Jump to content
IGNORED

8" floppy drive needed for text to speech source


Recommended Posts

I have found a collection of 8" floppies. They are backups of text-to-speech utilities as they were developed at TI over 1981-1983. The backups were made from a 990/10 running DX10.

 

So I am in need of a 8" floppy drive, to connect to my Kryoflux controller.  This is a proven solution for extracting a sector image. I would construct the 50-to-34 pin floppy cable. The files can then be extracted by Dave Pitts' disk utility, and used in his 990 simulator assembling under SDSMAC. . (Or extracted from floppy images as ordinary text file sources.)

 

Does anyone have an 8" drive they could lend me?  (I'm not too comfortable mailing the disks.)

 

The only lead I had was an offer for that TI WD500 (8" floppy, no hard drive) for $420 shipped. 
 

  • Like 2
Link to comment
Share on other sites

Nice find. I'd be interested in the backstory, if you're able to share it.

 

Are these single or double density disks? I don't have any 8" drives here right now, but I'm looking. I'd have bid on the FD1000 in SC a while back but had no way to go get it or the 990/1s.

 

There are people in and around Austin working with old 8" floppies, Shugarts in particular. Ask on  the #medialab channel on the classiccmp discord. You might reach out to SAMSAT in San Antonio as well.

 

Good luck,

jbdigriz

 

 

 

Edited by jbdigriz
  • Like 1
Link to comment
Share on other sites

@jbdigriz I’ll try #medialab. Spending my own $400 on this quest is deferred. I’m pretty sure they are double density. 
 

The story of these disks. I found them in an otherwise unmarked TI binder in my dad’s study. I guess that he had them when the TI Lubbock West building was re-absorbed in 84 or 85, and its speech lab shut down.  
 

They’re not his work on the Barcode Factory for the Magic Wand Speaking Reader, as I had hoped. The floppy binder had been with a shelf of manuals, probably since 1987. Only one of the manuals is from TI, but it  is the user’s guide and technical reference for   Barcode Factory. This package was prepared in 1981-82 to enable 3rd parties to edit the allophone strings from text-to-speech, and to print books with bar codes for the MWSR. (There was a Runoff manual next to that…)
 

It’s kind of chilling that I had pulled more 8” floppies from the West trash dumpster. And tried to cut them down to size. Disks were expensive then!  Those might have contained anything…
 

(TI Lubbock West was a small building near the Pyramid. The little I know is that from 1982 it had a speech lab. And also  the project to reverse-engineer an Intellivision, to see if any TI titles could be made for it.)
 


 

  • Like 2
  • Thanks 1
Link to comment
Share on other sites

Cool.  I can  see how you might not want to risk the disks to shipping. Ideally, though, if you could get them to someone with a 990 and an FD1000, like Dave Pitts, you might not need the Kryoflux, if and when you locate a drive.  BTW the folks in #medialab seem to be pretty down on Kryoflux. Issues with corruption on writing images to disk, closed source, etc. They recommend other products. I've never used any of them personally, so I can't say. 

 

ISTR, there were some ISA bus floppy controllers that had the Shugart SA800/801 interface and did 8" formats, maybe some of the retro hardware projects do also, using PCI(e) or USB. Have to look around. Pretty sure the TI disks will be soft-sectored. Klaus or Piero can probably tell you. Hard-sectored (SA801, which is just some dip switch settings on the drive as I understand it) will have sector marking holes in the disks, of course. That way though you could at least dd an image if the disks are in good shape, or even use Dave's utilities directly, once you find a drive. 

 

But, yeah, won't hurt to just set them aside for now. Good luck, though, and please keep us posted.

 

Thanks for the history,

jbdigriz

Edited by jbdigriz
  • Like 1
Link to comment
Share on other sites

On 8/26/2022 at 9:31 AM, jbdigriz said:

Cool.  I can  see how you might not want to risk the disks to shipping. Ideally, though, if you could get them to someone with a 990 and an FD1000, like Dave Pitts, you might not need the Kryoflux, if and when you locate a drive.  BTW the folks in #medialab seem to be pretty down on Kryoflux. Issues with corruption on writing images to disk, closed source, etc. They recommend other products. I've never used any of them personally, so I can't say. 

 

ISTR, there were some ISA bus floppy controllers that had the Shugart SA800/801 interface and did 8" formats, maybe some of the retro hardware projects do also, using PCI(e) or USB. Have to look around. Pretty sure the TI disks will be soft-sectored. Klaus or Piero can probably tell you. Hard-sectored (SA801, which is just some dip switch settings on the drive as I understand it) will have sector marking holes in the disks, of course. That way though you could at least dd an image if the disks are in good shape, or even use Dave's utilities directly, once you find a drive. 

 

But, yeah, won't hurt to just set them aside for now. Good luck, though, and please keep us posted.

 

Thanks for the history,

jbdigriz

I emailed Dave Pitts first.  I'll email again. 

 

The memo with the binder says:

 

"These DSDD floppies provide 1 mega-byte of formatted space. This corresponds to roughly 1000 ADUS of a T-50 disk."

 

"use DSDDINIT"

"use IV" to install volume

"use UV" to uninstall volume

 

below,  lowercase means pencilled notes:

 

INITIALIZE DSDD DISKETTE

UNIT NAME: DS05

VOLUME NAME: kmg.src kmg.misc kmg.

NUMBER OF VCATALOG ENTRIES: 31     100 for src

DEFAULT PHYSICAL RECORD SIZE: 288

INITIALIZE DISK SURFACE?: YES

LISTING ACCESS NAME:

 

 

I wonder if Listing Access Name is an optional catalog filename? Home computer used "DSK1." to access the listing. Home computer reserved 30 catalog entries to start with.

 

"The system will use two of the entries for VCATALOG itself and for the S$DIAG file"

 

"YES to initialize disk surface will clear the disk and analyze it for bad tracks; answering NO will simply initialize the allocation map, the volume name, and the VCATALOG directory, but leave the remaining information on the disk without file name attachment. The Listing access name is strictly optional."

 

"If you attempt to use a disk which has not been formatted and initialized, you can expect to wait up to 2 minutes for the system to respond to you with the error code.  ... It will require approximately 15 minutes to initialize a diskette for the first time or whenever the answer to INITIALIZE DISK SURFACE? is YES.   Initializing a disk without initializing the entire surface will require only about 30 seconds."

 

 

 

 

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

Haven't talked to Dave in a while; hope everything's ok with him. I'm getting 404's and content encoding errors on some of his 990 links. Probably he's just been busy with other things and hasn't had time to maintain the site; I can relate. Please let us know if you hear from him, though.

 

Regarding those instructions, I don't think you'll want to be initializing these floppies, unless they're just blank, but I'm not familiar with "DSDDINIT". Maybe just a DX10 synonym for "INV" (Initialize Volume). Or an SCI batch file invoking INV. However, the UNIT NAME of "DS05" suggests these instructions are maybe for a WD500 floppy. Still digging though for what UNIT NAME that would be for certain.  FD1000 would have a UNIT NAME of "DKxx" and INV doesn't work with those. TXMD will format a floppy in a DK device as a TXDS diskette, but here you'd be putting DX10 volumes on the floppy.

 

LISTING ACCESS NAME is just the output device for the command, defaults to whatever you're logged in on and issuing the commands from.

 

If you get hold of a drive, there are also several places that sell Shugart 8" 50-pin <-> 34-pin adapters for standard PC floppy controllers. They sell for around $25 or so. Search the usual places. Another scenario for using Dave's utilities.

 

Please do let us know if there's anything useful and shareable on the floppies, whenever you do get them read.

 

jbdigriz

 

ps. haven't forgotten about the 990/1 pics, just been stacked up here. Hopefully Real Soon Now(tm).

 

Edited by jbdigriz
  • Like 1
Link to comment
Share on other sites

Oh wow, I didn’t search for pre-made 50 to 34 pin Shugart adapters! The blog I linked to showed how the author made one. I figured on doing a quick PCB,  with the possibility to cut traces later. 
 

If I do somehow get the WD500 with 8” floppy, what brand of drive  would you expect TI put into it? Shugart? 
 

Im ignoring any leads on drives that are not DSDD or not 110V 60Hz (no way am I dealing with a spindle speed conversion!!!) I have one possible offer of a TEAC. 

 

 

 


 

 

Link to comment
Share on other sites

Well, not sure. The Shugart SA850 is basically the DSDD version of the SA800, but the doors look even less like the one in the WD500. Eject button and mechanism look similar, so I don't know.  Unfortunately the WD500 manuals don't seem to be online. I'll keep digging, though. 

 

I'm thinking anyway that that $400 WD500 without HD would be overkill for your purpose here, unless you're planning on getting a BS300 system or a 990 (which will need to have a TPBI controller board for the WD500 interface. The WD500 is basically a whole computer in itself with it's own formatter board for the drives that interfaces directly with the floppy. For the benefit of those reading along who aren't familiar.) So, probably best just to get a drive alone, if you can figure for sure which one it is, or one that's compatible, and talk to the drive with a PC using the utilities, or use a flux imaging tool or the like. 

 

Good luck with it.

jbdigriz

Edited by jbdigriz
  • Like 2
Link to comment
Share on other sites

8 hours ago, FarmerPotato said:

If I do somehow get the WD500 with 8” floppy, what brand of drive  would you expect TI put into it? Shugart? 

Got it. It's a Qume 842 (Data Trak 8), same as in the FD1000. I even have a photocopy of the 842 maintenance manual right here. Dunno why it didn't ring a bell earlier. My day for brainfarts, I guess.

 

Here's a useful site with more info and links to manuals, which are on bitsavers as well, iirc: https://www.parastream.com/support/qume-842.aspx

  • Like 2
Link to comment
Share on other sites

  • 2 weeks later...

Minor addendum: Further digging through various manuals reveals that some FD1000s were equipped with the the Shugart 850/851 drives, some with the Qume DT-8s. Likely that might be the case with the WD500s also. In any case, disks from one should be readable on the other, assuming correct alignment and index timing. Also, the Shugart drives seem to have held up better over the years, judging from reports of others who've used both in other machines.

 

@FarmerPotato knows all this, of course. I'm just removing a foot from my mouth here. :-)

  • Like 2
Link to comment
Share on other sites

  • 9 months later...

No luck so far reading the floppies. No luck reading the first two with Kryoflux. 
 

Note: these were written on a 990/10 with floppy drive. It makes 288-byte sectors. That's not impossible under "recommended" Shugart encoding, with MFM. It's just... weird. 
 

@ForgottenMachines tried reading both of them with his SuperCard controller. He has built an amazing rig of every kind of floppy drive! We got raw flux images (.SCP) but....

 

They don't decode as FM or MFM. Using HxC utility, only format AECD6000 makes partial sense out of them. Fails CRC. No data recovered, but recognizable tracks and sectors. 
 

Unfortunately, these big red smears, I'm guessing, are evidence the disks were de-Gaussed. AJ has seen many disks with recoverable damage, but never this bad.   
 

Whether that happened at TI or sometime later, who knows. I bet that the later floppies have this too.
 

So I started writing a flux decoder, from the SCP raw format (which HxC likes.)  

Merely looking at the flux signals, it sure looks like sector-sized pulse trains, with occasional glitches. (I used Audacity Import Raw to view them as PCM 16-bit samples. Totally wrong,  but revealing.) 
 

We suspect that the DX10 floppy format is like none used elsewhere.
 

 

 

  • Like 4
Link to comment
Share on other sites

If i get one of these 2 TFDC boards working then i'll lend it to you but at the moment 1 is failing it's self test.

The other one passes the self test but while it appears to polling the 4 drive selects it never turns the motor on.

I think this polling is part of the drive ID detection.  

I've started to disassemble the ROM code on the TFDC to see what it's trying to do.

 

Jim

 

 

  • Like 1
Link to comment
Share on other sites

On 6/26/2023 at 5:49 PM, FarmerPotato said:

(I used Audacity Import Raw to view them as PCM 16-bit samples. Totally wrong,  but revealing.) 

I dunno.  I have seen people using FFTs to visualize disk data before, which was elucidating and helpful.

  • Like 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...