Replacing floppy
in [Design]
|
From: Nico Coesel on 25 Apr 2010 10:01 John Tserkezis <jt(a)techniciansyndrome.org.invalid> wrote: >Adrian C wrote: > >> On 25/04/2010 03:53, George Jefferson wrote: >>> I have an old keyboard that uses floppy to hold its os. It takes about 3 >>> mins to fully load if that. Also the disk can become corrupt overtime. > >> Ah, Musical keyboard. I was initially puzzling over what computer >> keyboard would have such a need. > >>> I was thinking of replacing the floppy with a ssd type of system. What >>> I'm thinking is possible is to "hijack" the floppy interface cable and >>> simulate a floppy disk but provide a faster system. Essentially >>> emulating the floppy disk protocol, which I imagine it uses some >>> existing standard, but reading off a ssd/eeprom. > >> How about trying a SmartMedia flash card and one of these - >> http://en.wikipedia.org/wiki/FlashPath > > This won't work, as it needs drivers installed first. I'm guessing >since the FDD is the only data storage interface available, this isn't >going to be an option. > > > On the little reading I've done so far, the situation can get quite >complex. There ARE floppy interfaces that supply USB, or CF flash >replacement storage devices, however, they are priced quite expensively, >from what I've read $500-$1000+ US, depending on features. > > Also, I've read one person who had a similar problem with a musical >keyboard, where replacing the FDD with a standard PC model did not fix >the problem. True, we don't know if the problem wasn't with the disk, >however, there WERE various supplies of FDDs that appeared the same on >the outside, but the interface was at least a little different. There are many ways an FDD can be controlled. Older floppy drives usually have many jumpers to accomodate different controllers. -- Failure does not prove something is impossible, failure simply indicates you are not using the right tools... nico(a)nctdevpuntnl (punt=.) --------------------------------------------------------------
From: Mark Zenier on 25 Apr 2010 14:27 In article <hr16eu$90e$1(a)news.albasani.net>, Jan Panteltje <pNaonStpealmtje(a)yahoo.com> wrote: >The serial protocol is either MFM or FM, NRZ code, with CRC added for >each sector. Not NRZ. For both input and output, it's a short (100-300 ns) pulse for each flux transition. For the input, inside the drive it's the clock for a T flip-flop that (gated by the enables) drives the head. For output, it's a one-shot triggered by a peak detector. Mark Zenier mzenier(a)eskimo.com Googleproofaddress(account:mzenier provider:eskimo domain:com)
From: Mark Zenier on 25 Apr 2010 14:45 In article <hr0arr$64l$1(a)news.eternal-september.org>, George Jefferson <phreon111(a)gmail.com> wrote: >I have an old keyboard that uses floppy to hold its os. It takes about 3 >mins to fully load if that. Also the disk can become corrupt overtime. > >I was thinking of replacing the floppy with a ssd type of system. What I'm >thinking is possible is to "hijack" the floppy interface cable and simulate >a floppy disk but provide a faster system. Essentially emulating the floppy >disk protocol, which I imagine it uses some existing standard, but reading >off a ssd/eeprom. > >Am I on the right track? Here the biggest problem is probably getting the >protocol correct and the electronics would be rather simple? Probably can be >done with a pic and not much more... If you want it read-only, fed by flash, it wouldn't be too hard. A couple of outputs to fake the drive signals (index, track 0, etc). Inputs for step, direction, drive enable, motor on, etc. The datastream is documented in the floppy interface chip datasheet, or there are various ISO standards for floppy formats that tell you enough information. The main problem may be in reading in your original so you can save it to flash. I suspect this may be one of those rare 3 inch Hitachi hard shell diskettes. Few folks used them. Roland keyboards and the Amstrad computers. (Sony won that battle with their higher capacity 3.5 inch diskettes). As I remember, electrically, the 3 inch drives were the same as a double density (360kbyte) 5.25 drive. There is (or was) a web site for a guy in Silicon Valley who had a unit where you can replace the floppy drive (for various classic home compters) and up and download floppy images from a PC. Search for "Semi-Virtual Diskette" or "SVD". Mark Zenier mzenier(a)eskimo.com Googleproofaddress(account:mzenier provider:eskimo domain:com)
From: Jan Panteltje on 26 Apr 2010 15:06 On a sunny day (Sun, 25 Apr 2010 18:27:31 GMT) it happened mzenier(a)eskimo.com (Mark Zenier) wrote in <hr4f1h073i(a)enews2.newsguy.com>: >In article <hr16eu$90e$1(a)news.albasani.net>, >Jan Panteltje <pNaonStpealmtje(a)yahoo.com> wrote: >>The serial protocol is either MFM or FM, NRZ code, with CRC added for >>each sector. > >Not NRZ. For both input and output, it's a short (100-300 ns) pulse >for each flux transition. For the input, inside the drive it's the >clock for a T flip-flop that (gated by the enables) drives the head. >For output, it's a one-shot triggered by a peak detector. > >Mark Zenier mzenier(a)eskimo.com >Googleproofaddress(account:mzenier provider:eskimo domain:com) I think you are wrong: http://en.wikipedia.org/wiki/Modified_Frequency_Modulation <quote> As is standard when discussing hard drive encoding schemes, MFM encoding produces a bit stream which is NRZI encoded when written to disk. A 1 bit represents a magnetic transition, and a 0 bit no transition. <end quote> I should know, I designed a multi standard floppy controller board, and it works :-) It has one of the lowest error rates in the universe, ZERO. Want the diagram?
From: Grant on 26 Apr 2010 19:04
On Mon, 26 Apr 2010 19:06:09 GMT, Jan Panteltje <pNaonStpealmtje(a)yahoo.com> wrote: >On a sunny day (Sun, 25 Apr 2010 18:27:31 GMT) it happened mzenier(a)eskimo.com >(Mark Zenier) wrote in <hr4f1h073i(a)enews2.newsguy.com>: > >>In article <hr16eu$90e$1(a)news.albasani.net>, >>Jan Panteltje <pNaonStpealmtje(a)yahoo.com> wrote: >>>The serial protocol is either MFM or FM, NRZ code, with CRC added for >>>each sector. >> >>Not NRZ. For both input and output, it's a short (100-300 ns) pulse >>for each flux transition. For the input, inside the drive it's the >>clock for a T flip-flop that (gated by the enables) drives the head. >>For output, it's a one-shot triggered by a peak detector. >> >>Mark Zenier mzenier(a)eskimo.com >>Googleproofaddress(account:mzenier provider:eskimo domain:com) > >I think you are wrong: > > http://en.wikipedia.org/wiki/Modified_Frequency_Modulation > ><quote> > As is standard when discussing hard drive encoding schemes, > MFM encoding produces a bit stream which is NRZI encoded when written to disk. > A 1 bit represents a magnetic transition, and a 0 bit no transition. ><end quote> > >I should know, I designed a multi standard floppy controller board, >and it works :-) >It has one of the lowest error rates in the universe, ZERO. >Want the diagram? I would like to replace the floppy interface in a TDS3034 DPO with some USB or Ethernet interface. Emulating a floppy is a possible method? Would a PIC chip do that? Hang off the floppy drive cable, intercept and interpret the signals? Make the DPO think it's talking to a real floppy disk? Grant. -- http://bugs.id.au/ |