the computer has to process the data and send a new request, and by then the sector has already passed under the head. So the drive waits for it to come round again, and each read takes exactly one revolution. Time 20 reads with the computer's clock and you have the RPM.
10 REM DISK RPM CHECK - DRIVE 1
20 DIM B$(128)
30 FOR I=0 TO 4:READ A:POKE 1536+I,A:NEXT I
40 DATA 104,32,83,228,96
50 F=60:P=PEEK(53268)
60 IF INT(P/2)-8*INT(P/16)=0 THEN F=50
70 A=ADR(B$):POKE 769,1:POKE 770,82
80 POKE 772,A-256*INT(A/256):POKE 773,INT(A/256)
90 POKE 778,1:POKE 779,0
100 X=USR(1536)
110 T0=PEEK(18)*65536+PEEK(19)*256+PEEK(20)
120 FOR I=1 TO 20:X=USR(1536):NEXT I
130 T1=PEEK(18)*65536+PEEK(19)*256+PEEK(20)
140 IF PEEK(771)<>1 THEN PRINT "DISK ERROR ";PEEK(771):END
150 PRINT "RPM: ";INT(12000*F/(T1-T0)+0.5)/10
160 GOTO 110
How it works:
- Lines 30–40 put a five-byte machine-language routine in page 6 (address 1536):
PLA,JSR $E453,RTS. $E453 is DSKINV, the OS disk routine, which fills in the rest of the device control block itself. - Lines 70–90 set up that control block: drive 1 (location 769), command 82 = "R" for read (770), the buffer address (772–773) and sector 1 (778–779).
- Lines 50–60 check the GTIA's PAL flag (53268). The clock at locations 18–20 counts 60 ticks per second on NTSC machines and 50 on PAL.
- Line 100 makes one warm-up read so the motor is already at speed. The loop then times 20 reads and prints revolutions per minute to one decimal place. Line 140 stops with the SIO status code if a read fails.
Leave it running and watch several readings. A healthy drive gives steady numbers between about 280 and 296. Readings that wander from one line to the next point to a slipping belt or a feedback fault rather than a mis-set pot. In 1983, Atari service-centre owner Steve Switzer warned Antic readers that many speed programs don't update fast enough to show rapid fluctuation. Repeated short samples like these at least show drift.
Limits. Drives with track buffers, such as the Happy 1050 Enhancement and Speedy 1050, can answer a repeated read from memory without waiting for the disk, so this method gives meaningless results on them. Use their own diagnostic software, or switch them to slow/stock mode if they have that option. In an emulator the number only reflects the emulator's own timing model: atari800 reports about 282 for its virtual drive.
Atari also produced a 1050 Disk Diagnostics disk (1983), which includes a speed test. Copies circulate in the usual archives.
3. Adjusting the speed
Atari's own advice in 1983 was cautious. Antic told readers not to adjust their drive's speed themselves, and to take it to a repairer if a test showed a significant difference from 288 RPM. Today there's usually nobody to take it to. If you go ahead:
- Measure first. Only adjust if the drive is clearly outside 280–296 RPM, or you're trying to match a drive that wrote important disks.
- Unplug the drive and open the case. The drive mechanism sits on top of the main circuit board.
- Find VR2 on the board, in the motor/tach-feedback circuit. Turning it changes a control voltage on the feedback IC (U5), which raises or lowers the motor speed.
- Reconnect, run the RPM program with a test disk in the drive, and turn VR2 in very small steps with a non-metallic tool, waiting for a few readings after each tweak. Aim for 288.
- Close the case and confirm the readings again.
If VR2 makes no difference, the service manual points to the feedback components (VR2, U5, Q6) or an open tachometer winding in the motor, which means replacing the motor. If the speed is right but unstable, it lists a poor connection between plug P1 and socket J1, a slipping drive belt, wrong cone pressure on the disk, or worn motor bearings.
4. Cleaning the head
The 1050 manual's remedy for a dirty head is 91% isopropyl alcohol. Use a cotton swab, wipe gently and let it dry fully. Two cautions from experienced owners:
- Clean when errors suggest it, not as a ritual. A dirty head is one of the causes the manual lists for read, write and Error 144 problems.
- Keep alcohol off the pressure pad above the head, and don't scrub.
Also avoid abrasive cleaning disks: an engineer at Tandon, the 1050 mechanism maker, told Antic's Atari Clinic that they didn't recommend them. Ordinary "rubbing alcohol" can contain oils and other additives, so use isopropyl.
5. Atari's fault checklist
The 1050 Field Service Manual's symptom checklist is the best map of what goes wrong. In brief:
| Symptom | Likely causes (per the manual) |
|---|---|
| Motor doesn't start when the latch is closed (no activity light) | Diskette-enable switch on the mechanism; P1/J1 connection; motor-control parts (Q4, Q6, U5, VR2); motor |
| Speed too fast/slow, or not adjustable | VR2 setting; tach feedback parts (VR2, U5, Q6); open tachometer winding |
| Speed unstable | P1/J1 connection; slipping belt; cone pressure; feedback circuit; motor bearings |
| Head won't step, or steps to the wrong track | J15 connector polarity; stepper drivers (U2, U3, U7); binding band/pulley or carriage; track 0 sensor (Q5) |
| Intermittent read/write errors | Dirty or burned head; track alignment; head azimuth; belt slip; speed |
| Error 144 when writing or formatting | Write-protect circuit (U11, U13); erase or write circuit; dirty head |
| Always or never write-protected | Write-protect photo sensor or its circuit |
| No power or power light | Bridge diodes CR17–CR20; power adapter; +5 V regulator Q7; +12 V regulator Q8 |
| Drive doesn't answer as the right number | Drive select switch; PIA (U7) |
Many cures in the original manual simply say "replace drive mechanism", because that was the field-service policy then. Today the usual alternative is parts from a donor drive.
The drive number is set by the two switches on the back: both left is drive 1, black right and white left is drive 2, both right is drive 3, black left and white right is drive 4.
6. Upgrades and how they affect testing
Many 1050s were upgraded, so check inside before you trust a test result:
- ICD US Doubler (1984): a replacement ROM and extra RAM chip. Adds true double density (180K) and "UltraSpeed" fast transfer with SpartaDOS-style skewed formatting.
- Happy 1050 Enhancement (1985): its own 6502 processor and RAM, with whole-track buffering and Warp Speed; later ROMs also emulate the US Doubler.
- Speedy 1050 (Compy-Shop, 1986) and 1050 Turbo (1986) were especially popular in Europe. Modern boards such as Mega Speedy (2014) can switch between several of these personalities.
Buffered upgrades make the RPM program above meaningless, but they don't change the physical speed setting. The drive still needs to spin at 288 to read stock disks.
Troubleshooting
- The program prints
DISK ERROR 138. No answer from drive 1: check power, cable and the drive select switches. 139 means the drive refused the command. 144 means the read failed, so try another disk. - The reading is about double or half what you expect. Your machine is PAL but reports NTSC, or the other way round. Set
Fby hand (50 for PAL, 60 for NTSC). - Readings jump around by several RPM. Suspect the belt, the P1/J1 connector or the feedback circuit before touching VR2.
- The drive reads its own disks but not others'. Check the speed on both drives, as the manual suggests, then clean the head. Alignment problems come last.
Further reading
- Atari 1050 Disk Drive Field Service Manual (FD100330; Rev. 01 1983, Rev. 03 1985), especially the symptom checklist and the tach feedback waveform page.
- Sams Computerfacts CD13, Atari Model 1050 Disk Drive (1987): schematics, parts and preliminary service checks.
- Steve Switzer, "Atari Clinic", Antic, 1983, on speed, cleaning and cleaning disks.
- comp.sys.atari.8bit FAQ, section 3.2.2, for 1050 specifications, ROM revisions and every known upgrade.
- On this site: Caring for Atari 8-bit Hardware, SIO Explained and Page 6 usage.
Credits
- Picture: A 1050 with the top cover off: the drive mechanism sits over the main circuit board. — MOS6502, Wikimedia Commons (CC BY-SA 3.0)
- comp.sys.atari.8bit FAQ 3.2.2 — Atari 1050 specifications, ROMs, upgrades — Michael Current (comp.sys.atari.8bit FAQ)
- Antic v2n4 — Atari Clinic (Steve Switzer): speed programs, isopropyl, cleaning disks — AtariMagazines.com (Kevin Savetz)
- Antic v2n5 — I/O Board (don't adjust speed yourself; 288 RPM) — AtariMagazines.com (Kevin Savetz)
- DSKRPM.BAS (read-same-sector RPM method)
- AtariAge — MyDOS Enhanced Format thread (1050 Disk Diagnostics, Happy vs US Doubler); Atari 810 drive problem thread (cleaning, pressure pad) — AtariAge forum members
- Mapping the Atari — 18-20 RTCLOK, 769-779 DCB, 53268 PAL, 58451 DSKINV — AtariArchives.org
