Speaking of formatting RL packs, one thing that I've wondered about is whether the interleaved servo bursts have the same radial width as the data tracks. If they have the same width, then it seems to me that it would be possible to heavily modify an RL02 drive to make it into an RL02 pack servo writer. But if the servo bursts need to be physically wider to span adjacent tracks properly, then a servo writer would need special heads with the proper magnetic gap width.
Someone with a crashed pack (or unusable for some other reason) and some Magnasee should be able to give you an idea - 5440 technology was 100TPI and should be visible. I think the RL02 was just a 200TPI version of the same drive guts.
The drives in my collection which use 5440 style packs include some RL02 drives, and the 6045 drive in my DG Nova 3 system (one removable platter and one fixed one). The 6045 drive includes an optical position sensor in the head positioner, so I think that implies that it could format a bulk-erased platter?
The 6045 is the DG-built replacement for the re-branded Diablo 44, DG model 4234. RDOS treats them identically (SYSGEN question "Top loaders?") and they are the same device ID for bootstrap purposes (100033, Program Load). It is probably possible to infer which model you're talking to - the source for HIPBOOT (the "new generation" RDOS bootstrap from 1978 or so) will probably show you if it is possible or not. Packs physically compatible with the 4234 (correct number of sectors*) were formattable with a diagnostic utility.
Those were the days when you'd recover from a head crash by wiping out the inside of the drive with paper towels soaked in (a now-banned) cleaning fluid, then change the heads and do a fly-by-night alignment with a regular data pack and a 'scope - when you see bits, you know you're on the outermost track. Adjust for best amplitude and button it up.
I expect that the 5-over-5 removable/fixed design was responsible for more data loss due to operator error than anything else of that era. To back up the disk, you needed 2 scratch cartridges, and then the dance went like this:
o Put the scratch pack in the top. Copy the bottom (fixed) to the top.
o Remove the scratch pack and insert the production top pack. Copy the top to the bottom.
o Remove the production pack and put the second scratch pack in the top. Copy the bottom to the top.
o Remove the second scratch pack and put the first scratch pack in. Copy the top to the bottom.
o Remove the first scratch pack and reinstall the production pack.
o Try booting and see if you made any mistakes...
The RL02 drives require packs with pre-recorded servo bursts for positioning; does that imply that there's no need for a careful alignment to be performed after replacing heads in the RL02 drives? I haven't gotten deep enough into either kind of drive yet to know the answers to those assumptions for certain.
If they're using industry-standard 5440 sliders (even with custom heads), there is enough "slop" in the head alignment slot that you can adjust them off the usable area of the media. Plus, if you have the top and bottom heads unaligned, changing surfaces is going to take as long (or longer) than a 1-cylinder seek on the same surface.
* My first experience with buying 3rd-party 4234 disk packs (from Minicomputer Accessories - later INMAC [INternational Minicomputer Accessories Corporation]) didn't go well. I apparently got some hippie stoners (said my up-tight 18-year-old self) - they answered the phone "Customer serrrrrrvvvvvviiiiiccccee. Can I ... help? ... you?" They shipped replacements which were the same wrong ones that they originally shipped - took a few iterations to get the right packs.