8580 Common Devices
Look at Fred Spencer's PS/2 8580 Information Collection
   (He steals from me, I loot from him. Perfect)
162/163 Error Code Combination  (to Spencer's 8580 site)
Battery Orientation in Speaker Battery Assembly
   8580 Battery Cross Reference  (to Spencer's 8580 site)
   Time stands still (System time doesn't change)
8580 Floppy Types (to Spencer's 8580 site)
XGA-2 Install attempt failing after copy of ADF files  (to Spencer's 8580 site)
Repair of  PS/2 1.44mb Floppy (to Al Savage)
   Comments on Mitsubishi and Alps Floppy Drives
   Replacing Capacitors
Pinout for Speaker Battery Assembly Header
Clear a Power-On Password 
Open an 8580
   Remove Front Bezel
Open a Locked 8580 Case
>16MB on an 8580 under Linux
161/163 Error Code Combination Trivia

Comments on ALPS and Mitsubishi Floppy Drives
>There was some dust on the fdd controller below the fan, but I've seen pcb's in even worse condition. This machine seems to be on duty not for long time. I also
 cleaned the drive (mitsubishi type), but all didn't help.

Peter said
   Good. There were a lot dust-related errors on Mod. 80 - therefore I'd
explicitely mentioned it on my "Known Errors" page. Office and household
dust is often conductive and also causes a lot config lost and date/time errors.
The Mod. 80 PSU fan is over a very sensitive part of the planar.

>Ask somebody for the drive from an unused 8580.

   But "unused" does not automatically mean "in working condition". In the worst case they placed the computer there *because* it showed odd errors too. Worth a try nontheless.

>  How long is the half life time of these drives? They seem to die quite often!

  The ALPS drives had a much higher error rate than the Mitsu-drives. The ALPS are also much more wicked to take apart due to the amount of "0"-sized Philips screws. The Mitsu is a somewhat cheap design, but far more easier to service. Both drives are dust collectors due to the lack of a dust-shield covering the disk slot when no floppy is in the drive.
   The Mitsu 355 tends to collect dirt around the R/W-head actuator spindle -
which causes the head to stick and to improper positioning. The ALPS often
loose the head-damper plate (a piece of metal with some foam pad - glued on
the R/W head assembly (It's a remove-and-throw-away part).
   The Mitsu also tend to have "killer caps" installed - which cause 600-errors
(electronical errors). The repair / checkup is nicely displayed on HERE

Replacing Capacitors
Alan Douglas sounds off with:
>Unsoldering the caps is a little tricky. If you pull to strong or heat too long, you might risk to tear off the copper pad from the pcb. Thats what happend to me :-(  Fixing the cap on the thin copper lane was not easy!

   I found it easiest to cut the caps in two with flush-cutting diagonals, then remove the rubber seal and get access to the wire stubs from the top. 

>For replacment of the 22uF and 1uF caps, I used surface-mountable electrolytic caps of the same type. However, 0.22uF electrolytic caps are difficult to obtain, so I used ceramic caps instead.

   I also used ceramics there, and tantalum chips for the others.

   Just a note that I occasionally repeat: the best way of locating bad electrolytics is by checking ESR, effective series resistance.  A dandy little meter for this is a kit made by Dick Smith Electronics in Australia, and sold by various dealers around the world. 
   A good starting point is the designer's page.   It's generally under $50, and has received rave reviews in the sci.electronics.repair newsgroup and lsewhere.

Floppy Support Structure


Battery Orientation in Speaker Battery Assembly


The contacts on the battery point to the rear, where they press against the spring clips.

The Day Time Stood Still (Time does not Change)
j131 asked
   I am using this machine under linux (debian 2.2.r3) as a sort of occasional (in fact rather rarely turned on) home-file-server. The only symtom is that:
 _the time at powerup is the same as it was at powerdown_. Neither BIOS nor OS complains... Only me... There are no 161/163 errors.

Alfred Arnold replies with  "just the facts"
   Then your machine's battery might be shortly before death.  Typically, when battery voltage drops over time, first the 32 kHz oscillator in the RTC stops working (so time is standing still), but it is still high enough to keep the CMOS RAM contents intact.  You're probably not going to have to wait too long any more till the 161&163 errors come up...

Pinout for Battery/Speaker connector 
       6 pin Header
1 Gnd    4 NC 
2 Gnd    5 Clear 
3 +6v    6 NC

Clear a Power-On Password 

Model 50, 60, 70, 80 Override Jumper 
    1. Power-off the computer and unplug the power cord. 
    2. Remove the system-unit cover. 
    3. If the speaker assembly has pins, short pins 2 and 3 together. 
    4. If the speaker assembly does not have pins, insert a meter lead into connector 1 and short the other end of the lead to frame ground. 

  With the assembly shorted, power-on the computer. This erases the power-on password. Remove the short after POST is finished.

Open a 8580


   Use a coin or large standard screwdriver to turn the two large screwheads to the left. After they are both loose (they're captive screws, they won't come out), pivot the side cover on the bottom hinges until the top edge has cleared the case. Pull up and out.

   If the lock in the top center of the cover is locked, look at the next section.

Remove Front Bezel
  Open side cover. Look in the bottom front corner- pull the speaker/battery assembly out horizontally by grabbing the speaker assembly frame and pulling it.
  Don't pull the battery header out of the planar, if that happens, your configuration will disappear. Just put the loose speaker assembly down on the bottom of the case.
   Notice the white thermoplastic catch below the grille? Press it down and out towards the front of the case. The second catch is sort of hidden by the black foam around the grille. Feel for the inner edge of the foam (towards the inner side of the case) and you will feel the end of the second catch. Push it down and to the front. Once both catches are started (pushed level with the case) grab the bottom edge of the front bezel and pull it out. The bezel should come off.

NOTE This bezel is NOT like the 95 bezel where there are upper pivots. There is a single catch to the right of the power switch. Think of the two bottom catches and the top catch in the same way you'd pick up a bowling ball (which I haven't done since High School). These "fingers" grip the frame of the case and pull the bezel onto it. 

Open a Locked 8580 (and 8560/8565)
Tony Ingenoso  sez:

It's very easy if there's no card in the bottom slot, less easy if there is. 

1)  Flip the machine upside down 
2)  Remove tall front bezel 
3)  Loosen lower (now upper when upside down) slot retainer 
4)  Take a thin blade screwdriver and push the slot blank into the case 
5)  With a flashlight, you should now be able to see the front and rear retaining tabs on the bottom (now top) of the side cover. 
6)  Lever down on these retaining tabs with a long thin screw driver until they release. From the rear you'll be going in through the slot hole, from the front you'll be going in through one of the ~3/8" holes the bezel snaps into. It will be obvious which is the right one to use once you take a peek with the flash light. 
   Once the first tab is clear, pull the cover out enough to slip a shim in so it doesn't snap back in place again on you. Once both are clear, the cover will come off. 
   If there is a card in the bottom slot, it presents more of a problem. If it looks like something you don't mind wasting, loosen the retainer, pull on connectors to get it out of the MCA slot and jam/twist/break enough of the thin metal out of the way so you can get in with the long thin screwdriver. 

ALTERNATIVE APPROACH: -- drill a couple of small holes in the bottom of the case where the cover tabs are located, then use these holes to apply pressure to the tabs to get them to release. This doesn't involve having to waste something in slot 8, but it does involve mangling the case to a very minor degree. Use a depth gauge!!! If you drill through the retaining tab, there'll be nothing to push on ;-> 

Ed. Measured my 8580.  Front tab is 2.8" from front corner, 1.5" in, set depth to .25". Rear tab is 3.6" from back, 1.5" in, .25" deep.

>16MB on a 8580 under Linux
Ed Avis takes time out from the pub and mistakenly admits:
   I upgraded my 8580 from 12MB to 24MB RAM and Linux's ibmmca.c SCSI driver stopped working.  (I have a couple of Spock Primes in there.)
   To cut a long story short, this is caused by a DMA limitation on the 8580 (and possibly other early PS/2s) that restricts DMA to the lower 16Mbyte of RAM.  Michael Lang, the maintainer of the driver, was very helpful and after trying various bugfixes, suggested changing the flag 'unchecked_isa_dma' to 1.  I'm not sure what this does, but it fixed the problem.
   So if you have >16Mbyte RAM and you find that Linux crashes on boot with errors from IBM MCA SCSI, edit the file ibmmca.h in the kernel sources and change a couple of lines to read:

    unchecked_isa_dma: 1,                 /*32-Bit Busmaster */ \

instead of 'unchecked_isa_dma: 0'.  Hopefully this will become an option in later versions of the kernel - or maybe with the new 2.4 kernel and 4.0 ibmmca driver it is not needed.

161/163 Trivia
Peter "remembered" this-
   If the floppy *seems to start* booting (LED lights up, FDD whirrs for a second) after the beeps and 161 / 163 display but then "hangs there" press [F1] to force loading. The older Mod. 80 seem to have a *slighly* different BIOS that not always auto-loads even a valid reference disk. After 161/163 errors the first loading of the reference disk *may* take up to 90 seconds - because the machine BIOS uses the lowest transfer rate available, because it is uncertain about the type of FDD.

If the F1 fails and a second F1 brings you into IBM ROM BASIC suspect
- the floppy *not* being a valid reference disk (extraction fault ?)
- accidently snatched a defective floppy for the reference disk creation ?
- having a dirty / misaligned / defunctional FDD

9595 Main Page