Reading floppy disk data, part 3! itsa pulse party

OK so part 1 we got our wiring worked out and found an index pulse. Part 2 is we started getting MFM data coming out. Now we’re capturing pulses with gpio bitbanging, and storing all the pulse widths in a large memory array. each track has a 500Khz signal, and outputs data every 5 Hz, so we have a max of 100K samples worst case.

since we’re using a cortex arm with 192K~256K of RAM (the RP2040 has 264K!), its no biggie to store all the pulses in an array. here we are reading track 1 and binning the pulses: we’re seeing three pulse width bins stand out, about 40 count, 62 count and 85 count. but there’s also a couple ultra short pulses (25 count) and an ultra long pulse (~200 count).

each count is about 48ns-ish so that translates to 2us, 3us and 4us bins, with a few 1us and one or two 10us+. not exactly sure whats up with those outliers. do we ignore them? are they start-of-data markers? We also noticed that track 0 has way more unusually long or unbinned pulses, like almost 100 different values. its a little mysterious, we’ll have to investigate if there’s something special about track 0!

Video.

Part 2:

Part 1:

from Adafruit Industries – Makers, hackers, artists, designers and engineers! https://ift.tt/3Juw6yY
via IFTTT