Atari 1050 Service Manuals (Atari Field Service Manual and Sams Computerfacts CD13)

Year
1985
Type
Component
Category
Service documentation: 1050 specifications, diagnostic diskette, alignment and fault-finding
Manufacturer
Atari, Inc. / Atari Corp.; Howard W. Sams & Co.
Green cover of Sams Computerfacts CD13, 'Atari Model 1050 Disk Drive', showing the cream 1050 drive with its dark front bezel

Overview

Two service documents cover the Atari 1050 from the inside. Atari's own 1050 Disk Drive Field Service Manual (Revision 03 is dated 1985) was written for authorised service centres. It explains the theory of operation, Atari's diagnostic diskette and the alignment procedure, and gives symptom checklists, schematics for the Tandon-mechanism version, and a parts list. Howard W. Sams & Co.'s Computerfacts CD13, dated February 1987, is the independent version. It is shorter and practical, built around voltage readings, motor-winding resistances and a "check this chip by substitution" guide. Read together they cover most of what can go wrong with a 1050. A cleaned-up rebuild of the Atari manual also circulates, and its text matches the original.

The drive, by the numbers

Atari's specification appendix describes the 1050 as cheaper than the 810, fully 810-compatible in single density (FM), and able to read and write a double density using MFM. It also says the drive supports a new high-level command "requested by next generation Atari disk operating system".

Single density Double density (as specified)
Tracks 40 at 48 TPI 40 at 48 TPI
Sectors × bytes 18 × 128 26 × 128
Formatted capacity 92,160 bytes 133,120 bytes
Encoding / transfer rate FM, 125,000 bit/s MFM, 250,000 bit/s
Rotation 288 rpm ±3% 288 rpm ±3%
Track-to-track / settle 40 ms / 30 ms 40 ms / 30 ms

The "double density" here is the 1050's 26-sector enhanced density, which the manual says works only with DOS 3. Inside, the drive has a 6507 CPU, a ROM, RAM, a 6532 RIOT ("PIA" in the manual) that drives the four-phase stepper and reads the drive-select switches, and a 2793 floppy controller. The power supply turns 9 V AC from the external adaptor into regulated 5 V for the logic and 12 V for the motors and read/write circuits. A power-up circuit holds the SIO data output off while the drive initialises, so start-up noise never reaches the computer.

Atari's diagnostic routine

Power-up test. At switch-on, a healthy 1050 lights both LEDs, spins its motor, steps the head to track 0, then stops after a few seconds. In that time it has checked the RIOT, the 2793's registers, the ROM checksum and every RAM location, and has found track 0. If the head steps more than 50 tracks without seeing the track-0 sensor, or any other check fails, the drive enters a "hard failure" state. The spindle motor then cycles on and off every two seconds.

Diagnostic diskette (FD100690). The diskette boots on a 16K computer without a cartridge. It runs controller, invalid-command, write-protect, motor-start, motor-speed, head step-and-settle and track-00 tests, then opens a troubleshooting menu:

  • Speed calibration. Turn VR2 until one revolution takes 208.3 ms ±1 ms, which works out to 288 rpm.
  • Manual alignment with a Dymek analog alignment disk and a dual-channel scope. Use track 16 for the radial "cat's-eye" check: loosen the stepper motor and rotate it until the two lobes match, accepting 80–100%. Track 33 checks the head-load pad pressure (adjust in eighth-turns, and replace the pad after three-quarters of a turn). Track 34 checks head azimuth. If azimuth is off by more than ±18 minutes of angle, Atari says to replace the whole mechanism rather than adjust it.
  • Drive-select switch test, a 10-minute single-density burn-in, and a 2-hour double-density burn-in that logs errors onto the test disk for printing.

The manual includes an early tech tip, too. The first 1,700 drives had Rev. E or F ROMs, which returned a different error status from the 810's when reading some copy-protected disks. Service centres replaced the ROM at U10 with the Rev. 1 part and sent the old one back to Atari for reburning. (See 1050 firmware and mechanisms for later revisions.)

The Sams approach

CD13's tests were made on an Atari 800 running a short BASIC loop that writes D:SAMS.FIL again and again. Its checks:

  • No power. The pack should give about 10.9 V AC. Then look for 11.1 V at diodes CR17/CR18, 5 V at TP13 (regulator Q7) and 12 V at TP14 (regulator Q8, fed by 22.4 V).
  • Drive motor dead. Check the diskette-enable switch S3 and the connectors. Motor windings should read about 11 Ω (pins 1–2) and 186 Ω (pins 3–4). RIOT pin 11 should read 3.2 V with the motor off and about 0.1 V with it running.
  • Stepper dead. Look for 12 V on P15. Each winding should read about 33 Ω. Then check the phase waveforms at RIOT pins 18, 19, 21 and 22.
  • No communication. Check the drive-select switch S2, the cable, the RIOT (U7) and the amplifier IC U1. An "I/O failure" suggests the RAM (U8). A "boot error" suggests the controller (U13).
  • Erratic reading. Clean the head with 91% isopropyl alcohol on a lint-free cloth. Sams warns that cleaning diskettes may be too abrasive. Check the pressure cone for broken tabs, the belt tension and the speed. (Sams quotes 300 rpm, but Atari's figure is 288 rpm.)

Four Phillips screws in the base release the top cover.

Using them today

The Atari diagnostic disk, the Dymek alignment disk and a scope are rare now. The speed check, however, can be done with any modern speed-test program, and the target is Atari's 208.3 ms per revolution, not Sams' 300 rpm. The resistance and voltage figures above are the quickest way to tell a dead motor from a dead board. For drive upgrades see Happy 1050, Speedy 1050 and 1050 Turbo.

Credits