Last updated June 2 2008 - (c) Herb Johnson with permission of participants Introduction -------------- In Nov 2007, I had discussions with Mike Stein and Andrew Lynch of the Vector Graphic maillist about 100 TPI drives, as used by Vector Graphic and by Commodore. Turns out, Commodore built their own drive products, specifically their 8050 and 8250, from 100 TPI drives by OEM'ed (made by) Tandon, MPI and Micropolis. Commodore added their own electronics and removed the OEM's electronics. So in principle, the Commodore drives are candidates for conversion from Commodore use "back" to use as the respective Tandon, MPI or Micropolis drive models. Vector apparently used only Tandon or Micropolis drives. But the discussion was confusing while Mike, Andrew and I established most of the details. Bill Degnan, also on the Vector Graphic list, contributed additional discussion and information. Here's a summary of the discussion, with some editing and correction, in the form of a dialog. I'll add more information over time as well. - Herb Johnson Commodore drive mechanisms -------------------------------------- What I am talking about is *finding* a 100TPI drive in a Commodore [8050 or 8250] unit that is effectively missing some parts (by design) and, assuming you have those missing parts on a 48 or 96TPI drive of the same make and base model, stealing them off that drive and putting them on the 100TPI drive that is missing them. Maybe I haven't made it clear that the Commodore drives do NOT have ANY normal PC boards on them; there are two drives in the unit, of which one has no electronics whatsoever aside from the motors and head, and the other has a completely non-standard board that connects to both drives and interfaces to a separate microcomputer inside the drive case which in turn interfaces the drives to the computer via IEEE-488. There are no standard 34-pin cables or connectors anywhere like there are on most 5 1/4" drives. As to your Commodore information (boy, you've really been researching!) it's absolutely correct AFAIK, and I think that's what I had said (although I wasn't aware that they used two different series of Micropolis drives). Not counting the LP series which I know nothing about, [Commodore] made two 100TPI models, each with two drives: the single-sided 8050 and the double-sided 8250. They both used 100TPI drives with ~500MB per side, supplied by one of three manufacturers: Tandon, Micropolis and MPI, and apparently Micropolis supplied two different models of each giving eight different possible drives, all 100TPI. The difference between the 8050 and 8250 is that the 8250 has double-sided drives...The Micropolis drives are easily identified from the outside by their unusual two-way latch; the Tandon and MPI drives I've seen look identical and can really only be distinguished by looking into the slot and comparing it with [other drives] you have. - Mike Stein Specific drives used by Commodore -------------------------------------------- I found a Commodore document on line at http://www.zimmers.net/anonftp/pub/cbm/schematics/drives/old/8050/index.html Titled "Service Manual 8050 - 8250 Dual Disk Drives, June 1985, PN-314011-03" which has this information: File 8050-01.gif says: "8050 and 8250 are dual, 5-1/4 inch floppy disk drive units. the 8050 uses a 77 track, 100TPI, single headed drive. the 8250 ....100TPI double headed drive.... If you look at the other GIF'ed pages on this page, it says that the 8050 model used Micropolis, Tandon, and MPI "drive assemblies" A summary TXT file on 8050 and 8250 parts, by some guy William Levak, lists specific drive brands and models of drives used by Commodore. I added Commodore drive part numbers to Levak's list. 805003101 Micropolis 1006-II (8050) 805001601 Micropolis 1006-IV (8250) 825003901 Micropolis Safari (8050) 825003902 Micropolis Safari (8250) 805005901, 5902 MPI 101SM (8050) 805002901, 2902 MPI (8250) 805004701 Tandon (8050) 825001501 Tandon (8250) - Herb Johnson I have located a pair of Tandon drives from an 8050. I have confirmed that this is indeed a Tandon by matching the ROM used in the drive control board. Here are the cbm 8050 control board ROMS of a drive fitted with Tandon drives. location ROM # Part Number UL1 901482-07 901482-07 UH1 901482-06 901482-06 UK3 901483-04 901483-04 UK6 901487 901487 (Later, Bill said: "The MPI drives have the same ROMs as the Micropolis drives used in a given 8250. There is only one difference for the MPI and Micropolis drives: UK3 901483-03 901483-03 " Pics can be found at: http://www.vintagecomputer.net/commodore/CBM_8050 - Bill Degnan List of early Micropolis Drives ------------------------------- I discussed Micropolis 5.25-inch drives with Bill Degnan and Dan Roganti in April 2008. By reviewing the documentation we all had on Micropolis, we came up with this list. MICROPOLIS User Manual is Document # 100089-001, Revision 7 - MARCH 1978 MICROPOLIS (Maintenance manual) FLOPPY DISK SUBSYSTEM' schematics, release date is 9/77 Micropolis News newsletter, issue# 3, Jan.'79 MICROPOLIS Mod I = 35trk, 48tpi; lead screw is 4 threads per inch Mod II =77trk, 100tpi; lead screw is 8.33 threads per inch "formatted storage" is total of one or two drives. Dual drives used a Master board on one drive, an "add on" board on up to three additional drives. [subsystems with master card] 1057 - dual-disk, 35 trk, 48tpi, 287KB formatted storage (143KB per drive) 1053 mod II - dual-disk, 77trk, 100tpi, 630KB formatted storage (315KB per drive) 1053 mod I - dual-disk, 35trk, 48tpi, 287KB formatted storage (143KB per drive) 1043 mod II - single-disk, 77trk, 100tpi, 315KB formatted storage 1042 mod I - single-disk, 35trk, 48tpi, 143KB formatted storage 1041 mod II - single-disk, 77trk, 100tpi, 315KB formatted storage 1041 mod I - single-disk, 35trk, 48tpi, 143KB formatted storage Add-on slave drives - connected via Daisy-chain cable 1037 - addon dual-disk for 1057, 1033 mod II - addon dual-disk for 1053 mod II 1033 mod I - addon dual-disk for 1053 mod I 1023 mod II - addon single-disk for 1043 mod II 1022 mod I - addon single-disk for 1042 mod I 1021 mod II - addon single-disk for 1041 mod II 1021 mod I - addon single-disk for 1041 mod I Power supply configuration with A/C supply: 1057 (1037), 1053 (1033), 1042 (1022), 1043 (1023) without A/C supply: 1041 (1021), Accessories 1091-01 - Regulator kit for S-100 bus for 1021/1041 drives with Power cable B 1083-02 - Interface cable A, 2 connectors, controller to 1 drive 1083-03 - Interface cable B, 3 connectors, controller to 2 drives 1083-04 - Interface cable C, 4 connectors, controller to 3 drives 1083-05 - Interface cable D, 5 connectors, controller to 4 drives 100085, 100086 & 100087, S-100 floppy controller 1071 "B" model 4/1977) 100070, 100071, 10072, floppy drive single "A" card (7/1977) 100163, 100164, floppy drive single "B" card (5/1978) 100059, 100060, 100061 floppy drive dual "A" card (12/1977) Tandon TM-100 series drives; 96 TPI or 100TPI? Other differences ---------------------------------------------------------------- Tandon docs say the -3 and -4 are 96 TPI drives; the -3M and -4M are 100 TPI drives. -3's and 3M's are single sided, -4's and -4M's are double sided. But there ARE reports that some 80-track 100TPI Tandon drives were MISLABLED as -4's. - Herb Johnson, Mike Stein The Tandon TM-100 manual which was reprinted by Vector Graphic, has schematics marked "96/100TPI, Rework, Unipolar logic" or "...Bipolar logic". The differences are in the read write electronics, and have no effect on stepping or tracks per inch. - Herb Johnson Herb; yes, there are differences in the controller PCBs due to the different kinds of stepper motors used to move the drive head. Most of the TM100 drives I have seen are the unipolar variety, however, most of the TM100-4M drives I know of are the bipolar stepper motor for some reason. Not all, but most. The big difference as I can tell is that the bipolar stepper motors need an H bridge since they have a simpler motor wiring scheme than the unipolar stepper motors. Here is a link on the various types of stepper motors: http://www.cs.uiowa.edu/~jones/step/types.html There is a fundamental difference which prevents interchangeability between the unipolar and bipolar models. The bipolar stepper motor requires the H bridge and won't work with out it and vice versa. - Andrew Lynch Andrew, my apologies. I was distracted by the redrawing of the schematics between the unipolar and bipolar designs.... Indeed, the unipolar model has FIVE wires, the bipolar FOUR wires, as I read the schematic. Absolutely different in that way. If you have not done so, you might check the DC resistance between pairs of lines to the steppers. The "unipolar" model has all four coils connected to a common ground, while the "bipolar" has two pairs of coils each in parallel, without a common connection. I believe these can be distinguished "in the field". The stepper motor connector is labeled "P12" and it is one of many connectors to the board's "J4". - Herb Johnson [Later, Andrew wrote:] Herb, There are definite differences between the bipolar and unipolar design. The bipolar and unipolar stepper motors are wired differently. I have tried to rewire the bipolar into a unipolar configuration but it never worked no matter what combination I tried. They are not interchangeable and the reason is the H bridge chip (ULN29074B) present in the bipolar controller PCB and not present in the unipolar design. Apparently the bipolar stepper motor requires current in both directions while the unipolar only requires it in a single direction. Drive conversion to 100 TPI use --------------------------------------- I do understand that the drive mechs are what determines the step (and head) size; that's why I'm suggesting taking a 100TPI mech from a CBM drive and putting on it the electronics from a 96 or 48TPI version of the same basic model [and brand of] drive... if the 96 and 48TPI boards _are_ the same as the 100TPI boards. And when I talk about using a board from a different TPI version of the "same" drive, I mean it in the sense that a TM100-1 and a TM100-4 are both versions of the "same" basic model [Tandon] TM100 drive. I'm not talking about swapping boards between a TM100-2 and a -4; that wouldn't make sense, especially if the boards are the same. And I'm not talking about mixing parts from different manufacturers either. - Mike Stein I (Herb) asked Andrew what kind of drive failures he found. Andrew said: So far, I have found various TTL chips to have failed such as 74LS74 flip flops and a 75462 NAND driver. There have been other failures, some I have diagnosed and some not. Often just swapping the controller and/or linear servo PCBs around will recover good drives and will concentrate the bad parts together. Some TM100 drives will eventually just not be recoverable but since the TM100-1A/-2As (48tpi) are common enough, I can tolerate some percentage of dead drives. The TM100-4M's on the other hand (100tpi) are so rare as to be worth more expensive and exotic repairs. - Andrew Lynch Vector and 100TPI drives ------------------------------- Apparently Vector also used 100TPI drives from Tandon and Micropolis and it's quite likely that they are the same mechs as used by [Commodore] I thought [Vector] also used MPI drives but after another look around it seems that that's not the case. I don't have docs for the VG controller, but I suspect (and leave for you to confirm) that the VG controller specifically worked with the Tandon -3M & -4M drives, and the Micropolis controller worked with the Micropolis drives. - Mike Stein (Note: VG offered a Micropolis controller, designed to work with Micropolis drives. VG also had their own controller design, and VG reprinted the Tandon TM-100 manual for the -3, -3M, -4, -4M (96TPI and 100 TPI) Tandon drives. Mike later said he's working with Micropolis 1015-IIB and 1016-II drives. - Herb Johnson) Micropolis floppy drive interface -------------------------------------- But as I said, it's worth noting that the 34-pin interface on the Micropolis controller (and matching [Micropolis] drives) in a Vector Graphic computer is NOT the same as on a TM100. I only have docs for the Micropolis controller and a Tandon TM100; since you have a better selection, compare the usage of those pins by the Tandon TM100 and VG controller vs. the Micropolis versions. I think you'll find that Tandon followed the [Shugart] "standard" and used pin 6 for /DS4 while pin 34 was used for /READY (at least on the TM100-3 & 4; previous models such as the TM100s used by Cromemco and others did not have a signal on pin 34). The Micropolis controller on the other hand looks for drive ready on pin 6 and uses pin 34 for DS4. As I mentioned on the VG list, I ran across this while adapting 96TPI HD Panasonic drives (JU475-4, used in PCs), which use 6 and 34 for DS4 and RDY respectively, for use in my Vector MZ with its Micropolis controller (and also in my Cromemco systems). And no, before you ask, they are still 96TPI, not converted to 100TPI somehow... ;-) - Mike Stein I reviewed the Vector Graphic copy of the "Micropolis Disk Controller Board" document for the Micropolis S-100 controller, apparently used in Vector Graphic "System B". Here are the pinouts for the Micropolis floppy disk controller to the floppy drive; and pinouts for the TM-100-3 and -4 docs I have from VG. - Herb Johnson Micropolis controller Tandon TM-100-3 and -4 schematics 2 HD LD head load not used 4 WCS not used not used 6 RDY ready DS 3 <-- Shugart DS 4 (Mike) 8 SECP sector pulse index/sector 10 DS1 drive select 1 DS 0 12 DS2 drive select 2 DS 1 14 DS3 drive select 3 DS 2 16 MTRN motor on? motor on 18 DIRN direction direction 20 STEP step step 22 WDA write data write data 24 WRT write gate write gate 26 TRK0 track 0 sensor track 0 28 WPRT write protect write protect 30 RDA read data read data 32 HDSCT head select side one 34 DS4 drive select 4 not used <-- Shugart drive ready (Mike) - Herb Johnson Don Maslin and Vector Graphic ----------------------------- Don Maslin provided bootable diskettes for many CP/M and Z80 type computers throughout the 1990's and later, until he died. Relevant to this Vector Graphic discussion, Andrew Lynch found the reference below by Don in a post to "cctalk". - Herb Johnson http://www.classiccmp.org/pipermail/cctalk/1998-January/093114.html Don posted: "I have substituted a half-high 720k 5.25" drive for the full high 100tpi drive on a 16-hard-sector Vector Graphic quite successfully." This strongly suggests that Don Maslin understood that Vector Graphic systems used 100TPI drives. Andrew Lynch suggests Don may have used 96 TPI drives on his VG system to create disks (says Andrew) "readable ONLY on his machine". I disagree, I think it only suggests that Don COULD read or write BOTH 96 TPI and 100 TPI disks as a matter of convenience and requests. Why 77 tracks, not 80? ---------------------- I (Herb) asked Andrew why 77 tracks were first used on these drives. He said, "I believe they went to 77 tracks because that is what the 8" disks used. The 100tpi drives were capable of 80 tracks or more but I think they limited it to 77 to simplify the OS coding and make the 5.25" minidisks and 8" disks similar. VG used only 77 tracks and I think the documentation for all the 100tpi drives specified 77 tracks." - Andrew Lynch