Boot Disk SD Emulation

4th article in Staübli Reborn: Industrial Robot with a Modern Control System
RoboticsReverse EngineeringEmbedded
2024-8-26

In order to be comfortable investing in the robot I had to be confident I could get the robot booted in a timely manner.

Boot Disk / Image

Upon delivery one major issue with the robot was the 7 minute boot time and single operating boot disk. My first priority was ensuring that I didn't have a single point of failure.

I bought a generic floppy drive and used win32diskimager to create an image of the boot disk, and was greeted by an error message.

Attempting to burn any of the disk images I got with the robot was also not fruitful in producing a working image. The previous owner described a laborious process of getting a good "spec" (CNC parameters) describing the motion control characteristics - how many pulses per rotation on the encoder etc, so I assume there are many incomplete and partial images.

SCSI emulator

The previous owner described getting a SD card emulator working so I knew it was possible, but details were light. I bought a ZuluSCSI Laptop emulator and spent quite a bit of time playing with the configuration file.

The SIO card inside the machine had to be pulled and opened every time I wanted to burn a new SD image, and it has a ton of wires and jumpers attached to it. Each attempt was... thrilling. Was it going to work after reassembly?

Process

There were many simultaneous threads being pulled together:

  1. Research on the robots specific SCSI drive configuration
  2. Research on how to configure ZuluSCSI
  3. Reassembling a working boot disk

Some key milestones were being able to see the same error message booting a broken disk image from a floppy and booting the same broken image from the SD card - which indicated that the scsi emulator was successfully fooling the controller.

Getting the robot to boot from a different floppy disk was also valuable. With those two it was only a few more testing iterations to get the robot booting from the SD card.

In the end after about 6 disassembly/reassembly boot cycles and a bunch of untracked floppy attempts I derived the following conclusions and got the robot booting mostly consistently:

  1. The existing boot floppy "yellow" had a bad sector in the operating system and couldn't be fully imaged, but had working parameters
  2. The system images were bootable but had bad parameters
  3. Create a (new) working boot floppy "good spec" by starting with the bootable system image and copying just the parameters from the "yellow" boot floppy
  4. ZuluSCSI expects a filesystem image named HD0.img and a zuluscsi.ini file
  5. Create a system image from the "good spec" floppy
  6. Expand the system image to have more room for programs
  7. Ensure ZuluSCSI has external 5v power because the SCSI bus doesn't provide it (low current? busted?)

Emulator Configuration

[SCSI]
Debug = 0
SelectionDelay = 255
DisableStatusLED = 0

EnableUnitAttention = 0
EnableSCSI2 = 1
EnableSelLatch = 0
EnableParity = 0
MapLunsToIDs = 0

[SCSI0]
Type = 0
TypeModifier = 0
SectorsPerTrack = 18
HeadsPerCylinder = 2
Vendor = QUANTUM
Product = 'GO_DRIVE1       '
Version = '1.0'
Serial = '                '
BlockSize = 512

The entire configuration (with comments) is available here: zuluscsi.ini

Notes

Here are the raw notes from the sessions:

Creating new img file on sd card

https://github.com/ZuluSCSI/ZuluSCSI-firmware?tab=readme-ov-file#creating-new-image-files

fsutil file createnew HD1.img 1073741824

Copied config from ADEPT V PLUS HDD SETTINGS.xml to zuluscsi.ini

Discovered SD on 2022A (blue) card is unreadable - staublified? Was not able to boot with the SD

Test: USB power?

Attempt 1: partial success

With usb power it "recognized" the drive as unformatted It identified the drive status, etc When attempting to format it started "communication timeout" etc

Attempt 2: using the .img from the floppy

Attempted boot, after several seconds:

Invalid record type: SX

Next step:

  • Attempt to format
  • Attempt to pull "better" image from floppy

Attempt 3: tweaking img

Disable compression in winimg Copy settings from edit filesystem Increase disk size to 100mb

Invalid record SX

http://thetechpage.mirror.kb1max.com/cgi-bin/db/db.cgi_displayItem=966 some godrive specs

Attempt 4:

Empty formatted disk - based off yellow disk boot format? https://fejlesztek.hu/create-a-fat-file-system-image-on-linux/

dd if=/dev/zero of=HD0.img count=50 bs=1M
fdisk HD0.img

Command (m for help): **o**
Building a new DOS disklabel with disk identifier 0x46ac6035.

Command (m for help): **n**
Partition type:
  p primary (0 primary, 0 extended, 4 free)
  e extended
Select (default p): **<Enter>**
Using default response p
Partition number (1-4, default 1): **<Enter>**
First sector (2048-99999, default 2048):
Using default value 2048
Last sector, +sectors or +size{K,M,G} (2048-99999, default 99999): **<Enter>**
Using default value 99999
Partition 1 of type Linux and of size 47.8 MiB is set

Command (m for help): **t**
Selected partition 1
Hex code (type L to list all codes): **c**
Changed type of partition 'Linux' to 'W95 FAT32 (LBA)'

Command (m for help): **w**
The partition table has been altered!

Syncing disks

mkfs.vfat test.img

Attempt 5:

imgflpyd.exe a HD0.img

Also experienced a read error Going to attempt to fix the disk, but not on precious yellow disk.

Getting "good spec" disk, creating image backup Wrote yellow to goodspec Ran chkdsk fix Using imgflpyd to create image Attempt boot from chkdsk version

Invalid Record SX

Writing unmodified rosie boot image to goodspec disk "please work" otherwise yellow disk is MAGIC

Attempt 6:

Using StaubliRX90\V+\FLOPPY COPIES\WHAT I AM CURRENTLY RUNNING\USE THIS! ROSIE\ROSIE.imz burn to goodspec disk with winimage and attempt boot

SUCCESS! This image is good. Transfer to HD0.img and expand...

Used winimg to expand fs from 1.44mb to 144mb Set scsi2 on, since the drive supported it according to the refs Turned debug off Enabled prefetch because SPEED BBY

SUCCESS!! It booted off c

Unfortunately the calibration is wrong. Verified yellow disk has proper calibration

Attempting to merge them

Copied over config_c and calib from yellow disk into HD0 Still not working... Uh?!

copied adeptsys.sys and adeptsys.cfg

Now all files that seem relevant have the same modification date Ah invalid record SX! restoring adeptsys.sys and adeptsys.cfg from rosie.imz

Going a lil nuts here copied rosie.spc to config_c/CFG01_14.SPC

Works!

PreviousNext