POST Error Codes
General Error Codes
IML Error Codes (I99xxxx)
General SCSI Error Codes
Multiple POST Errors
Troubleshooting Dead System
Clearing the 96 8N1 code
This page is about one of the most annoying things in the PS/2 World - Error
codes. They come up when you need them the least and -often- don't tell much
why anything went wrong.
Now here's a collection of the most common error codes along with
some explanation and which action to take or which part to change.
POST Error Codes
|* *||A problem exists, check config. Run diags, if no error code displayed|
|_ *||Display or display adapter problem.|
|_ * *||Display or display adapter problem.|
|_ * * *||Display or display adapter problem.|
|* * * * * *||Continuous short beeps. Check keyboard.|
|___________||Continuous solid beep. System board.|
(* indicates a short beep, _ means a long beep)
General Error Codes
(from The Microchannel Enthusist's Page)
Major Error Code -----------+---+
Trailing Zeros -------- 0024 01XX
to fill 8 digits ||
Minor (diagnostic Error Code)----++
The Error Code is written in the 4 + 4 style as it appears on a PS/2 Mod.
95 LED-panel and in the Premium Line "extended 8-digits" form. The bold
part in the middle is the so called "Major Error Code" and -basically and
in large parts- identical with the PC/AT Error Codes and those used on
earlier PS/2 machines.
The 2 digits of the "Minor Error Code" is dependent on the type of the
error and -mostly- only specified when running the Advanced Diagnostics.
The minor error code is however often given at POST-Errors (after restarting
the computer) and might specify a particular error condition. In
case the minor error code is marked with "XX" in the follow-up error lists
it means "Don't care" and it can be any character.
Error messages generally consist of eight characters. If an error message
appears as a 12-character message, disregard the last four characters.
The following errors are all listed as eight-character messages. Messages
might appear on your screen as three-, four-, or five-character messages. When
this occurs, always add two zeros ["00"] to the
RIGHT of the code, and add enough zeros "0" to
the LEFT of the code to reach eight characters.
Error Code Format Examples:
00030100 - Three-character, Codes 300 - 999
00170100 - Four-character, Codes 1500 - 2999
01668000 - Five-character, Codes 15000 - 19999
Error Code Listing:
Codes 100 - 29999 and IML Code I9990
Error Code Tables:
These have to do with the Initial Microcode Load on the 76/77, 85, 90, 95,
and 500 systems.(and the TP700/720!) see HERE.
General SCSI Error Codes
Device Size Code -----------------+ +--------- Unit Reference Code
Adapter MCA Slot ----------------+| |+-------- Sense Key Code
Logical Unit Number ---------------+|| || +------ Additional Sense Code
||| || |
Physical Unit Number --------------+||| || |
|||| || |
Device Code Number ---------+--+ |||| || |
| | |||| ||++
| | |||| ||||
0210 401H 9231
Basically the SCSI Error Codes consist out of an 8-character output like a
General Error Code, but have an additional 4-character Error Return Code at the
end, which helps to specify the nature of the error.
There are small differences between SCSI Device Errors and SCSI
Adapter Errors. These are explained when necessary.
SCSI Error Code Tables
Refer to the General and SCSI
Error Codes above.
If multiple errors occur during POST, resolve them in the order that
they are presented.
Note: The correct procedure for using
the Personal System/2 reference diskette is to power on the system with the
diskette inserted in the diskette drive. Soft booting the reference diskette
(i.e. Ctrl-Alt-Del) may cause false errors as well as a false indication that
a power-on password is already present when you try to set one.
Ed. I can't remember having issues
from a warm boot while working on a PS/2, -BUT- if the system refuses to apply
configuration changes, cold boot.
Always cold boot and run Advanced Diagnostics (go into System Programs and
at the main menu do a Ctrl-A) before replacing components when trying to
resolve software problems. If diags don't fail, replacing components will
probably not solve the problem. Refer to the software vendor for possible
patches. The software may not be supported on the system.
Multiple POST Error Procedure
If more than one error code is displayed, diagnose the first error code
first. The cause of the first error code could cause other devices to fail.
This is especially important with configuration errors.
If no error codes appear, see if the error symptoms are listed in the Error
If an adapter consists of multiple FRUs (memory for example) remove the
optional FRUs one at a time to see if symptoms change before replacing the
External surge suppressors may be the source for hard to diagnose problems.
Dead System (no POST, screen blank, no beep)
The most likely cause is a device that is shorting out the power supply. An
improperly inserted memory module, a defective adapter or device can cause a
short circuit. To prevent damage to a power supply, the system board must
present a "Power Good" signal to the power supply in 150 ms or less. If this
does not occur, then the power supply shuts down internally.
Known dufus tricks- SIMM inserted backwards (with enough force ANYTHING is
possible). adapter not fully seated in expansion slot, power cord not plugged
in, unsupported/defective adapter (Non-IBM adapter, like ALR..), or with a
heavily loaded system, too many drives starting at once (overcurrent as the
drives attempt to spin up- leave motor start jumper open to start drives after
the controller interrogates them).
Troubleshooting a Dead System
Verify that power is on (power cord plugged in?). If the system has power,
then go to step 2.
Note: Intermittent and very difficult
to diagnose system problems, may be caused by line cords which are not
fully seated, or are too loose to make a tight connection.
When troubleshooting intermittent post errors, or any unusual, system
problems, (for example; system performs power-on reset unexpectedly during
operation) check the line cord for proper seating. Slight forming of the male
contacts in the system unit power supply connector may correct the problem
-OR- you can sometimes adjust the female contacts in the line cord plug.
Warning! Remove line cord from outlet
before working on line cord contacts!
Replacing the line cord may be necessary in some cases. Both ends of the
line cord should be checked.
- Remove all adapters, options, extra memory, etc.
Remove all external connections, KB, mouse, display, etc.
- Plug in, power on, listen for a beep. The beep indicates POST has run.
Multiple beeps may occur.
A) If no beep is heard, verify continuity through speaker.
If OK, replace system board memory and retry # 3.
If still no beep, verify PSU voltages. If voltages are OK, replace system board.
If voltages are incorrect, replace the power supply.
Note: On systems with LogicLock, the
switches used to detect unauthorized access might be put in the wrong position
during case re-installation.
Unit does not
run with cover removed (9556/57 9576/77)
B) If a beep occurs, reinstall adapters one at a time and return to step 3.
When something is added and the beep is no longer heard, the last item plugged
in is probably defective. At this point, it is not necessary to reconfigure the
system each time an adapter or device is added because we expect any beep.
Defective or weak batteries can cause loss of all setup information. If only
part of the setup is lost, the battery is probably NOT the cause. Inaccurate
time is usually caused by software, but some of the older models with the 6V
batteries are more prone to time slowing down.
Clearing 96 8N1 (From Peter Wendt)
IBM wrote in the HMM "Remove the battery, then wait 5 minutes..." (HMM,
September 1993, P/N 71G9316, page 300, "96 8N1 Error Message") but empirical
research showed that this doesn't work fairly often.
Shortening the battery connectors (with the battery removed of course) *and*
toggling the startup password jumper seems to be the only fast cure. There
seems to be a board logic, that ANDs the two conditions after a power-on and
deletes the entire setup from the CMOS.
At least during the "hot phase" when the Mod. 90s and 95s can be found in
larger amounts at the customers this procedure was the only one that worked in
a sufficiently fast way. And I had quite a lot 95s under service... (Some are
That worked any time I'd landed on the "ASCII-console mode" during fiddling
with SOD- / bad-DMA platforms and Kingston Turbochips :-) Moving only the
jumper or only isolating the battery often did *not* cure it. Which is slightly
Info about the password jumper HERE.