IBM PS/2 Model 50 Troubleshooting Guide
.jpg)
Fault-finding on the IBM PS/2 Model 50 (machine type 8550: the 8550-021 and the Model 50 Z, 8550-031 and 8550-061) follows IBM's Micro Channel checkout procedure and symptom-to-FRU index, published in the October 1994 PS/2 Hardware Maintenance Manual (HMM). The same procedure covers the Models 50, 55, 56, 57, 60, 65, 70, 76, 77, 80, 85, 90 and 95.[1] The numeric error-code tables below are shared with the Model 60, Model 70 and Model 80 guides. Power-supply test points, the battery test and the fixed-disk notes are in the IBM PS/2 Model 50 Maintenance Guide.
Before you start
[edit | edit source]IBM's general checkout for Micro Channel machines begins the same way every time: power off the computer and all external devices, check every cable and power cord, remove any diskettes, power on the external devices and then the computer. A good start gives one or two short beeps followed by readable instructions or the main menu. If a POST error code appears, press Pause and write it down before pressing F1.[1] After a clean POST the Model 50 gives one short beep and hands over to the bootstrap loader. If no operating system is found it shows the Insert Diskette icon, and pressing F1 without a diskette starts IBM Cassette BASIC.[2]
Rules from IBM's procedures that change the order of work:
- The most likely cause is always listed first in each FRU list. If several error codes appear, diagnose them in the order they appear, because the later ones may be false errors. If a POST error comes with the wrong beep pattern, diagnose the error code first.[1]
- Start the Reference Diskette with a cold boot, with the diskette in drive A at power-on. IBM's note, reproduced by Ardent Tool, says a warm boot of the Reference Diskette can give false errors and a false indication that a power-on password is already set.[3]
- Run Advanced Diagnostics before replacing parts. From the Reference Diskette main menu, press Ctrl+A. If the diagnostics pass, replacing hardware is unlikely to help.[1][3]
- Where an adapter is made up of more than one FRU (a card with memory on it, for example), remove or exchange the parts one at a time before replacing the whole adapter.[1]
- A surge suppressor between the machine and the wall is on IBM's list of things to remove when a fault cannot be found.[1]
- The Model 50 keeps its configuration in 64 bytes of battery-backed CMOS RAM, set from the Reference Diskette. Configuration is stored against the slot an adapter occupies, so put a replacement adapter back in the same slot, or run Set Configuration again.[2]
POST beep codes
[edit | edit source]| Beeps | FRU / action, most likely first |
|---|---|
| One or two short beeps with readable instructions or the main menu | Normal |
| One long and one short | Display adapter (the system-board VGA on a Model 50), system board, video memory, bus adapter, power supply |
| One long and two short | Display adapter, video memory, system board, adapter video memory, bus adapter, power supply |
| One long and three short | Display adapter, video memory, system board |
| One or two beeps with a blank or unreadable display, or a blinking cursor | Display adapter, video memory, system board, adapter video memory, display, bus adapter |
| Continuous beep | System board, power supply |
| Continuous beep and a blank display | Display adapter, video memory, system board |
| Repeating short beeps | Check the keyboard for a stuck key; then system board |
IBM also lists the ACPA/A audio adapter against some of these patterns; it applies only to machines fitted with one.[1]
Numeric POST error codes
[edit | edit source]IBM's index lists every code as eight characters. The screen often shows a shorter form: add two zeros to the right of a three-, four- or five-character code, then pad with zeros on the left to eight characters, so 301 is 00030100 and 1701 is 00170100. If a twelve-character code appears, ignore the last four characters. An X in a code means any digit.[3][1] The tables give the short form and IBM's full form. FRUs that IBM marks for the Models 90 and 95 only (processor board, processor cache, memory riser) are left out.
System board, 101–159
[edit | edit source]| Code | IBM's note | FRU / action |
|---|---|---|
| 101 (000101XX) | Interrupt failure; possibly a bad battery or processor | Run Advanced Diagnostics |
| 102 (000102XX) | ROM checksum or timer error; possibly a bad battery or processor | Run Advanced Diagnostics; system board |
| 103, 104, 107 | System board | |
| 105 | Command not accepted | Information only |
| 106 | Converting logic test failure | Run Advanced Diagnostics |
| 107, 108, 109 | Interrupt or memory failure | System board memory, system board, any adapter, bus adapter |
| 110 | Check memory | Memory module kit, system board |
| 111 | Adapter memory, expansion adapter, system board | |
| 112, 113 | Possible timeout error | System board |
| 114 | ROM error | Any adapter, bus adapter |
| 115 | 80386 protect mode failure, or BIOS checksum error | System board |
| 116 | Possible read/write problem | Run Advanced Diagnostics |
| 118 | Previously detected error; run Advanced Diagnostics | System board memory |
| 119 | 2.88 MB diskette drive installed but not supported | Run Automatic Configuration; system board |
| 120 | Possible processor self-test failure | System board |
| 121 | A problem was detected during memory remapping | Information only |
| 00012201 to 00012208 | Data error; possible system board failure | Run Advanced Diagnostics |
| 129 | Processor board, cache kit, system board | |
| 130 (00013000) | POST could not start the operating system. Is the boot drive in the selectable drive sequence? | Check the drive sequence |
| 131 | System board | |
| 132, 133, 134 | DMA extended-register, verify and arbitration errors | Run Advanced Diagnostics |
| 152 | Real-time clock error; not always a hardware failure | Run Advanced Diagnostics |
| 10x and 11x not listed (00010XXX, 00011XXX) | System board |
Codes 00013001 to 00013003, 00014905 to 00014909 and 156 concern selectable startup sequences, tamper-evident switches and key locks on later models. The Model 50's security is a cover lock and a power-on password.[2][1]
Configuration and battery, 160–199
[edit | edit source]| Code | IBM's note | FRU / action |
|---|---|---|
| 160 | System board ID not recognised; possible system board failure | Run Advanced Diagnostics |
| 161 | Battery. IBM's caution: the lithium battery can cause fire, explosion or severe burns, and another type of battery could ignite or explode | Battery, system board, bus adapter |
| 162 | Be sure all devices are powered on; check the Enable/Disable settings. Configuration changed? If so, run Automatic Configuration | Any device, battery |
| 163 | Date and time error | Set the date and time |
| 160, 164, 165, 167, 169 (00016000 etc.) | If setting the configuration, date and time does not cure it, check the devices list | Set configuration and features; system board |
| 166 | Reseat all adapters | Run Advanced Diagnostics |
| 168 | Real-time clock error; not always a hardware failure | Run Advanced Diagnostics |
| 171 | Battery, system board, bus adapter | |
| 172 | System board | |
| 173 | Possibly a weak battery | Set configuration and features |
| 174 | If the configuration has changed, run Automatic Configuration; otherwise run Advanced Diagnostics | Any device, system board, bus adapter |
| 191 | 82385 cache test failed | Run Advanced Diagnostics |
| 199 | User indicated the configuration is invalid | System board |
| 1xx not listed (0001XXXX) | System board, 486DX33 upgrade (Models 70 and 80), any adapter, bus adapter |
The other codes in this range (169XX processor configuration, 170 ASCII console, and 175 to 189 for security EEPROMs, tamper switches, privileged-access passwords, vital product data and remote IPL) belong to later machines with those features (HMM pp. 36–38).[1]
The Model 50 battery is IBM part 72X8498, a 6 V lithium pack, in battery and speaker assembly 33F5950. IBM retain tip H024809, reproduced by Ardent Tool for the Models 60, 65 SX and 80, which use the same battery and assembly, says intermittent 161, 162 and 163 errors can come from low battery voltage or a faulty assembly, and that the battery should be replaced if it reads below 5.5 V DC out of the assembly.[1][2][4][5] The full test is in the maintenance guide.
Memory, 200–299
[edit | edit source]| Code | IBM's note | FRU / action |
|---|---|---|
| 201 to 205 | Reseat system board memory | System board memory, system board |
| 20x, 217 (00020XXY, 000217XY) | System board memory, system board | |
| 210, 211 | Check memory | System board memory, system board |
| 214, 215, 216, 221, 226, 235, 240, 241 | Check memory: it might be the wrong type or in the wrong slot | System board memory, system board; then bus adapter |
| 225 | Check memory: it might be the wrong type or the wrong speed | System board memory, system board; then bus adapter |
| 221 | ROM-to-RAM parity error | System board |
| 231 | Expanded memory option | |
| 245, 246 | Check memory | System board, system board memory |
| 251 | Memory location changed on the memory expansion option | Information only |
| 255 | Check memory | System board memory, system board |
| 290 (00029000) | Unsupported combination of ECC and parity modules | Correct the combination, run Automatic Configuration, rerun Advanced Diagnostics |
| 291 to 294, 298 | Checksum value mismatch | Run Automatic Configuration, then rerun Advanced Diagnostics |
| 295, 296 | Check memory for an unsupported configuration or modules | System board memory |
On the Models 50, 55, 60, 65, 70 and 80, IBM finds a failing memory kit by running the memory test and then exchanging kits one at a time with a known-good one; if the kits are not the fault, suspect the riser card or memory expansion adapter, then the system board. Only on the Models 53, 56, 57, 76, 77, 85, 90 and 95 does a digit of the 2xx code identify the socket.[1]
The Model 50 reallocates memory in 1 MB blocks. An error in the first 1 MB on the system board stops the machine with an error message. An error above that is reported once, its address is stored, and later starts map the bad area to the top of memory without a message, at the cost of capacity. Up to two such errors are tolerated; a third stops reallocation and all errors must then be fixed. Advanced Diagnostics shows the stored errors.[2]
Keyboard, ports, display, diskette and coprocessor, 300–999
[edit | edit source]| Code | IBM's note | FRU / action |
|---|---|---|
| 301, 302 | Keyboard cable, system board | |
| 303, 304 | System board, keyboard, keyboard cable | |
| 305 | Keyboard voltage error. If the system has no fuse, replace the system board | Fuse, keyboard cable |
| 306 | Wrong keyboard attached? | Check for an unsupported keyboard |
| 307 | Keyboard, keyboard cable | |
| 320 | System board | |
| 401 to 405 | System board | |
| 5xx | Display adapter (any type) | |
| 601 | Diskette, diskette drive, system board | |
| 602 | Invalid boot record | Diskette |
| 604 | Unsupported drive or cable; the drive must be a media-sense type | Diskette drive, system board, diskette drive cable |
| 605 | Diskette drive error | Run Advanced Diagnostics |
| 606, 607, 610, 621 to 624, 630 to 633, 640 to 643, 650 to 654, 656 to 660 | Generally media errors. Try a known-good diskette; if the error returns, replace the drive | Diskette, diskette drive, system board |
| 655 | System board | |
| 662 | Wrong drive type installed | Information only |
| 663 | Wrong media type in the drive | Information only |
| 668 | Diskette drive | |
| 6xx not listed | Diskette drive, system board, diskette drive cable | |
| 7xx | Math coprocessor. With a 486 processor, erase COPROC.DGS from the backup Reference Diskette, restore the system partition from it and rerun Advanced Diagnostics | Math coprocessor, 486DX33 upgrade (Models 70 and 80), system board |
The Model 50's keyboard and mouse fuse is on the system board (F1, 3 A, IBM FRU 72X8561), so a 305 on this machine points first at that fuse.[1][2] The math coprocessor must be a 10 MHz 80287. Fitted the wrong way round the machine will not run, and IBM warns that the wrong coprocessor can cause intermittent errors. The notch faces the rear of the system unit on the Type 1 board (8550-021) and the left side on the Type 2 board (Model 50 Z).[2]
Other codes that turn up on these machines: 11xx is the system-board serial port, 14xx the parallel port and printer, 8601 to 8604 the pointing device, 8001 to 8012 the pointing device after a Model 50 or 55 system-board upgrade, and 00850XXX to 00852XXX the 80286 Memory Option Adapter/A and its memory (HMM pp. 42–45).[1]
ST-506 fixed disk and video, 1500–2999
[edit | edit source]| Code | IBM's note | FRU / action |
|---|---|---|
| 1701, 1703, 1704, 1714, and any 17xx not listed below | Hard disk drive, ST-506 cable, ST-506 hard disk adapter, system board, power supply | |
| 1702 | Hard disk adapter | |
| 1705 to 1708, 1710 to 1713, 1715 to 1717, 1726, 1735, 1750 to 1755, 1757, 1780 to 1782, 1790, 1791 | Read/write problem. Check the drive type is supported; if it is, try a low-level format; if the error continues, replace the drive | Format the drive; hard disk drive |
| 1801 to 1807 | System board | |
| 186x (00186XXX) | Set configuration and features; battery | |
| 18xx not listed | System board expansion unit | |
| 2401, 2402 | If the screen colours change | Display |
| 2401, 2402 | If the screen colours are correct | System board, display |
| 2409 | Display | |
| 2410 | System board |
On the Model 50 the 17xx codes apply to the 8550-021's 20 MB ST-506 drive (FRU 72X8522) and its adapter.[1][6]
ESDI fixed disk, 10400–10499
[edit | edit source]| Code | IBM's note | FRU / action |
|---|---|---|
| 10450, 10451, 10490, 10491, 10492, 10499 | Read/write error (10490 and 10491: drive 0 or 1 read error; 10499: drive controller error) | Run Advanced Diagnostics; riser card if installed, hard disk drive, system board |
| 10452 | Seek test error | Run Advanced Diagnostics |
| 10453 | Wrong drive type? | Information only |
| 10454 | Sector buffer test error | Run Advanced Diagnostics |
| 10455, 10456 | Controller error | Run Advanced Diagnostics |
| 10459 | Drive diagnostic command error | Information only |
| 10461 | Drive format error | Run Advanced Diagnostics |
| 10462 | Controller seek error | Run Advanced Diagnostics |
| 10464 | Hard drive read error | Run Advanced Diagnostics |
| 10467 | Drive non-fatal seek error | Run Advanced Diagnostics |
| 10468 | Drive fatal seek error | Run Advanced Diagnostics |
| 10469 | Drive soft error count exceeded | Run Advanced Diagnostics |
| 10470, 10471, 10472 | Controller wrap error | Run Advanced Diagnostics |
| 10473 | Corrupt data; a low-level format might be required | Information only |
| 10480 | ESDI hard disk drive, drive cable, controller, system board | |
| 10481 | ESDI drive D seek error | Run Advanced Diagnostics |
| 10482 | Drive select acknowledgement bad | Run Advanced Diagnostics |
| 10483 | ESDI hard disk adapter, system board | |
| 104xx not listed | Hard disk drive, ESDI hard disk adapter, hard disk cable, power supply |
The codes are not a continuous run: IBM's index has no 10457, 10458, 10460, 10463, 10465, 10466, 10474 to 10479, 10484 to 10489 or 10493 to 10498.[1] The integrated drives of the Model 50 Z, 55 SX, 70 and P70 combine the disk and a Micro Channel controller in one unit; Peter Wendt describes them as technically MFM RLL 2,7 drives with the controller built in; IBM's BIOS treats them as ESDI, so their faults report in this range.[6]
Dead system
[edit | edit source]IBM's no-beep symptoms, with the most likely FRU first:[1]
| Symptom | FRU / action |
|---|---|
| No beep, power light off, fan not running | Power supply, control/speaker assembly, system board, any device or adapter, bus adapter |
| Power light off, fan not running, computer not working | Power supply, system board |
| Power light off, fan running, computer not working | System board, power supply |
| Power light off, fan running, computer works | Control/speaker assembly |
| No beep, fan running, power light on, hangs in POST with no message | System board, any device or adapter, bus adapter |
| No beep, fan running, power light on, hangs in POST with a message | System board, power supply |
| No beep and the computer otherwise works | Control/speaker assembly, system board |
| Memory count hangs during POST | System board memory, memory option if fitted, system board, processor |
| Supply seems to fail at power-on, but disconnecting the optional hard disks cures it | Remove the motor-start jumper from one or more option drives |
On the Model 50 the speaker is part of the battery and speaker assembly.[2] The green power-good light is lit by the supply's Power Good signal, which the supply asserts only once +5 V, +12 V and −12 V have passed their under-voltage sense levels of +4.5 V, +10.8 V and −10.2 V; if the light is off, IBM's reading is that the supply is not working properly.[2] The 8550 supply has no fan of its own. The fan is a separate assembly, FRU 72X8515, plugged into the system board.[1][7]
The procedure Ardent Tool reproduces for a machine with no POST, a blank screen and no beep:[3]
- Check the line cord at both ends. IBM notes that loose or badly seated line cords cause intermittent faults that are very hard to find, including unexpected resets during operation. Take the cord out of the wall before re-forming any contacts, and replace the cord if in doubt.
- Remove all adapters, options and extra memory.
- Remove all external connections: keyboard, mouse and display.
- Plug in, power on and listen. A beep means POST has run.
- If there is no beep, check the speaker for continuity. If it is good, replace the system board memory and try again. If there is still no beep, measure the supply voltages: if they are right, replace the system board; if not, replace the power supply.
- If there is a beep, refit the adapters one at a time, powering on after each. The item fitted just before the beep stops is the suspect. There is no need to reconfigure between steps.
Ardent Tool gives the usual cause as a device shorting the supply, such as a memory module in backwards or a defective adapter, and states that the system board must return Power Good to the supply within 150 ms or the supply shuts down. Its contributors add adapters not fully seated, unsupported adapters, and on heavily loaded machines too many drives spinning up at once.[3] Supply test points and IBM's limits are in the maintenance guide.
If the voltages are correct and nothing else has found the fault, IBM's last procedure is to remove parts one at a time, powering on and reconfiguring each time: external devices, surge suppressor, modem, printer, mouse and non-IBM devices, each adapter, the drives, the memory kits, the bus adapter and the coprocessor. A Micro Channel machine needs at least 1 MB of memory to run. If the fault stays with everything removed, suspect the system board.[1]
Configuration errors
[edit | edit source]IBM's action for 162 is to check that every device is powered on and its Enable/Disable setting is right, and to run Automatic Configuration if the configuration has changed; for 165 it is to set the configuration and features, then suspect the system board.[1] The Model 50 stores configuration against the slot each adapter occupies, so an adapter moved to another slot needs Set Configuration run again.[2] Automatic Configuration can change customised settings, so note them first with View configuration and restore them afterwards.[1] If an installed adapter is missing from the Advanced Diagnostics devices list, IBM's first suspect is that its files were never copied from the adapter's option diskette (Copy an option diskette); the others are a defective adapter or device, or one the diagnostics do not recognise.[1]
If the configuration is lost because the battery was flat or removed, IBM's instruction is to run the restore configuration program from the user's backup copy of the Reference Diskette where custom settings were used, or Automatic Configuration where the defaults were used.[2] A machine that forgets only part of its setup probably does not have a battery fault; Ardent Tool's note is that a weak battery loses all of it.[3]
The clock on these machines is interrupt-driven and drifts with interrupt activity. IBM's check for "real-time clock loses time" is to read the date and time in Set Features: if they are right there, the problem is software. Some network software also resets the workstation clock from the server.[1] A clock that stands still while the machine is off, with no 161 or 163 yet, is the battery nearing the end of its life.[5]
Fixed disk will not spin up
[edit | edit source]A drive that has stood unused may not start because the heads have stuck to the platters. Peter Wendt's advice, published by Ardent Tool: try a few power cycles, or unplug the system unit, shake it up and down a few times and turn it sharply clockwise, then power on. Never lift the front and drop it on the desk. However it starts, copy the data off straight away. He names the Seagate ST-177I 60 MB drive in some Model 55 SX and Model 70 machines as prone to this.[8] Before moving the machine, select Move the computer on the Reference Diskette to park the heads.[2]
Common symptoms
[edit | edit source]| Symptom | Probable cause | Action |
|---|---|---|
| Click at power-on, then nothing | Supply shut down because something is loading a rail | Dead system procedure: strip to the system board and add parts back |
| 161 at every start | Battery flat or assembly faulty | Test the 72X8498 battery out of the assembly; replace below 5.5 V[5] |
| Intermittent 161, 162 and 163 | Low battery voltage, faulty battery and speaker assembly | As above; H024809[5] |
| 162 or 165 after changing cards | Configuration no longer matches the hardware | Configuration errors |
| Machine will not run after a coprocessor was fitted | 80287 in the wrong way round, or wrong speed | Check the notch and the part number[2] |
| Intermittent errors after fitting a coprocessor | Wrong speed coprocessor | Fit a 10 MHz 80287[2] |
| 2xx after changing memory | Wrong type, wrong slot or wrong speed | Check the module against the machine's own specification; exchange kits one at a time[1] |
| Diskette read/write errors, 6xx | Media, drive, or failed capacitors on the drive board | Try a known-good diskette, then the drive; see IBM PS/2 Model 50 Capacitor Replacement Guide[5] |
| Diskette LED stays on | Diskette drive | Replace or repair the drive[1] |
| Hard disk LED stays on | Hard disk drive, then system board | [1] |
| Unexpected resets during use | Line cord loose at either end, then the supply | Reseat or replace the cord[3] |
| 2401 or 2402 | Display or system board | If the screen colours change, the display; if the colours are correct, the system board, then the display[1] |
| One or more keys do not work | Keyboard, keyboard cable, system board | [1] |
Related pages
[edit | edit source]- IBM PS/2 Model 50
- IBM PS/2 Model 50 Maintenance Guide
- IBM PS/2 Model 50 Capacitor Replacement Guide
- IBM PS/2 Model 60 Troubleshooting Guide
- IBM PS/2 Model 70 Troubleshooting Guide
- IBM PS/2 Model 80 Troubleshooting Guide
- IBM Personal System 2 Hardware Maintenance Manual (S52G-9971-02, October 1994)
- IBM PS2 Model 50 Hardware Maintenance Reference
- Capacitor Failure Symptoms
- Recommended Tools
References
[edit | edit source]- ↑ 1.00 1.01 1.02 1.03 1.04 1.05 1.06 1.07 1.08 1.09 1.10 1.11 1.12 1.13 1.14 1.15 1.16 1.17 1.18 1.19 1.20 1.21 1.22 1.23 1.24 1.25 1.26 1.27 IBM, IBM Personal System/2 Hardware Maintenance Manual, S52G-9971-02, October 1994. Hosted on this wiki as IBM Personal System 2 Hardware Maintenance Manual (S52G-9971-02, October 1994) (PDF). General Checkout (Micro Channel), pp. 19–25, including the devices list and "Copy an option diskette", p. 23; Symptom-to-FRU Index (Micro Channel): no-beep symptoms pp. 27–28, beep symptoms p. 29, miscellaneous symptoms pp. 30–32, numeric error codes pp. 33–51; combined FRUs, p. 108; how to use the index and error-code format, pp. 114–115; memory problems, p. 125; power supply voltage check, p. 138; real-time clock problems and running Automatic Configuration, p. 150; undetermined problems, p. 183; Model 50 parts, p. 240.
- ↑ 2.00 2.01 2.02 2.03 2.04 2.05 2.06 2.07 2.08 2.09 2.10 2.11 2.12 2.13 IBM, IBM PS/2 Model 50 Hardware Maintenance Reference, S64F-3989-00, March 1990. Hosted on this wiki as IBM PS2 Model 50 Hardware Maintenance Reference (PDF). Security (cover lock and power-on password), 1.1; CMOS RAM and restoring lost configuration, 1.2.2; keyboard and pointing-device fuse, 1.2.8; parking the heads, 2.2; power supply and under-voltage sense levels, 3.1; POST, 3.2; memory reallocation, 3.3.4; battery caution (use only 72X8498), 6.2; adapters and slot-dependent configuration, 6.4; math coprocessor orientation and speed, 6.10.
- ↑ 3.0 3.1 3.2 3.3 3.4 3.5 3.6 Ardent Tool of Capitalism, "PS/2 Error Codes", content collected by Louis F. Ohland, Peter H. Wendt and others, maintained by Tomáš Slavotínek. Error-code format and padding rule; IBM's notes on cold-booting the Reference Diskette, multiple POST errors and running Advanced Diagnostics before replacing parts; the dead-system procedure, the line-cord note and the 150 ms Power Good requirement; the contributors' list of common causes; and the note that a weak battery loses all of the setup while a partial loss is probably not the battery.
- ↑ Ardent Tool of Capitalism, "PS/2 RTC/CMOS Batteries", based on content by Bob Eager and Peter H. Wendt. Models 50, 50 Z, 60, 70, P70 and 80: FRU 72X8498, a 6 V two-cell lithium pack in the battery/speaker assembly.
- ↑ 5.0 5.1 5.2 5.3 5.4 Ardent Tool of Capitalism, "60, 65 SX, and 80 – Common Devices". IBM retain tip H024809, "Low Battery Voltage May Cause Config Errors": intermittent 161, 162 and 163 errors from low battery voltage or a defective speaker/battery assembly 33F5950; remove battery 72X8498 from the assembly, measure on the 12 V DC range, replace below 5.5 V DC. Alfred Arnold on the clock stopping before the CMOS contents are lost.
- ↑ 6.0 6.1 Ardent Tool of Capitalism, "8550 HDD Adapters". The Model 50 drive adapters and risers, and Peter Wendt's explanation that the Model 50 Z's integrated drives are a Micro Channel controller and disk in one unit, not true ESDI.
- ↑ Ardent Tool of Capitalism, "8550 / 8570 / 756x – Power Supply". The 8550 and 50 Z supply has no built-in fan and a lint screen on its underside.
- ↑ Ardent Tool of Capitalism, "Known Problems – Experience, RETAIN Tips, ECAs". Peter Wendt, "Hard-Drive Doesn't Spin up – 'Stiction' Issue".