• Please review our updated Terms and Rules here

Restoring an old IMSAI 8080

sponaugle

Member
Joined
Jan 30, 2019
Messages
26
Location
Portland, Oregon, USA
I'm working on restoring an old IMSAI 8080 I recently acquired. The system itself looks to have been sitting for the last 30 years or so. I removed everything and checked the power supply first, which measured perfect. I installed the cpu card and a ram card and the system came up without any problems. I was able to enter some code via the front panel which executed correctly.

The system had some other cards that look to have been 'stored' inside the unit. There were three floppy disk controller cards, plus two SIO cards and a 64K DRAM card along side a 64K SRAM card.

I also got two 8 inch floppy disk drives, which were a bit rough. One has a broken door latch, and both were pretty dirty. They were in an 'enclosure' but without the lid, so they also had what looked like 30 years of dust! I took the better drive of the two apart and cleaned it, cleaned the heads, lubed a few places, and checked the power supply (these drives use +24V, +5, and -12, plus 120V for the spindle motor).

I connected the drive up to a PC using the new TexElec 4 floppy controller and didn't have any problems getting it working. It is a single sided drive, and I was able to read and write a new single sided disk (77 tracks).

I have about 40 8" disks from the original IMSAI system that I'd like to be able to read. Using IMD I was able to scan and read the disks, but I'm curious about the format.

1699424608605.jpeg

Most of them (including the ones labeled System Backup) show up as 77 tracks, with track 0 being 500k MFM, Single Density, 26 sectors/track, 128 bytes/sector, while the rest of the tracks (1-76) are 500k MFM, Double Density, 26 sectors/track, 256 bytes/sector.

A small portion of them show up as track 0 being 500k MFM, Single Density, 26 sectors/track, 128 bytes/sector, while the rest of the tracks (1-76) are 500k MFM, Double Density, 15 sectors/track, 512 bytes/sector.

Is it common that in 8" CPM formats the first track is single density, while the rest of the tracks are double density? I could not find any formats in 22DISK that looked like this. Any other tools that I should try using to be able to read this format?

1699424593211.jpeg
 
For a start, the standard CompuPro 8" CP/M floppy format always used single density 128 bytes/sector on track 0. All other tracks could be either SD 128 bytes/sector and 26 sectors/track, or DD with a choice of 256/512/1024 bytes/sector arranged as 26/15/8 sectors/track respectively.
 
First off, nice acquisition.

Yes, from memory, 8" floppies sometimes used a single density track 0 but (of course) there are many floppy controllers and machines out there and some of the manufacturers had other ideas.

The general gist was that track 0 was always single density with a known geometry so it could always be read and would contain information about the remainder of the disk format.

Dave
 
Thanks for the info! I dug a bit more, and indeed it is mentioned in a few places that some of the Crememco setups have the first track single density.

I would like to be able to read/write these disks with 22DISK, so I set out to build a config. I can read these entire disks just fine in IMD, and I also installed ANADISK so I could look at the sectors and guess the layouts.

1699494556582.jpeg

That worked, and I was able to browse the disks and find the directory table starting on cylinder 2 as expected.

The sectors looked as I expected, with the expected pattern that seems to be common on 8" disks in terms of sector number sequence.

1699494636472.jpeg

I used this data to create a config for 22DISK:

1699494664738.jpeg

It is very close to some other configs, which is a good sign. However, when I try to use 22DISK with this config to read the disk (which I can directly read with ANADISK on this same setup), it somehow can't read the sectors?

1699494762386.jpeg

I can hit I, and it will go to the next sector, but again can't read it. (actually it says 'sector not found'). I'm not sure what could cause this problem, since the definition has the correct MFM,HIGH ( Double density, 500kpbs), which works correctly in ANADISK and IMD.

I did set the drive config for 22DISK correctly (using DISKETTE.CFG)

1699494875177.jpeg

Any ideas on what could be the problem? It seems like perhaps it is trying to read the disk at single density or perhaps at 250kbps, although the .DEF config shows MFM, HIGH. Just to be sure I tried all 4 combinations of FM,MFM, LOW , and HIGH with no success.

I did see that the line in my config:

SIDE1 0 1,7,13,19... has a first parameter that is the according to manual:

"The first term after the SIDE1 keyword is the logical side ID found in each ID address mark for that disk."

I suppose it is possible the side ID address mark is not 0 for some reason. I'm not sure how to determine that. I did use the ANADISK Shift-F2 function to read the diagnostic information, which appears to be the raw data including the sector gaps, but I didn't see any way to find the address ID mark.

I wish it were possible to have the 22DISK tools (the CPM part at least) run against an IMD image instead of the actual physical drive).
 
Well, I have been somewhat successful, but I'm not sure why. It *seems* like if I use CTOD first to copy a file, then CDIR works. But if I do CDIR before doing CTOD, I get the sector not found... I only noticed it because I use CMENU to format a new blank disk, the used CMENU to copy a file to the disk, and then the directory worked...then I swapped to the other old disk and the CTOD worked.
I'm not sure why I can't do a CDIR without having to do a CTOD first... hmm

1699504958875.jpeg
 
Back
Top