Це відео не доступне.
Перепрошуємо.
How Data is Encoded and Stored on Floppy Disks
Вставка
- Опубліковано 7 сер 2024
- A two part series: In this part i'll take a look at how data is encoded and stored, and read back on a floppy disk.
Make some notes, because the next video is all about copy protection!
Ever wanted to know how a floppy disk works?
Small Note: At (09:00) The Sync Word would actually decode to A9, but its called the "A1 Sync Word" If you decode/encode it you can still see that its not possible to make this sequence - Thanks to Keir Fraiser for pointing this out.
If you want to support my channel then consider becoming a Patreon!
Patreon: / robsmithdev
Discord: / discord
Chapters:
Welcome (00:00)
Introduction and Background (00:12)
Physical Storage (00:58)
Auto-Gain (02:27)
MFM Encoding (03:38)
PLL (06:53)
Index Signal (07:52)
SYNC Mark/Word (08:24)
Write Precompensation (09:05)
What is DrawBridge? (09:48)
Summary (10:23)
Music:
rotate picture by rush
technology by axel of brainstorm (1992)
Copyright © RobSmithDev 2021-2022
robsmithdev.co.uk
Where appropriate, all efforts were made to trace & obtain copyright
holders permissions to reproduce their material in this video.
If you have any questions regarding content please get in touch.
#floppydisk #mfm #encoding #vintagecomputing
You dont know how far i had to scroll past "content" to finally get to your video actually describing how floppys work.
Edit: re-read your message - glad you found it useful
I love learning
Absolutely brilliant. Thank you for all your hard work on the Drawbridge, Rob, and making Amigans’ dreams come true! Inspired by my A500, I did a degree in electrical engineering (1996-1999), and then ended up working in software. WHAT A WASTE! Glad to see you putting your knowledge to a good cause.
Very informative Rob. A real Floppy boss!
Bravo, finally a video with the correct explanation! Considering how important this invention was to human history and the vast amount of data stored using it there is surprisingly little correct information out there. I suppose you will explain the problems with too many zeros in a row in your second part, hence the reason for the need for encoding. The way I understand it is that a flux transition is like a 3D gaussian curve. When moving from one transition going up and the other going down the distance between them dictates how many zeros between the ones. But the more zeros inserted, the more the slope between the gaussian peeks becomes shallower and shallower. The more shallow it becomes the higher the auto gain amplifier ramps up. If there are too many zeros then the slope can become so shallow it almost becomes a constant straight line. Induction stops in the presence of a constant field so this is bad. The next amplifier after the auto gain one is configured as a differentiator amplifier that differentiates the signal and detects zero crossings by detecting when the flux reverses (ie the slope at a true flux reversal at the top of the gaussian peek is flat). The derivative of a shallow sloping, almost constant signal can be close to zero. So with the first amp with its auto gain maxed and the second differentiating close to zero, any noise can easily cause a random incorrect zero crossing to be sent through to the one shot and be incorrectly detected as a one. Some copy protection did this on purpose. When implementing Pi1541 floppy drive emulator I had to introduce random flux reversals to emulate this side effect of the two op-amps used in real drives. I'm developing a new emulator for shugart drives that is like a DrawBridge/SCP/Zoomfloppy and a Gotek combined together. It is a small device that sits behind a real floppy drive. It can intercept the signals from the computer and emulate the drive or it can pretend to be the computer and image disks or it can simply let the signals go through and allow you to use the real floppy as usual. All controllable wirelessly via a cell phone app.
Nice to hear that I got my facts right! You perhaps go into more detail than I planned, but yes your understanding of the amplifiers is the like mine. I guess in the video if I’d shown a threshold line for the signal level it would have covered the second stage, but I think it’s enough for the majority of what i was trying to get across. Your project sounds interesting. Do you want have a link about it?
I understand jack shait about flpopy disk operations oszyllos... somthing etc. but i managed with the Arduino and the DrawBridge project to
copy my Amiga stuff onto the PC. It was a mess made out of jumper wires, Arduino, floppy drive, FTDI break outboard, power supply companioned
by a enormous amount of excitement and maybe a bit of fear, caused of the possible magic smoke. There is nothing magic about the smoke it stinks
and is expensive! And i didn't see any that day! Thank you for your phenomenal work and sharing. 😀.
Letting the magic smoke out can be “educational” lol - glad you managed to grab your Amiga stuff!
Yes I got it right, 5 min tho to Wright it down for the first time
I did not know about the sync mark and I'm pleasantly surprised. It reminds me a bit of an overlong UTF-8 sequence, in a way.
The MFM system seems similar to how the old TRS-80 saved data to tapes. It recorded each bit over 4ms divided into 1ms intervals. The first, second and fourth interval were always the same. The first interval was a 1ms pulse, followed by a 1ms pause, followed by either another 1ms pause for a 0 bit or a 1ms pulse for a 1 bit, and the fourth interval would always be a 1ms pause.
The beauty of this scheme is that it would automatically re-synch to the next clock pulse on each bit.
Unfortunately, the whole thing was "bit-bashed" and there was an error in the timing routines which made reading back from tapes somewhat "flakey" which played a large part in earning the computer its nickname of "Trash-80" - a name which persisted even after the tape reading routine had been adjusted to compensate for the original error.
Ah that sounds like FM encoding, the predecessor to MFM, interesting!
Excellent informative video. Looking forward to the copy protection one.
Excellent info Rob...I have to admit to being a bit of a floppy addict and self described Archivist.Amongst my 'devices' so far I own a Kryoflux a Supercard Pro and a couple of Greaseweazles...and recently added a Shugart style full height 5.25 inch drive....and now I can finally go through my collection of about 700 old floppies already dumped from PET C64 VIC20 AMIGA APPLE II and PC to establish what can be reformatted and repurposed. Like LGR I'm endlessly fascinated by this medium.Keep up the great work Rob.
This is fantastic Rob, loving your channel and of course your awesome drawbridge, just updated to latest firmware and also it`s great that Tony has now incorporated Drawerbridge into Winuae it works a treat, and I love that you improved the USB floppy works well on USB2 even though as we know its underpowered as I am a cheapskate and use an old Laptop lol ,obviously using USB3 is the best way to go but thanks for looking after us tightwads :)
Great video. Thanks.
Just found your videos today. “Instant subs” are very rare for me (:
Thanks, and appreciated! If you have any feedback I’d love to hear it
Outstanding. Instant subscribe!
Interesting video, congratulations!
@2:00 How would you address then the flippy disks (5.25") we used on single-sided 8-bit Atari and Commodore drives (among others), where the second side was easily accessible by punching a write protection notch and just rotating the disk? In such case, the tracks were EXACTLY aligned, almost ideally - and everything worked, and still works. Disks recorded 30 years ago are still in perfect shape (if stored correctly).
So that video was made a few years ago, and whilst I’m still not 100% sure, on a 3.25” drive the upper and lower heads probably are writing in the same spots just offset by one cylinder. If this is true then I guess the level of magnesium they produce must be just enough and localised enough to only affect one side of the disk. If that’s the case then that explains how the flippy ones worked too. Every day is a school day and it’s amazing we’re (me included) still learning new things about this obsolete technology. I had to ask for help on that one
@@RobSmithDevOffsets are true for double head drives, also 5.25" ones (were the difference is 6 or 7 tracks actually). My guess is, it's to avoid interference between the tracks themselves. But the material, that the disk is made of, is enough to isolate the data.
How was Amiga High-Density floppy disk format different from PC HD? Were Amiga clock intervals the same for DD & HD?
Amiga and PC use different sector encoding formats but aside from that, the Amiga couldn’t handle the higher data rate so used special drives that when detecting a HD disk would spin the disk at half the speed meaning the data came in at DD speeds. An Amiga HD disk just had twice as many sectors compared to a PC. Also because the Amiga didn’t index align it’s data it could fit more on it, whereas a PC sector locations were precisely timed to start after some period from the index pulse. Because of this the spaces between sectors were larger to make them easier to find
But what about controllers without PLL? I.e. WoZ drive e.t.c. They measure time from the edge of the pulse, and have pretty good margin 2,3,4uS /*giving +/-1/4 us tolerance out of largest 4u interval they get +/-10% too, but actually there could be more like 1/3 even 1/2.5! and still read the interval right*/
Can you tell probably I've missed something?
Woz wasn’t using MFM encoding - I’d be surprised if there wasn’t some kind of PLL there somewhere though. Measuring from pulse to pulse does work quite well anyway - found that out first hand, but the PLL helps with poor accuracy of rotational speed.
Could you please help me? I'm a mum going back to work after 19 years of brining up kids and working for my husband. Suddenly, I need my CV! It's on a 19 year old floppy! If the floppy is not able to be read by the drive (it's a 20 year old 3.5" floppy) as it's making a clicking sound (I think it's not spinning), is it still possible to get the data of the disk? There's a bit of corrosion on the outside.
Hi, can you turn the disk yourself and is it easy? If not, then the disk is probably too bad to be used.
If it turns easily (could be a fault with the drive), then slide the shutter back on the disk, and hold it up to the light. If you cant see anything through it (or a small consistant amount of light) then yes its more likely. If however you can see little pin spots of light, or strange marks on the surface (almost looks like water marks) the chance is less likely. Either way the drive you're using will probably need cleaning (the drive heads).
If it's in IBM format, then these days you can still buy an external USB powered 3.5" drive which runs off a laptop USB port. So long as the disk itself is still readable and it's in IBM format, it should be readable through such a drive.
Is this MFM Amiga specific? ua-cam.com/video/OAswAoNLJhs/v-deo.html
Can AMIGA read PC floppy written by IBM MFM i8271?
Hi, not Amiga specific just anything that uses MfM. Was widely used on floppy disks including PC floppy disks and early hard drives too. Yes Amigas can read PC floppy disks
@@RobSmithDev Wow! I red i8271 data-sheet many times, it looks pretty complicated - different markers with missed clocks. Your MFM looks 100 times simpler, well done! Thank you
Data!