Jump to content

IBM PS/2 Model 30 Troubleshooting Guide: Difference between revisions

From RetroTechCollection

Josh (talk | contribs)

5 user groups
10,877 edits
undo
VisualWikitext
Create comprehensive IBM PS/2 Model 30 / Model 50-60 / Model 70-80 page set with full PS/2 POST code reference, MCA architecture detail, capacitor lists, ECA recalls, and SMD electrolyte leakage diagnosis per RTC style guide
ย 
Rebuild against Ardent Tool's 8530 planar and power pages: 'the Model 30 8086 has no hardware RTC' is wrong (MM58167AN at U16 with a soldered 3 V cell); delete the unsourced 'Model 55 PSU is a drop-in replacement' claim, the 3 A keyboard fuse rating and code 2408; add IBM's own four service tips (H043664 cursor keys / 501 errors and the U9/U15 'Japan' marking check, H067282 bus adapter battery voltage where IBM's own HMS prints the wrong pins, H007157 the fixed disk / board / diagnostics comb...
ย 
(2 intermediate revisions by the same user not shown)
Line 1: Line 1:
This guide covers diagnostic procedures and the complete IBM PS/2 numeric POST error code reference applicable to the '''[[IBM PS/2 Model 30]]''' (machine type 8530 — Model 30 8086 and Model 30 286). The Model 30 uses the same PS/2 BIOS POST architecture as the rest of the PS/2 line, sharing audio beep codes and Major/Minor numeric error codes with the [[IBM PS/2 Model 25 Troubleshooting Guide|Model 25]]. This guide focuses on faults specific to the Model 30 (ISA-bus, proprietary ST-506 HDD, MCGA / VGA distinction).
<templatestyles src="Template:StyledTable/styles.css" />
[[File:IBM PS-2 Model 30 (photo).jpg|thumb|right|300px|IBM PS/2 Model 30. Source: Wikimedia Commons.]]


== Audio Beep Codes ==
Fault diagnosis for the '''[[IBM PS/2 Model 30]]''' (machine type 8530) โ€” an '''ISA''' desktop, in an 8086 form and a 286 form that share almost nothing electrically.


{| class="wikitable styled-table" style="width:100%; text-align:center;"
== Which machine do you have ==
|+'''Model 30 power-on audio codes'''
ย 
! Pattern !! Meaning
{| class="wikitable styled-table" style="width:100%; text-align:left;"
|-
|+'''The two Model 30 planars'''
| 1 short || POST OK
! Variant !! Processor !! RTC and battery !! Notes
|-
| 2 short || Numeric error displayed; read screen
|-
| 1 long + 1 short || Display or display adapter problem (MCGA / VGA)
|-
| 1 long + 2 short || Display or display adapter problem (VGA)
|-
| 1 long + 3 short || EGA / display adapter problem
|-
| Continuous short beeps || Keyboard problem &mdash; check fuse, cable, 8042 controller
|-
|-
| Continuous solid beep || System board failure
| '''8530-001 / -002 / -021''' (8086) || Intel 8086-2 at 8 MHz (U10) || '''NS MM58167AN RTC at U16''' with a 32.768 kHz crystal at Y4, and a '''3 V battery soldered to the board''' โ€” IBM assigns it no FRU || "P-PLANAR" P/N 61X8823. MCGA video: 72X8300 memory controller at U11, 72X8205 video formatter at U12, IMSG171S RAMDAC at U5, 8 KB SRAM at U38. F1 and F2 are the mouse and keyboard fuses<ref name="ard_30p1" /><ref name="ard_batt" />
|-
|-
| Click then no beep || PSU started but Power Good not asserted in 150 ms &mdash; suspect shorted tantalum
| '''8530-E01 / E0R / E21 / E2R / E31''' (286) || 286 at 10 MHz (ZM36) || '''Dallas DS1287 RTC/CMOS module at ZM35''', IBM FRU 8509237 โ€” the whole module is the replaceable part || VLSI VL82C1xx chipset. VGA controller '''15F6864''' at ZM18, 8042 or 8742 keyboard controller at ZM33, ยตPD765AC-2 FDC at ZM19, F1 the fuse, '''J13 the password override'''. The PLCC-CPU planar is FRU 34F0046 (37F6654) / 33F8700<ref name="ard_30p286" /><ref name="ard_batt" />
|-
| No display and no sound || PSU fold-back protection, shorted planar capacitor, or PSU dead
|}
|}


== Numeric POST Error Codes ==
'''The 8086 Model 30 does have an RTC.''' An earlier version of this page said it did not, and therefore could not report CMOS or clock errors. IBM's own planar list puts a National Semiconductor MM58167AN at U16 with its own crystal and a soldered 3 V cell.<ref name="ard_30p1" /><ref name="ard_batt" />


The Model 30 uses the standard PS/2 numeric error format. Codes specific to the MCA subsystem (which the Model 30 does not have) are absent &mdash; the Model 30 only triggers the ISA-class codes (1xx through 24xx in the main range, plus 86xx for the pointing device).
'''15F6864 is a VGA controller, and it is on the 286 planar''' โ€” not on the 8086 planar, which has no VGA at all. It is also the VGA controller on the 8525-286 and the 8570 Type 1 planar.<ref name="ard_30p286" />
== Power supply โ€” IBM's own figures ==


=== 1xx &mdash; System Board ===
The Model 25 and Model 30 families are '''ISA''' machines, and IBM serviced them under the Personal Computer maintenance library rather than the PS/2 one. That is why their supplies document '''four''' DC rails, including โˆ’5 V, where the Micro Channel machines document three.<ref name="ard_2530pwr" />


{| class="wikitable styled-table" style="width:100%; text-align:center;"
'''The supply is manually switchable''' and must be set to either the 100โ€“125 V ac or the 200โ€“240 V ac range '''before''' the power cord goes into an outlet.<ref name="ard_2530pwr" />
|+'''Major code 1xx &mdash; system board'''
ย 
! Code !! Meaning
=== Minimum under-voltage sense levels ===
|-
ย 
| 101 || Interrupt failure / processor error
When the machine has been off for five seconds or more and is then switched on, the supply generates a '''Power Good''' signal that resets the system logic. Its presence means the supply is working and that the minimum under-voltage sense levels have been established:<ref name="ard_2530pwr" />
|-
ย 
| 102 || Timer 0 failure or BIOS ROM checksum error &mdash; reseat the ROM if socketed
{| class="wikitable styled-table" style="width:70%; text-align:left;"
|-
|+'''Under-voltage sense levels'''
| 103 || Timer interrupt failure
! Output !! Minimum under-voltage sense level
|-
| 104 || 8259A controller error (Model 30); protected mode failure (Model 30 286)
|-
| 105 || Last 8042 keyboard controller command not accepted
|-
| 106 || Converting logic test (planar)
|-
| 107 || NMI test failure
|-
| 108 || Timer bus test failure
|-
| 109 || DMA (8237A) test error
|-
| 110 || Planar parity error
|-
| 111 || I/O parity error
|-
|-
| 121 || Unexpected hardware interrupt
| +5.0 V || '''+4.5 V'''
|-
|-
| 151 || Bad battery (Model 30 286 only)
| โˆ’5.0 V || '''โˆ’4.3 V'''
|-
|-
| 152 || CMOS error / power-on password installed
| +12.0 V || '''+10.8 V'''
|-
|-
| 161 || CMOS configuration empty &mdash; '''Model 30 286: replace battery'''
| โˆ’12.0 V || '''โˆ’10.2 V'''
|-
| 162 || CMOS checksum mismatch &mdash; run SETUP from Reference Diskette
|-
| 163 || Time and date not set
|-
| 164 || Memory size error &mdash; run SETUP
|-
| 199 || User-indicated configuration not correct
|}
|}


==== Model 30 286 161 / 162 / 163 cluster ====
'''Power Good itself should read +3.5 to +5 V with the machine running correctly.''' If it does not, remove all adapters, cables and extra memory and try again; if Power Good is then valid, a device was shorting. Refit them one at a time.<ref name="ard_2530pwr" />


This cluster on a Model 30 286 indicates a dead Dallas DS1287 / DS12887 RTC battery. Replace per [[IBM PS/2 Model 30 Maintenance Guide]] and re-run SETUP.
This is more useful than a general rail check, because it tells you what the supply itself is deciding: a rail that reads plausibly on a meter but is below its sense level will not let POST start at all.


The Model 30 8086 does '''not''' report 161 / 162 / 163 because it has no hardware RTC.
=== Where to put the probes ===


=== 2xx &mdash; Memory ===
IBM's own connector measurements, with pin 1 towards the rear of the machine:<ref name="ard_2530pwr" />


{| class="wikitable styled-table" style="width:100%; text-align:center;"
{| class="wikitable styled-table" style="width:90%; text-align:left;"
|+'''Major code 2xx &mdash; memory'''
|+'''Models 25, 25 SX and 25-286 โ€” connector J7'''
! Code !! Meaning
! โˆ’ lead !! + lead !! V dc min !! V dc max
|-
|-
| 201 || Memory test failed &mdash; data miscompare or parity error
| P7-5 || P7-10 || +4.8 || +5.2
|-
|-
| 202 || Address line 0&ndash;15 failure
| P7-9 || P7-5 || +4.5 || +5.4
|-
|-
| 203 || Address line 16&ndash;23 failure (Model 30 286 only)
| P7-5 || P7-3 || +11.5 || +12.6
|-
|-
| 207 || ROM failure
| P7-4 || P7-8 || +10.8 || +12.9
|-
| 211 || Base 64 K on I/O channel failed
|-
| 215 / 216 || Base 64 K on daughter card failed
|-
| 225 || Wrong / defective / slow memory
|}
|}


The Model 30 8086 has 640 KB on planar; the failing bank cannot be in expansion (there is no expansion bank). A 201 error therefore identifies the failing chip on the planar by bit pattern. The Model 30 286 has two MMK SIMM sockets, and the failing bank is identified by the leading hex digits of the 201 code.
{| class="wikitable styled-table" style="width:90%; text-align:left;"
ย 
|+'''Models 30 and 30-286 โ€” connectors P3 and P4'''
=== 3xx &mdash; Keyboard ===
! โˆ’ lead !! + lead !! V dc min !! V dc max
ย 
{| class="wikitable styled-table" style="width:100%; text-align:center;"
|+'''Major code 3xx &mdash; keyboard'''
! Code !! Meaning
|-
| 301 || Keyboard did not respond / stuck key. The XX value preceding "301" is the scan code of the stuck key. Also occurs if keyboard plugged into mouse port
|-
|-
| 302 || User-indicated keyboard test failure
| P3-5 || P4-4 || +4.8 || +5.2
|-
|-
| 303 || Keyboard or 8042 controller error
| P4-3 || P3-6 || +4.6 || +5.5
|-
|-
| 304 || Planar (8042) error / CMOS does not match keyboard
| P4-1 || P3-3 || +11.5 || +12.6
|-
|-
| 305 || +5 V fuse blown on planar (3 A) &mdash; replace fuse and inspect keyboard / mouse for shorted cable
| P3-4 || P4-2 || +10.8 || +13.2
|}
|}


==== 301 with X-scan-code ====
Note that these are '''pairs of pins''', not a rail measured against chassis ground โ€” IBM specifies both leads for each reading.
== Dead system ==


If the Model 30 reports "XX 301" at POST and the XX is a hex scan code, an object resting on a key is the most likely cause. If multiple keys appear stuck, also see 8602. On the Model 25 / Model 30 family, the keyboard and mouse ports are interchangeable at the 8042; plugging a mouse into the keyboard port (or vice versa) raises a 301 + 8603 pair.
IBM's stated most likely cause of no POST, a blank screen and no beep is '''a device shorting out the power supply'''. '''The system board must present Power Good within 150 ms or the supply shuts itself down internally''' โ€” which is the mechanism behind the "click and then nothing" symptom.<ref name="ard_err" />


=== 4xx &mdash; Monochrome / MCGA display ===
# '''Check the line cord at both ends.''' IBM singles this out: intermittent faults, including unexpected power-on resets during operation, come from line cords that are not fully seated or too loose to make a tight connection. Remove the cord from the outlet before working on its contacts.<ref name="ard_err" />
# '''Measure Power Good''' โ€” it should be +3.5 to +5 V. If it is not, remove all adapters, cables and extra memory and retry.<ref name="ard_2530pwr" />
# Remove all external connections.
# Power on and listen for a beep.
# '''No beep:''' check speaker continuity; if good, replace the planar memory and retry; if still silent, measure the rails against IBM's figures above โ€” correct rails mean the planar, wrong rails mean the supply.
# '''A beep:''' refit adapters and memory one at a time. When the beep stops, the last item fitted is the suspect.


* 401 &mdash; MCGA / monochrome adapter test failure (planar VGA on Model 30 286 also raises 4xx on certain error paths).
IBM's own list of the usual causes includes '''a SIMM inserted backwards''' โ€” which has genuinely been seen in service โ€” an adapter not fully seated, the power cord not plugged in, an unsupported or defective adapter, and, on a loaded machine, too many drives spinning up at once.<ref name="ard_err" />
ย 
== Reference Diskette and Advanced Diagnostics ==
=== 5xx &mdash; Colour / MCGA / VGA ===


* 501 &mdash; CGA / MCGA / VGA colour test failure.
Neither machine has a built-in SETUP screen; configuration and diagnostics come from the Starter Diskette and the Advanced Diagnostics diskette.


=== 6xx &mdash; Diskette ===
* '''Always cold-boot''' โ€” insert the diskette before powering on. Soft-booting it can report false errors '''and''' falsely indicate that a power-on password is already set.<ref name="ard_err" />
* '''Ctrl-A''' from the main menu enters Advanced Diagnostics, which produces the minor codes.
* '''If several errors appear, resolve them in the order presented''' โ€” the first can cause the others.<ref name="ard_err" />
* '''The diskettes are machine-specific.''' A Model 30 diskette is not interchangeable with a Model 25 one, and the 8086, 286 and SX variants each have their own.
* '''Run Advanced Diagnostics before replacing anything.''' If the diagnostics do not fail, replacing parts probably will not help.<ref name="ard_err" />
== Battery ==


{| class="wikitable styled-table" style="width:100%; text-align:center;"
{| class="wikitable styled-table" style="width:85%; text-align:left;"
|+'''Major code 6xx &mdash; diskette'''
|+'''What this machine takes'''
! Code !! Meaning
! Variant !! Battery !! IBM FRU
|-
| 601 || Drive or controller test failed
|-
| 602 || Boot record on diskette invalid
|-
| 606 || Verify function failed
|-
|-
| 607 || Diskette write-protected
| 8530 (8086) || '''3 V battery, soldered to the board''' || None assigned
|-
|-
| 610 || Initialisation failed
| 8530-286 || '''Dallas RTC/CMOS module''' (DS1287) โ€” replace the whole module || 8509237
|-
| 611 || Timeout
|-
| 612 || Bad NEC FDC status
|-
| 613 || Bad DMA
|-
| 621&ndash;626 || Seek / CRC / address mark / record / data compare failures
|-
| 650 || Drive speed error
|-
| 651&ndash;656 || Format / verify / read / write / controller / drive failures
|-
| 657&ndash;660 || Write-protect or change-line signal stuck
|}
|}


The Model 30 with the proprietary ST-506 hard disk reports 6xx codes only for the floppy. 17xx codes cover the hard disk.
For the 286: the DS1287 is obsolete and is replaced by the pin-compatible '''DS12887''', or the lead-free '''DS12887+'''. The 'A'-suffix parts are fine too. '''Do not use the DS12C887''' โ€” it is the Year-2000-compliant variant but uses the century byte differently and will not work in a PS/2. The module can also be reworked with an external cell instead of replaced.<ref name="ard_batt" />


=== 7xx &mdash; Math coprocessor ===
'''Defective or weak batteries cause the loss of all setup information. If only part of the setup is lost, the battery is probably NOT the cause.'''<ref name="ard_batt" />


* 7xx &mdash; 8087 (Model 30 8086) or 80287 (Model 30 286) math coprocessor error. The 80286 BIOS does not test the 80287 unless software invokes it.
== IBM service tips for this machine ==


=== 9xx / 10xx &mdash; Parallel port ===
Four IBM tips survive for the Model 30 and all four are worth knowing before you start swapping parts.<ref name="ard_30p1" />


Same code structure as PC/AT and Model 25.
=== Cursor keys and 501 POST errors (H043664) ===


=== 11xx / 12xx &mdash; Serial port ===
'''On a limited number of 8530-001 and -021 machines''', using the cursor movement keys makes unexpected characters appear on the display, and a '''501''' POST error appears once two or more option cards are fitted (the 3270 and Token Ring half-size adapters among them). The problem may be intermittent and '''may not be detected by Advanced Diagnostics''' โ€” so a clean diagnostics run does not clear it.


Same code structure as PC/AT and Model 25.
'''The check is visual, and it is a printed-marking check, not a part number check.''' Inspect the planar modules at:


=== 17xx &mdash; Fixed disk ===
* '''U15''' โ€” next to the speaker. P/N 61X8965.
* '''U9''' โ€” next to the math coprocessor socket. P/N 72X8203.


{| class="wikitable styled-table" style="width:100%; text-align:center;"
'''Good chips have the word "Japan" printed on them. Affected chips do not.''' If either module lacks it, the fix is a new system board, FRU P/N '''64F3453'''. Machines with a serial above '''55-00HV000''' left the factory with the new board. If a math coprocessor is fitted, move it across to the new board.
|+'''Major code 17xx &mdash; fixed disk'''
ย 
! Code !! Meaning
=== Bus adapter battery voltage (H067282) ===
|-
ย 
| 1701 || Controller POST error
'''Early releases of the 8530 HMS print the wrong pin numbers.''' On the system board maps, page 0100-4, the pins labelled "0" and "39" in figure one should be "1" and "40". '''The correct measurement is at pins 38 and 39''' โ€” the second and third pins from the right edge. '''Replace the bus adapter card if the voltage between the correct pins is less than 2.5 V.'''
|-
ย 
| 1702 || Controller timeout
This is exactly the class of error this wiki exists to catch, and it is IBM's own.
ย 
=== Fixed disk, system board and diagnostics must match (H007157) ===
ย 
Successful fixed-disk operation on the 8530-001 and -021 depends on the right combination of system board, hard disk and Advanced Diagnostics version. '''Formatting with the wrong level of Advanced Diagnostics causes fixed disk errors.'''
ย 
{| class="wikitable styled-table" style="width:100%; text-align:left;"
|+'''IBM's required combinations'''
! Machine !! System board !! Fixed disk !! Diagnostics
|-
|-
| 1703 || Seek failure
| 8530-021 below serial #2500000 || 61X8907 || 61X8929 || 8530 Advanced Diagnostics, any version
|-
|-
| 1704 || Adapter or drive error
| 8530-021 above serial #2500000 || 33F8474 || 6128285 || 8530 Advanced Diagnostics '''v1.05'''
|-
|-
| 1780 || Drive 0 timeout at IPL
| 8530-001, all || 33F8474 || 6128285 or 6128279 || 8530 Advanced Diagnostics '''v1.05'''
|-
|-
| 1781 || Drive 1 timeout at IPL
| 8530-E21 only || 27F4069 || 6128285 || 8530-286 Advanced Diagnostics '''v1.02'''
|-
|-
| 1790 / 1791 || Format not recognised / data cable off
| 8530-E31 and E01 || 34F0046 || 6128285 or 6128279 || 8530-286 Advanced Diagnostics '''v1.02'''
|}
|}


On the Model 30 8086 with the famously failure-prone ST-506 drive, 1701 / 1780 with the drive present is most commonly drive stiction or drive failure. Gentle rotation of the drive case (powered off) may free a stuck spindle for one boot; replace the drive at the earliest opportunity.
'''Fixed disks 6128285 and 6128279 must not be fitted to any 8530-021 below serial number 2500000''', and machines above that serial, plus all -001 machines, must be formatted only with v1.05.
ย 
=== 8515 and 8516 displays are not supported on the 8086 machines (H094152) ===
ย 
An 8086 Model 30 (-001, -002, -021) connected to an '''8515''' or '''8516''' display shows compressed and distorted video. '''That is not a fault.''' The supported displays are:
ย 
* '''8086 machines (-001, -002, -021):''' 8503, 8504, 8507, 8512, 8513, 8514.
* '''286 machines (E01, E21, E31, E41):''' 8503, 8504, 8512, 8513, 8514, 8515, 8516.


=== 24xx &mdash; Integrated VGA (Model 30 286) ===
=== ECA 023 ===


* 2401 &mdash; Planar VGA error.
The Model 30 also has an ECA 023, "Erroneous/False Display Characters". Read it alongside H043664 above, which describes the same class of symptom.<ref name="ard_30p1" />
* 2402 &mdash; Diagnostic video error.
* 2408 &mdash; VGA ground circuit fault.


=== 86xx &mdash; Pointing device ===
== Keyboard and 301 errors ==


* 8601 / 8603 / 8604 &mdash; planar (8042) or pointing-device error. Most commonly a mouse plugged into the keyboard port.
An '''XX 301''' where XX is a hex scan code usually means something is resting on a key. IBM's FRU order for 301 is the keyboard cable, then the system board.<ref name="ard_err" />


== Reference Diskette and Advanced Diagnostics ==
For '''305''', IBM's order is the keyboard port fuse, the keyboard cable, then the system board โ€” '''and if there is no fuse in the system, replace the system board.''' No rating is published for the fuse and none is given here.<ref name="ard_err" />
ย 
The keyboard and mouse sockets are identical Mini-DIN-6 connectors wired differently; swapping them produces a 301 and a pointing-device error together.
ย 
== Power-on password ==
ย 
'''J13''' is the password-override connector on the 286 planar. Jumper it momentarily during power-on and the BIOS clears the password.<ref name="ard_30p286" />


The Model 30 8086 and the Model 30 286 use '''different''' Reference Diskettes. Always cold-boot the Reference Diskette (insert before powering on) &mdash; warm-booting may report false errors. From the main menu, '''Ctrl-A''' enters Advanced Diagnostics which produces minor (NN) codes giving further FRU-level detail.
== Hard disk ==


The Reference Diskette is also where SETUP is performed; the Model 30 BIOS has no built-in setup screen.
* '''17xx with the drive fitted''' โ€” often stiction on the original ST-506. Rotate the drive gently around its spindle axis with the power off, refit and retry; then image the drive and replace it.
* '''Before anything else on a -001 or -021, check the H007157 combination table above.''' A machine that has been "repaired" with a mismatched board, drive or diagnostics version will produce disk errors that no amount of further part-swapping will fix.
* IBM's FRU order for the general 17xx group ends with '''the power supply''', so a sagging supply is a documented cause of disk errors.<ref name="ard_err" />


== Dead System Isolation ==
== Diskette drive ==


If the Model 30 shows no signs of life:
'''6xx codes are mostly media errors''' in IBM's own reading: try a known-good diskette, and if the error returns, replace the drive. Reseat the single-cable connector and check pin 1 orientation. A continuous activity light with no read means the cable is on backwards โ€” power off at once.<ref name="ard_err" />


# Verify the wall outlet and mains lead.
== Display ==
# Open the system unit.
# Verify the PSU output rails at the planar power connector with a multimeter (+5 V, +12 V, &minus;5 V, &minus;12 V).
# If rails are absent, suspect the PSU. The Model 55 PSU is a drop-in replacement.
# If rails are present but the system is dead, remove all ISA cards and the optional hard drive.
# Power on with planar, riser and (Model 30 286) one SIMM only. Listen for a beep.
# If no beep, suspect bank-0 memory on planar (Model 30 8086 has integrated DRAM &mdash; suspect the DRAM chips at U-positions on the planar) or the SIMM (Model 30 286).
# Re-add cards one at a time until the failing card is identified.


The PS/2 PSU asserts Power Good within 150 ms or shuts down internally. A shorted tantalum on the planar can pull the PSU into fold-back protection and produce a "click then nothing" symptom.
IBM's discriminator for the 24xx codes is whether the screen colours change: if they do, suspect the display; if they do not, suspect the system board. The general display-adapter family is 0005 XXXX.<ref name="ard_err" />


== Common Symptoms ==
== What has been removed from this page ==


* '''Continuous beep, no display''' &mdash; shorted tantalum or keyboard fuse blown.
* '''The hand-transcribed numeric POST tables''', replaced by the single reference on the [[IBM PS/2 Model 50 Troubleshooting Guide#Numeric POST error codes|Model 50 guide]]. They gave PC/AT-era meanings as PS/2 codes โ€” 104 as an 8259A error or a protected-mode failure, 107 as an NMI test, 108 as a timer bus test, 109 as an 8237A DMA test, 121 as an unexpected hardware interrupt, 151 as a bad battery and 162 as a checksum mismatch. IBM's protected-mode code is 115; 121 is a memory-remapping note and is information only; there is no 151; and 162 is a devices-powered-on check, with 173 being the checksum code.
* '''161 / 162 / 163 cluster at every boot (Model 30 286)''' &mdash; dead Dallas RTC battery.
* '''The beep-code table''', which added an "EGA / display adapter" pattern IBM does not distinguish and two rows that are not beep codes.
* '''Drive does not detect any diskette''' &mdash; PS/2 single-cable connector seated badly, or drive logic-board capacitor leak.
* '''"The Model 30 8086 has no hardware RTC"''' โ€” it has an MM58167AN at U16.
* '''ST-506 hard drive not detected''' &mdash; drive failure (very common on the original ST-506 fitted to 8530-021 / E21 / E31).
* '''"The Model 55 PSU is a drop-in replacement"''' โ€” no source was found for this, and IBM publishes the Model 25/30 supply's own figures, which are above.
* '''Random crashes under load''' &mdash; aged PSU electrolytics. Recap.
* '''A 3 A rating for the keyboard fuse''', and '''2408 "VGA ground circuit fault"''', which is not in IBM's 24xx family.
* '''VGA shows wrong colours (Model 30 286)''' &mdash; planar VGA gate array (15F6864) failing.


== Related Pages ==
== Related pages ==


* [[IBM PS/2 Model 30]]
* [[IBM PS/2 Model 30]]
* [[IBM PS/2 Model 30 Maintenance Guide]]
* [[IBM PS/2 Model 30 Maintenance Guide]]
* [[IBM PS/2 Model 30 Capacitor Replacement Guide]]
* [[IBM PS/2 Model 30 Capacitor Replacement Guide]]
* [[IBM PS/2 Model 25 Troubleshooting Guide]] &mdash; sister all-in-one
* [[IBM PS/2 Model 25 Troubleshooting Guide]] โ€” the all-in-one sibling, same supply document
* [[IBM PS/2 Model 50 Troubleshooting Guide]] โ€” the numeric error code reference
* [[Capacitor Failure Symptoms]]
* [[Capacitor Failure Symptoms]]
* [[Recommended Tools]]


== References ==
== References ==


* [https://www.ardent-tool.com/trouble/ps2error.html PS/2 Error Codes], Ardent Tool of Capitalism. Authoritative source for the complete 1xx&ndash;25xx PS/2 numeric error code listing, the Power-Good-in-150-ms PSU spec, and dead-system isolation procedure.
<references>
* [http://museodelcomputer.org/parts/ibm/ps2_50z/docs/error_codes.htm PS/2 Error Codes], Museodelcomputer.org. Reference for the audio code patterns and minor code structure.
<ref name="ard_err">''PS/2 Error Codes'' and the linked error-code pages, [https://www.ardent-tool.com/trouble/ps2error.html Ardent Tool of Capitalism]. A community reference reproducing IBM's own symptom-and-FRU lists; content credited on the site to Louis F. Ohland, Peter H. Wendt, David L. Beem, William R. Walsh, Tatsuo Sunagawa, Tomรกลก Slavotรญnek, Jim Shorney, Tim N. Clarke, Kevin Bowling and others, maintained by Tomรกลก Slavotรญnek. Source for the beep-code patterns, the numeric error-code list and its FRU/action column (including the keyboard voltage error's "if no fuse in system replace system board" note, the 17xx FRU list ending with the power supply, the 6xx media-error reading and the 24xx colours-change discriminator), the diagnostics procedure and the dead-system procedure with the 150 ms Power Good requirement and the line-cord advice. This wiki's transcription of the code list lives on the [[IBM PS/2 Model 50 Troubleshooting Guide]].</ref>
* [https://www.dosdays.co.uk/computers/IBM%20PS2%20Model%2030/ibm_ps2_model_30.php IBM PS/2 Model 30 &mdash; DOS Days]. Reference for the proprietary ST-506 interface, ISA-bus expansion, MCGA / VGA distinction and the (failure-prone) original hard drive.
<ref name="ard_batt">''PS/2 RTC/CMOS Batteries'', [https://www.ardent-tool.com/misc/batteries.html Ardent Tool of Capitalism], based on content by Bob Eager and Peter H. Wendt. Source for the per-variant cross-reference (Model 30 8086: 3 V battery, soldered, no FRU assigned; Model 30-286: Dallas RTC module FRU 8509237), for the Dallas module being a DS1287 replaceable by the DS12887 or DS12887+ and the 'A'-suffix parts but '''not''' the DS12C887, and for a weak battery losing all of the setup information while a partial loss is probably not the battery.</ref>
* IBM, ''IBM Personal System/2 Hardware Maintenance Manual'' (S30G-2207, October 1994). FRU listings and error-code references.
<ref name="ard_2530pwr">''8525 / 8530 Power'', [https://www.ardent-tool.com/8530/Power.html Ardent Tool of Capitalism]. Source for the manually switchable 100โ€“125 V ac / 200โ€“240 V ac input, the Power Good behaviour after five seconds off, the minimum under-voltage sense levels (+5.0 โ†’ +4.5 V, โˆ’5.0 โ†’ โˆ’4.3 V, +12.0 โ†’ +10.8 V, โˆ’12.0 โ†’ โˆ’10.2 V), the Power Good reading of +3.5 to +5 V and the remove-everything test that follows from it, and the J7 and P3/P4 connector measurement tables reproduced here.</ref>
* IBM, ''IBM Personal System/2 Model 30 Reference Diskette''. Source for SETUP, Advanced Diagnostics (Ctrl-A).
<ref name="ard_30p1">''8530 Planar (8086)'', [https://www.ardent-tool.com/8530/Planar.html Ardent Tool of Capitalism]. Source for the "P-PLANAR" P/N 61X8823, the Intel 8086-2 at 8 MHz at U10, the '''NS MM58167AN RTC at U16''' with the 32.768 kHz crystal at Y4, the MCGA devices (72X8300 at U11, 72X8205 at U12, IMSG171S RAMDAC at U5, SRM2264M 8K ร— 8 SRAM at U38), F1 and F2 as the mouse and keyboard fuses, and the full text of IBM's tips ECA 023 (erroneous/false display characters), H007157 (the fixed disk / system board / Advanced Diagnostics combination table with its FRU numbers and serial breakpoints), H043664 (cursor keys and 501 POST errors โ€” the U9 72X8203 and U15 61X8965 "Japan" marking check, replacement board FRU 64F3453 and the 55-00HV000 serial breakpoint) and H094152 (the 8515 and 8516 displays not being supported on 8086 machines, with the supported display lists for both variants).</ref>
<ref name="ard_30p286">''8530 286 Planar'', [https://www.ardent-tool.com/8530/Planar_286.html Ardent Tool of Capitalism]. Source for the VLSI VL82CPCAT chipset, the DS1287 RTC/CMOS at ZM35, the 15F6864 VGA controller at ZM18, the 8042/8742 keyboard controller at ZM33, the ยตPD765AC-2 FDC at ZM19, the 286-10 at ZM36, J13 as the password override, and the PLCC-CPU planar FRU 34F0046 (37F6654) / 33F8700.</ref>
</references>


{{Navbox-IBMComputers|state=collapsed}}
{{Navbox-IBMComputers|state=collapsed}}

Latest revision as of 10:40, 27 September 2026

IBM PS/2 Model 30. Source: Wikimedia Commons.

Fault diagnosis for the IBM PS/2 Model 30 (machine type 8530) โ€” an ISA desktop, in an 8086 form and a 286 form that share almost nothing electrically.

Which machine do you have

[edit | edit source]
The two Model 30 planars
Variant Processor RTC and battery Notes
8530-001 / -002 / -021 (8086) Intel 8086-2 at 8 MHz (U10) NS MM58167AN RTC at U16 with a 32.768 kHz crystal at Y4, and a 3 V battery soldered to the board โ€” IBM assigns it no FRU "P-PLANAR" P/N 61X8823. MCGA video: 72X8300 memory controller at U11, 72X8205 video formatter at U12, IMSG171S RAMDAC at U5, 8 KB SRAM at U38. F1 and F2 are the mouse and keyboard fuses[1][2]
8530-E01 / E0R / E21 / E2R / E31 (286) 286 at 10 MHz (ZM36) Dallas DS1287 RTC/CMOS module at ZM35, IBM FRU 8509237 โ€” the whole module is the replaceable part VLSI VL82C1xx chipset. VGA controller 15F6864 at ZM18, 8042 or 8742 keyboard controller at ZM33, ยตPD765AC-2 FDC at ZM19, F1 the fuse, J13 the password override. The PLCC-CPU planar is FRU 34F0046 (37F6654) / 33F8700[3][2]

The 8086 Model 30 does have an RTC. An earlier version of this page said it did not, and therefore could not report CMOS or clock errors. IBM's own planar list puts a National Semiconductor MM58167AN at U16 with its own crystal and a soldered 3 V cell.[1][2]

15F6864 is a VGA controller, and it is on the 286 planar โ€” not on the 8086 planar, which has no VGA at all. It is also the VGA controller on the 8525-286 and the 8570 Type 1 planar.[3]

Power supply โ€” IBM's own figures

[edit | edit source]

The Model 25 and Model 30 families are ISA machines, and IBM serviced them under the Personal Computer maintenance library rather than the PS/2 one. That is why their supplies document four DC rails, including โˆ’5 V, where the Micro Channel machines document three.[4]

The supply is manually switchable and must be set to either the 100โ€“125 V ac or the 200โ€“240 V ac range before the power cord goes into an outlet.[4]

Minimum under-voltage sense levels

[edit | edit source]

When the machine has been off for five seconds or more and is then switched on, the supply generates a Power Good signal that resets the system logic. Its presence means the supply is working and that the minimum under-voltage sense levels have been established:[4]

Under-voltage sense levels
Output Minimum under-voltage sense level
+5.0 V +4.5 V
โˆ’5.0 V โˆ’4.3 V
+12.0 V +10.8 V
โˆ’12.0 V โˆ’10.2 V

Power Good itself should read +3.5 to +5 V with the machine running correctly. If it does not, remove all adapters, cables and extra memory and try again; if Power Good is then valid, a device was shorting. Refit them one at a time.[4]

This is more useful than a general rail check, because it tells you what the supply itself is deciding: a rail that reads plausibly on a meter but is below its sense level will not let POST start at all.

Where to put the probes

[edit | edit source]

IBM's own connector measurements, with pin 1 towards the rear of the machine:[4]

Models 25, 25 SX and 25-286 โ€” connector J7
โˆ’ lead + lead V dc min V dc max
P7-5 P7-10 +4.8 +5.2
P7-9 P7-5 +4.5 +5.4
P7-5 P7-3 +11.5 +12.6
P7-4 P7-8 +10.8 +12.9
Models 30 and 30-286 โ€” connectors P3 and P4
โˆ’ lead + lead V dc min V dc max
P3-5 P4-4 +4.8 +5.2
P4-3 P3-6 +4.6 +5.5
P4-1 P3-3 +11.5 +12.6
P3-4 P4-2 +10.8 +13.2

Note that these are pairs of pins, not a rail measured against chassis ground โ€” IBM specifies both leads for each reading.

Dead system

[edit | edit source]

IBM's stated most likely cause of no POST, a blank screen and no beep is a device shorting out the power supply. The system board must present Power Good within 150 ms or the supply shuts itself down internally โ€” which is the mechanism behind the "click and then nothing" symptom.[5]

  1. Check the line cord at both ends. IBM singles this out: intermittent faults, including unexpected power-on resets during operation, come from line cords that are not fully seated or too loose to make a tight connection. Remove the cord from the outlet before working on its contacts.[5]
  2. Measure Power Good โ€” it should be +3.5 to +5 V. If it is not, remove all adapters, cables and extra memory and retry.[4]
  3. Remove all external connections.
  4. Power on and listen for a beep.
  5. No beep: check speaker continuity; if good, replace the planar memory and retry; if still silent, measure the rails against IBM's figures above โ€” correct rails mean the planar, wrong rails mean the supply.
  6. A beep: refit adapters and memory one at a time. When the beep stops, the last item fitted is the suspect.

IBM's own list of the usual causes includes a SIMM inserted backwards โ€” which has genuinely been seen in service โ€” an adapter not fully seated, the power cord not plugged in, an unsupported or defective adapter, and, on a loaded machine, too many drives spinning up at once.[5]

Reference Diskette and Advanced Diagnostics

[edit | edit source]

Neither machine has a built-in SETUP screen; configuration and diagnostics come from the Starter Diskette and the Advanced Diagnostics diskette.

  • Always cold-boot โ€” insert the diskette before powering on. Soft-booting it can report false errors and falsely indicate that a power-on password is already set.[5]
  • Ctrl-A from the main menu enters Advanced Diagnostics, which produces the minor codes.
  • If several errors appear, resolve them in the order presented โ€” the first can cause the others.[5]
  • The diskettes are machine-specific. A Model 30 diskette is not interchangeable with a Model 25 one, and the 8086, 286 and SX variants each have their own.
  • Run Advanced Diagnostics before replacing anything. If the diagnostics do not fail, replacing parts probably will not help.[5]

Battery

[edit | edit source]
What this machine takes
Variant Battery IBM FRU
8530 (8086) 3 V battery, soldered to the board None assigned
8530-286 Dallas RTC/CMOS module (DS1287) โ€” replace the whole module 8509237

For the 286: the DS1287 is obsolete and is replaced by the pin-compatible DS12887, or the lead-free DS12887+. The 'A'-suffix parts are fine too. Do not use the DS12C887 โ€” it is the Year-2000-compliant variant but uses the century byte differently and will not work in a PS/2. The module can also be reworked with an external cell instead of replaced.[2]

Defective or weak batteries cause the loss of all setup information. If only part of the setup is lost, the battery is probably NOT the cause.[2]

IBM service tips for this machine

[edit | edit source]

Four IBM tips survive for the Model 30 and all four are worth knowing before you start swapping parts.[1]

Cursor keys and 501 POST errors (H043664)

[edit | edit source]

On a limited number of 8530-001 and -021 machines, using the cursor movement keys makes unexpected characters appear on the display, and a 501 POST error appears once two or more option cards are fitted (the 3270 and Token Ring half-size adapters among them). The problem may be intermittent and may not be detected by Advanced Diagnostics โ€” so a clean diagnostics run does not clear it.

The check is visual, and it is a printed-marking check, not a part number check. Inspect the planar modules at:

  • U15 โ€” next to the speaker. P/N 61X8965.
  • U9 โ€” next to the math coprocessor socket. P/N 72X8203.

Good chips have the word "Japan" printed on them. Affected chips do not. If either module lacks it, the fix is a new system board, FRU P/N 64F3453. Machines with a serial above 55-00HV000 left the factory with the new board. If a math coprocessor is fitted, move it across to the new board.

Bus adapter battery voltage (H067282)

[edit | edit source]

Early releases of the 8530 HMS print the wrong pin numbers. On the system board maps, page 0100-4, the pins labelled "0" and "39" in figure one should be "1" and "40". The correct measurement is at pins 38 and 39 โ€” the second and third pins from the right edge. Replace the bus adapter card if the voltage between the correct pins is less than 2.5 V.

This is exactly the class of error this wiki exists to catch, and it is IBM's own.

Fixed disk, system board and diagnostics must match (H007157)

[edit | edit source]

Successful fixed-disk operation on the 8530-001 and -021 depends on the right combination of system board, hard disk and Advanced Diagnostics version. Formatting with the wrong level of Advanced Diagnostics causes fixed disk errors.

IBM's required combinations
Machine System board Fixed disk Diagnostics
8530-021 below serial #2500000 61X8907 61X8929 8530 Advanced Diagnostics, any version
8530-021 above serial #2500000 33F8474 6128285 8530 Advanced Diagnostics v1.05
8530-001, all 33F8474 6128285 or 6128279 8530 Advanced Diagnostics v1.05
8530-E21 only 27F4069 6128285 8530-286 Advanced Diagnostics v1.02
8530-E31 and E01 34F0046 6128285 or 6128279 8530-286 Advanced Diagnostics v1.02

Fixed disks 6128285 and 6128279 must not be fitted to any 8530-021 below serial number 2500000, and machines above that serial, plus all -001 machines, must be formatted only with v1.05.

8515 and 8516 displays are not supported on the 8086 machines (H094152)

[edit | edit source]

An 8086 Model 30 (-001, -002, -021) connected to an 8515 or 8516 display shows compressed and distorted video. That is not a fault. The supported displays are:

  • 8086 machines (-001, -002, -021): 8503, 8504, 8507, 8512, 8513, 8514.
  • 286 machines (E01, E21, E31, E41): 8503, 8504, 8512, 8513, 8514, 8515, 8516.

ECA 023

[edit | edit source]

The Model 30 also has an ECA 023, "Erroneous/False Display Characters". Read it alongside H043664 above, which describes the same class of symptom.[1]

Keyboard and 301 errors

[edit | edit source]

An XX 301 where XX is a hex scan code usually means something is resting on a key. IBM's FRU order for 301 is the keyboard cable, then the system board.[5]

For 305, IBM's order is the keyboard port fuse, the keyboard cable, then the system board โ€” and if there is no fuse in the system, replace the system board. No rating is published for the fuse and none is given here.[5]

The keyboard and mouse sockets are identical Mini-DIN-6 connectors wired differently; swapping them produces a 301 and a pointing-device error together.

Power-on password

[edit | edit source]

J13 is the password-override connector on the 286 planar. Jumper it momentarily during power-on and the BIOS clears the password.[3]

Hard disk

[edit | edit source]
  • 17xx with the drive fitted โ€” often stiction on the original ST-506. Rotate the drive gently around its spindle axis with the power off, refit and retry; then image the drive and replace it.
  • Before anything else on a -001 or -021, check the H007157 combination table above. A machine that has been "repaired" with a mismatched board, drive or diagnostics version will produce disk errors that no amount of further part-swapping will fix.
  • IBM's FRU order for the general 17xx group ends with the power supply, so a sagging supply is a documented cause of disk errors.[5]

Diskette drive

[edit | edit source]

6xx codes are mostly media errors in IBM's own reading: try a known-good diskette, and if the error returns, replace the drive. Reseat the single-cable connector and check pin 1 orientation. A continuous activity light with no read means the cable is on backwards โ€” power off at once.[5]

Display

[edit | edit source]

IBM's discriminator for the 24xx codes is whether the screen colours change: if they do, suspect the display; if they do not, suspect the system board. The general display-adapter family is 0005 XXXX.[5]

What has been removed from this page

[edit | edit source]
  • The hand-transcribed numeric POST tables, replaced by the single reference on the Model 50 guide. They gave PC/AT-era meanings as PS/2 codes โ€” 104 as an 8259A error or a protected-mode failure, 107 as an NMI test, 108 as a timer bus test, 109 as an 8237A DMA test, 121 as an unexpected hardware interrupt, 151 as a bad battery and 162 as a checksum mismatch. IBM's protected-mode code is 115; 121 is a memory-remapping note and is information only; there is no 151; and 162 is a devices-powered-on check, with 173 being the checksum code.
  • The beep-code table, which added an "EGA / display adapter" pattern IBM does not distinguish and two rows that are not beep codes.
  • "The Model 30 8086 has no hardware RTC" โ€” it has an MM58167AN at U16.
  • "The Model 55 PSU is a drop-in replacement" โ€” no source was found for this, and IBM publishes the Model 25/30 supply's own figures, which are above.
  • A 3 A rating for the keyboard fuse, and 2408 "VGA ground circuit fault", which is not in IBM's 24xx family.
[edit | edit source]

References

[edit | edit source]
  1. โ†‘ 1.0 1.1 1.2 1.3 8530 Planar (8086), Ardent Tool of Capitalism. Source for the "P-PLANAR" P/N 61X8823, the Intel 8086-2 at 8 MHz at U10, the NS MM58167AN RTC at U16 with the 32.768 kHz crystal at Y4, the MCGA devices (72X8300 at U11, 72X8205 at U12, IMSG171S RAMDAC at U5, SRM2264M 8K ร— 8 SRAM at U38), F1 and F2 as the mouse and keyboard fuses, and the full text of IBM's tips ECA 023 (erroneous/false display characters), H007157 (the fixed disk / system board / Advanced Diagnostics combination table with its FRU numbers and serial breakpoints), H043664 (cursor keys and 501 POST errors โ€” the U9 72X8203 and U15 61X8965 "Japan" marking check, replacement board FRU 64F3453 and the 55-00HV000 serial breakpoint) and H094152 (the 8515 and 8516 displays not being supported on 8086 machines, with the supported display lists for both variants).
  2. โ†‘ 2.0 2.1 2.2 2.3 2.4 PS/2 RTC/CMOS Batteries, Ardent Tool of Capitalism, based on content by Bob Eager and Peter H. Wendt. Source for the per-variant cross-reference (Model 30 8086: 3 V battery, soldered, no FRU assigned; Model 30-286: Dallas RTC module FRU 8509237), for the Dallas module being a DS1287 replaceable by the DS12887 or DS12887+ and the 'A'-suffix parts but not the DS12C887, and for a weak battery losing all of the setup information while a partial loss is probably not the battery.
  3. โ†‘ 3.0 3.1 3.2 8530 286 Planar, Ardent Tool of Capitalism. Source for the VLSI VL82CPCAT chipset, the DS1287 RTC/CMOS at ZM35, the 15F6864 VGA controller at ZM18, the 8042/8742 keyboard controller at ZM33, the ยตPD765AC-2 FDC at ZM19, the 286-10 at ZM36, J13 as the password override, and the PLCC-CPU planar FRU 34F0046 (37F6654) / 33F8700.
  4. โ†‘ 4.0 4.1 4.2 4.3 4.4 4.5 8525 / 8530 Power, Ardent Tool of Capitalism. Source for the manually switchable 100โ€“125 V ac / 200โ€“240 V ac input, the Power Good behaviour after five seconds off, the minimum under-voltage sense levels (+5.0 โ†’ +4.5 V, โˆ’5.0 โ†’ โˆ’4.3 V, +12.0 โ†’ +10.8 V, โˆ’12.0 โ†’ โˆ’10.2 V), the Power Good reading of +3.5 to +5 V and the remove-everything test that follows from it, and the J7 and P3/P4 connector measurement tables reproduced here.
  5. โ†‘ 5.00 5.01 5.02 5.03 5.04 5.05 5.06 5.07 5.08 5.09 5.10 PS/2 Error Codes and the linked error-code pages, Ardent Tool of Capitalism. A community reference reproducing IBM's own symptom-and-FRU lists; content credited on the site to Louis F. Ohland, Peter H. Wendt, David L. Beem, William R. Walsh, Tatsuo Sunagawa, Tomรกลก Slavotรญnek, Jim Shorney, Tim N. Clarke, Kevin Bowling and others, maintained by Tomรกลก Slavotรญnek. Source for the beep-code patterns, the numeric error-code list and its FRU/action column (including the keyboard voltage error's "if no fuse in system replace system board" note, the 17xx FRU list ending with the power supply, the 6xx media-error reading and the 24xx colours-change discriminator), the diagnostics procedure and the dead-system procedure with the 150 ms Power Good requirement and the line-cord advice. This wiki's transcription of the code list lives on the IBM PS/2 Model 50 Troubleshooting Guide.


Retrieved from "https://wiki.retrotechcollection.com/index.php?title=IBM_PS/2_Model_30_Troubleshooting_Guide&oldid=14769"