famicomaster2
Experienced Member
Haha, 200MB is an extreme stretch. For 2 heads at 5mbit (9x1024byte) sectors, most controllers will top out at 18MB formatted. A general limit of 1024 cylinders can be expected, as most software from this era that deals directly with the drives won't accept more than that. However, drives did exist with more cylinders than that, up to 1924 as far as I'm aware. That would garner you 33MB. If you jump to the typical controller-imposed limit of 2048 cylinders and use an RLL controller to get 26x512byte sectors, the limit would still be 52MB. Very few drives ever exceeded 100MB in MFM-land, and most of them were actually RLL and just had a crapload of surfaces. See: Maxtor XT series.I kind of wonder how hard it would be to use a single modern platter (to make it more simple) from a malfunctioning drive with some low tech heads and controller to make a "new" MFM drive. New drives with massive storage are complicated beasts but if you dial down the design where say a 2TB platter just needs to be 200 MB spinning at 3600 RPM or less you might be able to hack something up in a garage.
Not to mention that these controllers make no account for servo tracking, so you would end up writing your own servo data if you plan to use a voice coil and therefore lose one of your surfaces, 50% of the disk's capacity.
All this being said though, I highly support this idea on a lower level. I have had some mediocre success with making my own platters, and almost all components used in such drives are still made today in some capacity. I have been learning to sand cast to make my own chassis (though obviously this could be done with just aluminum stock and gasket maker, as proved by JCT). Building a PCB would be trivial, as many drives of the era were completely "dumb" like this JCT 100 or the Seagate ST-506, and use all off the shelf 74 series logic chips and passives. Going further you could probably integrate the entire drive plus a few parts with a single modern microcontroller.
The last real hurdle, for me at least, is the heads. Good luck finding "low tech" heads, and good luck making your own! 3D printing sliders is one thing, and stamping twinarms out of sheet steel would not be impossible, and I'm sure the microcoaxial cable and amps still exist in some form - But winding ferrite heads, even on the largest scale is by no means a trivial task.
If you plan to pursue this idea further, please PLEASE DM me or make a thread about it. I have been working towards that exact goal for a while now.