13/07/2021 minuszerodegrees.

net

  Home 

IBM 5170  -  POST  -  Some on-screen errors

IMPORTANT:  The following is specific to the IBM 5170 (IBM AT).

  Certain portions will not be applicable to clones that were made of the IBM 5170.

NOTE:  Execution of tests is not always done in numerical order.

  For example, the hard drive tests are done in order of 1782, 1780, 1781, 1790, 1791

Error As shown on screen      Description

101-System Board
  101 IBM describes this as 'INTERRUPT FAILURE'.

Error

102-System Board
  102 IBM describes this as 'TIMER FAILURE'.

Error

103-System Board
  103 IBM describes this as 'TIMER INTERRUPT FAILURE'.

Error

104-System Board
  104 IBM describes this as 'PROTECTED MODE FAILURE'.

Error

105-System Board
  105 IBM describes this as 'LAST 8042 COMMAND NOT ACCEPTED'.

Error

106-System Board
  106 IBM describes this as 'CONVERTING LOGIC TEST'.

Error

107-System Board
  107 IBM describes this as 'HOT NMI TEST'.

Error

108-System Board
  108 IBM describes this as 'TIMER BUS TEST'.

Error

109-System Board
  109 IBM describes this as 'LOW MEG CHIP SELECT TEST'.

Error

IBM describes this as 'DEAD BATTERY'.

Note that simply fitting a good battery is alone, not enough to remove the 161
error.  One also needs to then complete the SETUP procedure.  The combination of
those two actions should remove the 161 error.

161-System Options
  161
Not Set-(Run SETUP)
Some possible causes:

• Battery is dead/low - replace it and then run the SETUP procedure.

• A dead/low battery was replaced with new but the SETUP procedure has yet to
be done.

• Faulty motherboard.

  162 162-System Options

Not Set-(Run SETUP) IBM describes this as 'CMOS CHECKSUM ERROR', but that is only a partial
description.

Either, a SETUP configuration (CMOS SETUP) checksum problem, or some other

configuration issue.

Some possible causes:

www.minuszerodegrees.net/5170/post_errors/5170_post_errors.htm 1/6
13/07/2021 minuszerodegrees.net

• A hiccup of some kind has corrupted SETUP configuration (CMOS SETUP).

 Correct this by running the SETUP procedure.

• Battery is dead/low.  Some 5170 motherboards, for unknown reason, don't

produce a 161 error.  Check/replace battery, and then run the SETUP procedure.

• A CGA video card is fitted, but the SETUP configuration (CMOS SETUP) is set for
MDA (monochrome), or 'special' (EGA/VGA).

• An MDA ( monochrome) video card is fitted, but the SETUP configuration (CMOS
SETUP) is set for CGA, or 'special' (EGA/VGA).
 

IBM describes this as 'CLOCK NOT UPDATING', but that is only a partial
description.

Either, there are invalid clock (date/time) values in the CMOS/RTC chip, or the
clock in the CMOS/RTC chip is not ticking over (not advancing).

163-Time & Date Not

  163
Set-(Run SETUP) Some possible causes:

• A hiccup of some kind has corrupted the clock values.  Correct this by running
the SETUP procedure.

• The SETUP procedure was started but not finished.

• Motherboard failure (in CMOS/RTC related circuitry).

The amount of base RAM or expansion (extended) RAM discovered by the POST is
different than the amounts recorded in SETUP configuration (CMOS SETUP).

Some possible causes:

164-Memory Size
  164 • User has added RAM but failed to then perform the SETUP procedure.

Error-(Run SETUP)
• User ran the SETUP procedure, and during that procedure, entered incorrect
amounts for the RAM.

• Faulty RAM (faulty in a way that unfortunately, leads the POST to believe that
there is less RAM than fitted).

RAM related problem.

Displayed on screen in the format of "AAAAAA  BBBB  201-Memory Error",

where AAAAAA is the hexidecimal address (in bytes) of the failing word and BBBB
shows which bits in the word have failed.

You can use the diagram here to calculate the failing bit/s, and the address as a
KB figure.

  201 201-Memory Error

Sometimes, you may see an address that is one word beyond the actual failing
address, e.g.  080002 instead of 080000

NOTE: If the BBBB portion (failing bits) contains many "F", example: "FFFE",
then it is highly likely that non-existing RAM is being tested.

For example (one of many), this would occur if the machine SETUP is
configured for 640K of RAM, but only 512K is fitted.

For example (one of many), a RAM card has failed completely.

202-Memory Address Memory related problem concerning address bits 0 to 15.

  202
Error IBM's description is at here.

203-Memory Address Memory related problem concerning address bits 16 to 23.

  203
Error IBM's description is at here.

  301 301-Keyboard Error

Keyboard problem.

If the error is preceeded by a byte, e.g.  "21 301-Keyboard Error", that indicates
that a key is stuck down on the keyboard.  The byte indicates which key.

www.minuszerodegrees.net/5170/post_errors/5170_post_errors.htm 2/6
13/07/2021 minuszerodegrees.net

If instead, the error is just "301-Keyboard Error", that indicates a general

keyboard error.

Some possible causes:

• No keyboard.

• Wrong type of keyboard - AT class keyboard required - PC/XT class keyboards

will not work.

• Faulty keyboard.

• Faulty motherboard (specifically the keyboard circuitry on it).

• In certain AT clones, a keyboard error can be caused by the front panel

keyswitch being in the 'locked' position.

• In certain AT clones, a keyboard error can be caused by a blown keyboard fuse

on the motherboard.  (Note: The IBM 5170 does not have a keyboard fuse.)

The POST interrogated the keyboard controller chip on the motherboard to get the
status of the 5170's keylock switch (on front panel).

It was indicated that the keylock is in the locked position.

Some possible causes:

• Keylock is in the locked position.

302-System Unit • Keylock is in the unlocked position, but the keylock switch is faulty  (faulty in a
  302
Keylock is Locked way that has closed the contacts on motherboard connector J20).

• Partially faulty keyboard controller chip on motherboard  (faulty in a way that

results in a read of its 'P17' pin being reported incorrectly).

To rule out the keylock, simply remove the plug on motherboard connector J20, as
shown removed in the right-hand photo at here.  If the 302 error no longer
appears at power on of the 5170, it means that the keylock is the 302 cause.

Power off the IBM 5170, then remove the keyboard, then power the IBM 5170
back on.

303-Keyboard Or If the 303 error has gone, it means that the keyboard is faulty (faulty in a
  303
System Unit Error particular way).

If the 303 error remains, it means that the motherboard is faulty (faulty in a
particular way).

IBM describes this as 'KEYBOARD CLOCK LINE HIGH'.

Power off the IBM 5170, then remove the keyboard, then power the IBM 5170
304-Keyboard Or back on.

  304
System Unit Error If the 304 error has gone, it means that the keyboard is faulty (faulty in a
particular way).

If the 304 error remains, it means that the motherboard is faulty (faulty in a
particular way).

A known cause: Having only a CGA video card fitted, but motherboard switch SW1
  401 401-CRT Error
is incorrectly set to the MDA position (rear).

A known cause: Having only an MDA (monochrome) video card fitted, but
  501 501-CRT Error
motherboard switch SW1 is incorrectly set to the 'CGA' position (front).

  601 601-Diskette error

Some possible causes:

• POST can't find the floppy portion of the IBM Fixed Disk and Diskette Drive
Adapter

• Wrong type of twisted cable is being used - hard drive type instead of floppy
drive type.  (photo of difference)

• Floppy drive A: - faulty

• Floppy drive A: - does not have power to it (either +5V or +12V, or both).

• Floppy drive A: - jumpered incorrectly for the 5170

• Floppy drive A: - a make-model that is incompatible with the 5170

  

www.minuszerodegrees.net/5170/post_errors/5170_post_errors.htm 3/6
13/07/2021 minuszerodegrees.net

• BIOS issue:

Are you seeing the 601 after removing the IBM Fixed Disk and Diskette Drive
Adapter and substituting it with something else ?

There is something in the second and third BIOS revisions of the 5170 that
expects the IBM Fixed Disk and Diskette Drive Adapter.

The first BIOS revision does not have the issue.

602-Diskette Boot
  602 IBM describes this as 'DISKETTE BOOT RECORD IS NOT VALID'.

Record Error

The 5170's POST, upon seeing that a hard drive is included in the SETUP
configuration (CMOS SETUP), asked the hard disk controller to perform an 'internal
diagnostic' (see 1782 error).  That passed.  The POST then asked the controller to
issue a 'Recalibrate drive' command to the first hard drive (C:).  The 1780 error
was displayed because that command failed.

Some possible causes:

• Control cable (34 pin) is missing.

• Control cable (34 pin) is no longer seated correctly on controller or drive.

• Control cable (34 pin) is faulty.

• On the controller, control cable (34 pin) accidentally connected to J1 (for floppy
drives) instead of J5 (for hard drives).  See diagrams at here.

• On the controller, control cable (34 pin) is connected to J5 upside down.  See
1780 1780-Disk 0 Failure
diagrams at here.

• Drive Select jumper on hard drive is in wrong position for the type of control
cable being used.  See diagrams at here.

• A twisted control cable is being used but is the wrong type - floppy drive type
instead of hard drive type.  (photo of difference)

• Hard drive does not have power to it (either +5V or +12V, or both).

• Hard drive is faulty.

• Hard drive controller card is faulty (even though it earlier passed its 'internal
diagnostic' [test 1782]).

• Hard drive controller card is in an 8-bit slot (example: 5170 slot 7) instead of a
16-bit slot.

• Hard drive controller card has 'dirty' edge connectors (ones that affect this test).

1781 1781-Disk 1 Failure Per 1780 error, but for second hard drive (D:)

 
The 5170's POST, upon seeing that a hard drive is included in the SETUP
configuration (CMOS SETUP), asked the hard disk controller to perform an 'internal
diagnostic'.  The 1782 error was displayed because, either the controller
responded in the negative, or the controller did not respond within a minute.

Some possible causes:

• Hard drive controller card missing.

1782-Disk Controller
1782 • Hard drive controller card is not seated properly in its slot.

Failure
• Hard drive controller card has 'dirty' edge connectors (ones that affect this test).

• Wrong type of hard drive controller card is fitted (example: XT class controller is
fitted).

• Faulty hard drive controller card.

• Hard drive controller card is incorrectly jumpered (example: set for secondary
address range, not primary).

1790 1790-Disk 0 Failure The 5170's POST, upon seeing that a hard drive is included in the SETUP
configuration (CMOS SETUP), asked the hard disk controller to perform an 'internal
diagnostic' (see 1782 error).  That passed.  The POST then asked the controller to
issue a 'Recalibrate drive' command to the first hard drive (see 1780 error).

That passed.  The POST then asked the controller to read certain sectors on the
first hard drive.  The 1790 error was displayed because one of those reads failed.

Some possible causes:

• Data cable (20 pin) is missing.

• Data cable (20 pin) is no longer seated correctly on controller or drive.

• Data cable (20 pin) is faulty.

• On the controller, data cable accidentally connected to J3 (for D:) instead of J4

www.minuszerodegrees.net/5170/post_errors/5170_post_errors.htm 4/6
13/07/2021 minuszerodegrees.net

(for C:).  See diagrams at here.

• On the controller, data cable is connected to J4 upside down.  See diagrams at

here.

• A new hard drive was fitted but its low-level format does not match that of the
controller.  Low level format the drive after fitment.      <----- Common, if new
drive fitted

• Hard drive has fewer cylinders or heads than the drive C: type in the SETUP
configuration (CMOS SETUP) indicates.

• Hard drive C: is faulty.

• Faulty hard drive controller card (specifically the drive data processing area).
• Hard drive C: is getting +12 volts, but not the hard drive controller card.  See
note 1.

• The hard drive controller card is not getting -12 volts.  See notes 2 and 3.

1791 1791-Disk 1 Failure Per 1790 error, but for second hard drive (D:)

Intermittent and unlike below, there is no '1' nor '2' following 'PARITY CHECK'.

A virus is known to display this.

  PARITY CHECK
(The IBM BIOS in an IBM 5170 always displays a 1 or a 2 after PARITY CHECK.)

A parity error occurred when an address in motherboard RAM was read.

PARITY CHECK 1

 
xxxxx 'xxxxx' will either be the address where the parity error occured, or it will be five
question marks if the address could not be determined.

A parity error occurred when an address in expansion card hosted RAM was
read.

PARITY CHECK 2
'xxxxx' will either be the address where the parity error occured, or it will be five
 
xxxxx question marks if the address could not be determined.

PARITY CHECK 2 with five question marks is also known to occur in the specific
situation stated at here.

Note 1 Maybe the +12V connection to the motherboard is bad.  Try remove/refit connector P8.
   
Note 2 Check -12 volts out of power supply.

This check should not be required if you have the genuine 5170 power supply, because the corresponding
lack of a POWER GOOD signal would not have allowed the motherboard to start.

   
Note 3 Maybe the -12V connection to the motherboard is bad.  Try remove/refit connector P8.

www.minuszerodegrees.net/5170/post_errors/5170_post_errors.htm 5/6
13/07/2021 minuszerodegrees.net

www.minuszerodegrees.net/5170/post_errors/5170_post_errors.htm 6/6