• Please review our updated Terms and Rules here

IBM PS/2 Model 70 Reference Disk/Boot Behavior Questions

ataritoobin

Experienced Member
Joined
Jul 17, 2008
Messages
64
Location
Bay Area, CA, USA
Hello!

I've been working on getting an IBM Model 70-E61 up and running, but am (surprise!) having some issues getting the Reference Disk booted up.

The drive originally in the machine was a ragged Mitsubishi 15F7503 1.44MB 3.5" drive with the top head sheared off. On boot, I'd get the expected 161/163 error.

I found another Mitsubishi 15F7503 and cleaned the heads, cleaned up the worm gear, and recapped it.

I replaced the battery and get the expected 161/163 error on boot.

A downside here is I understand the only way I have access to writing the Reference Diskette floppy using the rf7080a.img source at the moment is less than ideal:
  • Windows 10 with external USB floppy and WinImage
  • dd in Terminal on MacOS 10.11 with USB floppy
  • DOS 5 on Virtual Box on MacOS 10.11 with USBASPI.SYS configured to use the external USB floppy drive. With this setup I tried RAWRITE. If I use REFSTAMP.COM afterwards, it reports that the signature is already stamped. This is the same if i write the disk with WinImage, dd or RAWWRITE. If I try running REFSTAMP on a freshly formatted disk, it will stamp the signature and report "Disk Stamped OK" (so it says).
  • That said, I can create a disk using rf7080a.img and have matching checksums with the source image file when I create an image of the resulting disk back into a file. I can see and access the contents on the disk as expected (even on a Quadra 700).
When I try to boot the Model 70, it will access/initialize the drive, give the 161/163 error, beep twice, drive will seek 4 times, then throw the red “not OK” + IBM manual. If I hit F1, the floppy will seek four more times, then go to BASIC.

I figured I’d give the TexElec PS/2 to standard floppy adapter a shot with a Sony standard 3.5" PC floppy drive. I was pretty surprised when it gave the exact same result: it will access/initialize the drive, beep twice, seek 4 times, then throw the red “not OK” + IBM manual. If I hit F1, it will seek four more times, then go to BASIC.

For the heck of it, I cleaned up the original drive, sans top head, recapped it and had the same exact result.

If I make a DOS bootable diskette, I get the same result as well, which I'd expect since the system CMOS isn't configured.

At this stage, assuming my Reference Disk is good (which perhaps isn't a sure thing with USB floppy drive, etc?), where would I go next? Is this "seek 4 times" behavior what you see on an unconfigured Model 70 and you try to boot with a non-Reference Disk? I’d imagine it’s possible some floppy disk controller chip might be bad?

Thank you!
 
Last edited:
I surfed around a bit... did you make sure the floppy controller leads on the system board are clean? Not sure if you have a Type I or Type II
https://ardent-tool.com/8570/8570_Type1.html look for U89, NEC D72065L
https://ardent-tool.com/8570/8570_Type2.html look for U18 NEC D72065L

If it is a Type I, the FDC is close to the PSU fan intake, and the FDC leads may have some grunge on them. This issue is better known on 8580s, but any system can get it.

The Type II FDC is towards the front of the case, but the vents in the front allow air to travel over it.

A Q-Tip and alcohol [rubbing or denatured] will do the trick.

161/163, configuration, yep. No 6xx error codes? That's what I'd expect with a bad floppy drive. If the refdisk was bad, I'd expect an error message, asking you to insert a refdisk and hit F1...
 
Last edited:
Re-read the 8570 RETAIN tips...I remembered the 8570 uses a drive riser to connect the FDD [and HD!]

Pull the drive riser, buff the gold contacts that plug into the system board socket. Use a folded dollar bill [or bank note] to clean the drive riser contacts. Do NOT use an eraser [gummy residue] or abrasives [wear gold plating AND leave abrasive residue!]. Use compressed air [low pressure], computer duster, or a vacuum [IBM preferred, but the others do just fine] to blow out the drive riser socket on the system board. Insert / remove the drive riser into/from the system board socket three to four times [should buff up socket contacts].

Reassemble, insure drive riser is firmly and squarely seated in system board socket. Also czech cleanliness of FDC leads. Clean as needed.

Another issue to look into is contaminated heads. Not as likely... Use a brand of floppy that is good [IBM recommended... IBM branded floppies...] and let the system search a few times. Since you probably can't do a "DIR", you might have to boot three or four times with the "scratch" diskette.

Brings up a thought - how clean is the floppy drive used for creating the refdisk? Scratch diskette to clean that one as well [if you neeed to]....
 
Last edited:
Thank you for the tips! I appreciate it!

Yep, strangely I never saw any 6XX error codes.

Mine is the type Type I motherboard with the NEC D72065L @ U89. I cleaned up the chip and area with alcohol as you suggested and inspected for any cracked solder joints or the like.

I blew out the edge connectors with air and cleaned all edge connectors with the bill method.

I also cleaned the heads on the USB drive and tried again with the same results.

For fun, I pulled up MacTools on the Quadra and confirmed I could see the "3y" in the disk header required for the Reference Disk to boot. I tried changing and writing a different value back, then put the 3y back in and wrote it to disk just in case the alignment or something else might be off with the USB drive, but had the same result on the PS/2

I'll give it a final go with a Gotek at some point this week, but have the sneaking suspicion it's a problem with some circuitry on the motherboard :(
 
Last edited:
With the Gotek, I got the same "seek 4 times" (via LED flash), then it throws the red “not OK” + IBM manual. If I hit F1, it will seek four more times, then go to BASIC :(

EDIT: Strangely with the Gotek this evening, it would take longer and longer to "try" beyond the usual 4 seeks and eventually I got a "Non system disk or disk error" message which I hadn't seen before. I kept retrying and eventually it actually finally loaded the configuration program, but was super unreliable and fail any time it tried to read from the Gotek (e.g. "try backup copy" error).

Eventually it locked up, the Gotek seek light was stuck on, and on reboot the system totally won't POST whatsoever, even with all drives/riser removed, etc. If drives are plugged in, seek light is always on. Power supply voltages check out ok.
 

Attachments

  • PS2.jpg
    PS2.jpg
    156.1 KB · Views: 3
Last edited:
I tried checking the ROMs, and they compared OK to the images found online, and the behavior is the same with the ROMs and the CPU removed. I guess I'm in the market for a replacement motherboard until I have time to take a closer look at this. That leaves the case and the power supply (knock on wood) as the only really useable parts from the original machine :crazy:
 
I'm having the exact same issues with a model 50.
Could you get it working?
 
I wound up buying another Model 70 (Type 2) and it too had the exact same problem attempting to read disks. I replaced the D72065L chip and was able to get it working. Not fun though since it's in an awkward place on the Type II motherboard :(

However, I never resolved the issue with the original Type I motherboard that no longer shows any signs of life.
 
Hey, sorry to resurrect two years old thread, but I think resurrection is kind of a theme here anyway. So to help anyone stumbling upon the same as me and this thread author:

I have PS/2 50z with the same symptoms: 161/163, no boot.

I had:
1. Memory check: Ok
2. 301
3. 161
4. Floppy init noises
5. 163
6. Nothing happens, floppy light on indefinitely

Solved 301 keyboard with getting proper ps/2 one, usb with adaptor did not work

Then installed new CRP2 battery from Amazon

Still the same, sans 301

Replaced two SMD caps in fdd, got infamous “not ok, consult manual”. Progress!

Tried every variant of reference diskette creation, it was not it. dd on mac or WinImage in win 11 is fine.

Cleaned drive heads and all contacts between fdd-riser board-system board with rubbing alcohol.

It worked! It booted, set everything up, fdd test succeed. No errors on boot now, one beep.

Hope it helped someone. Never too clean with that old stuff.
 
Tried every variant of reference diskette creation, it was not it. dd on mac or WinImage in win 11 is fine.

Cleaned drive heads and all contacts between fdd-riser board-system board with rubbing alcohol.

It worked! It booted, set everything up, fdd test succeed. No errors on boot now, one beep.

Hope it helped someone. Never too clean with that old stuff.

When you say set everything up, were you able to get your reference disk to read?
Recently went through this due to a dead battery in one of my P70s and I had to create a reference disk using DOS 5.0 on an older PC with a real 3.5 disk drive.
The doc supplied with the file even mentioned that simply viewing the disk's directory under Windows would destroy it.
I was under the impression there would be no way to create a reference disk under Windows or especially with a 3.5" USB drive.
 
My PS/2 machine boots from reference diskette created with WinImage on Windows 11 64bit via SmartDisk usb fdd (supposedly Mitsumi D353FUE, given markings on the label). I know for sure because after I tried to "Copy option diskette" during setup and it prompted me to "Insert backup Reference Diskette to save option(or something)" I used original (not original original from 1987, but one I created with dd, I think) Reference and after that it failed to work with "BAD REFERENCE DISKETTE". So I went on and formatted and written it again with WinImage, and it worked fine.

I even went to lengths of installing windows 8 in VirtualBox to run refstamp (before I recapped fdd and was not sure my diskette is good) and it said all good, already patched.
 
I successfully created Model 70 reference disks on Windows 10 using "RawWrite for Windows" using a USB floppy drive.
 
Back
Top