On Tue, Mar 23, 2010 at 6:31 PM, Pete Turnbull <pete at dunnington.plus.com> wrote: Thanks, both! I have a few bare drives knocking around, some of which are, I hope, elderly enough that they have DS jumpers. If not, I don't think I fancy trying to make this cable up... 8?( -- Liam Proven ? Profile: http://www.linkedin.com/in/liamproven Email: lproven at cix.co.uk ? GMail/GoogleTalk/Orkut: lproven at gmail.com Tel: +44 20-8685-0498 ? Cell: +44 7939-087884 ? Fax: + 44 870-9151419 AOL/AIM/iChat/Yahoo/Skype: liamproven ? LiveJournal/Twitter: lproven MSN: lproven at hotmail.com ? ICQ: 73187508

On Tue, Mar 23, 2010 at 2:19 PM, Gene Buckle <geneb at deltasoft.com> wrote: I'm arriving late to the conversation, so maybe I missed something. High density floppies have a square hole opposite the right protect tab up in the corner. Most 3.5" floppy drives have a switch that detects this hole and does (here's my hardware knowledge black hole) something different with the media. 720k double density floppies don't have this hole. So you should be able to simply cover the holes with tape or disable the switch the right way and the drive will treat the media like a double density disk. read/write head is thinner which I believe causes problems when you put that disk back into a real double density drive and one of the drives is slightly misaligned. Also, I think the actual media may be different, as in different magnetic properties. There's my highly technical explanation of making HD disks work like DD disks. :-P brian

On Tue, 23 Mar 2010, Gene Buckle wrote: 3.5" drives don't change speed; they always run at 300 RPM. That's why HD diskettes hold 1.44MB instead of 1.2MB. The average PC HD 5.25" *also* doesn't change speed, but runs at 360 RPM instead of the usual 300 RPM. That's why you need a data separator that runs at 300 KHz instead of 250 KHz to read a DD diskette in the average PC HD 5.25" drive. The DEC RX33 5.25" drive *does* change speed. It runs at 300 RPM for DD media and 360 RPM for HD media. -- roger ivie rivie at ridgenet.net

On Tue, Mar 23, 2010 at 3:32 PM, Roger Ivie <rivie at ridgenet.net> wrote: Ordinary "PC" 3.5" drives don't change speed, but Amiga high-density drives do - they spin at 150 RPM when high-density media are inserted, maintaining the double-density data rate. This is because it was judged more economical to make a custom drive than to design a new Paula chip (for those that don't know, unlike in a PC, with the Amiga, there is no single "floppy controller" per se - the necessary functions are performed by a combination of CIA I/O pins, a massive shift register in the Paula chip and logical miniterms applied by Agnes to decode/encode the data to/from MFM). One advantage of changing the drive vs changing the chip is that any Amiga can take an HD drive, not just new models (the HD drive came in around the same time as the A3000, IIRC). The official drive may be the Chinon FZ-357A, but I am not in the room with an Amiga to verify that. It's unlikely to run across one of these drives outside of an Amiga context, but they did exist (and I do have a couple of them - they even work under emulators like A-Max and Shapeshifter to read high-density (FDHD) Mac disks). That's interesting to note, considering the RX33 is a specifically-strapped Teac FD55GFR (usable on a "regular" PC when suitably re-strapped). -ethan

On Tue, Mar 23, 2010 at 7:19 PM, Gene Buckle <geneb at deltasoft.com> wrote: I wasn't planning on trying to format, access or otherwise use 1.4MB disks as 1.4MB. Just format them to /less/ than capacity and treat 'em as normal. The machine is a Sinclair Spectrum, although I also have a QL it would be fun to revive this way. On neither is there any option to select /how/ you format the disks, I believe. Both only support 80-track DS DD drives & disks. The Spectrum+DISCiPLE put 780K or so on a disk, I think. I've never tried to attach my QL to its disk interface... -- Liam Proven ? Profile: http://www.linkedin.com/in/liamproven Email: lproven at cix.co.uk ? GMail/GoogleTalk/Orkut: lproven at gmail.com Tel: +44 20-8685-0498 ? Cell: +44 7939-087884 ? Fax: + 44 870-9151419 AOL/AIM/iChat/Yahoo/Skype: liamproven ? LiveJournal/Twitter: lproven MSN: lproven at hotmail.com ? ICQ: 73187508

CORRECTION: SOME (NOT ALL) "1.2M" drives switched back and forth between 300 and SOME (NOT ALL) "1.2M" drives stayed at 360RPM, requiring 300K and 500K "360K" drives were 300 RPM, requiring 250K bits per second

HP used them _a lot_ which is where I mostly come across them, but other manufacturers (for example ACT/Apricot) used them too. Why the 'ich'? They are not difficult to keep going, -tony

On 23 Mar 2010 at 21:26, Pete Turnbull wrote: Ever use much PC98 Japanese gear? They use 3.5" HD floppies that spin at 360K, just like 5.25" HD and their 8" cousins. It makes a lot of sense--the format doesn't change, just the physical size of the medium does. You see these a lot on Japanese and Korean CNC and lab gear, as well as the veneered and generated NEC PC9801 series of personal computers. Most makes of 3.5" drives can be jumpered or switched (although it may not be documented very well) to run at 360 RPM and some--such as most USB floppies and Imation Superdrives can detect the recording density and switch automatically. Related to this, I received two floppies in for "copying" a couple of weeks ago--a 3.5" DSHD and a 32 sector 8", with the admonition that both floppies contained exactly the same information. (Hear the head scratching?). The recording method was clearly MFM, but with unique address marks and track layout. The hard-sectored 8" turned out to be a red herring--the layout of data on the track didn't follow the sectoring at all--I suspect that someone jumpered the 8" drives for HS operation so that they could use a stock of old 32-sector media. That way, you still get only one index pulse (but 32 sector pulses on a different pin). But the 3.5", while the spitting image of the 8", wasn't identical. It had been recorded at 300 RPM, so there was about 2000 bytes of information on the last half of each track that duplicated that on the first half--and, predictably, the last of the duplicated sectors was truncated at the index. Probably someone's hack. Personally, I'd rather stick with the 8" media, as it's far more robust than the 3.5" stuff. If 3.5" drives were required, I'd run them at 360 RPM. Cheers, Chuck

On 23 Mar 2010 at 23:59, Pete Turnbull wrote: Some Samsung and Mitsubishi drives document "1.6MB support" jumpers, so you'd want that. But if you're like me, I have buckets full of Teac FD235-HF and -HG drives, particularly those with a drive select jumper and nothing else. On these, find the pads labeled S3 and S4. You'll find a 0-ohm SMT resistor installed on S4 and an ohmmeter will show that it's connected to the +5 line. If you check the unpopulated S3 pads next to it (usually), you'll find one is connected to one of the S4 pads and the other is connected to pin 2 of the drive interface header. The usual way of adapting these is to replace the S4 jumper with a 1K SMT resistor and jumper the S3 pads with a 0-ohm resistor. This gives the spindle speed control to pin 2, which is an option if your system uses pin 2 as a density select pin. I usually remove the S4 jumper completely and run a wire between the pad that's not connected to +5 and ground. This hard-configures the 360RPM spindle speed without regard to pin 2. --Chuck

I am not sure it applies to all FD235s -- Teac are well-known for producing very different drives with slightly different model numbers (I can't rememebr how many toally different FD55s I've worked on...). The service manual I have for the FD235 doesn't explain the jumpers (annoyingly!), it just gives scheamtics where some of the jumpers just l;ink to pins on an ASIC. But it does show a 300/360rpm control input to _some_ of the spindle motor schematics, but not all of them. -tony

Subject: Re: Reusing old floppy *drives* The 5.25" FD55GFR was found in various qualities. In my conversion systems I used them for e.g. AS400. At one point, I got a series of failures on those drives. It showed that the drives made in India couldnt change reliably between 300 and 360 RPM, but the FD55GFR-142U series from Japan had no problem.

On 23 Mar 2010 at 18:31, Pete Turnbull wrote: D'oh! I picked up the closest pinout chart at hand, and wouldn't you know it--it was a PeeCee pinout. 10-11-12 it is. A mini-twist. --Chuck

Err, the IBM twisted cable did more than that. It also swaped MotorOn/ with (IIRC) DS0, so that the MotorOn/ signals for the 2 drives on the cable ended up on differnt pins at the controller end. That way the cotnroller could control the 2 motors independanty. Ulness the cotnroller is expecting that, an IBM OPC twisted cable is not going to work. What will work is a cable that leaves all signals apart that each drive select output of the cotroller ends up on the 'active' drive select signal (normally DS1/ for a PC drive) of a differnt drive. I am not convinced that was the real reason. After all, with the 5.25" full-height drives in the PC you had to fit a terminating ressitor package to the last drive on the cable (And yes, IBM did get that right). Anyone who could fit/remove resstor packs correctly could set jumpers. I still think the real reason was to have separate motor-on signals for the 2 drvies. The fact that it allowed all drives to be strapped the same way is a bonus. -tony