23 Jun
2018
23 Jun
'18
2:17 p.m.
Hi folks,
For the last few weeks I?ve been helping the owner of a PPG2.2 synthesizer get the
?computer? side of it, a Waveterm A, running.
The Waveterm is a 6809-based single board computer running FLEX that was designed to use
8? drives, then at some point they modified it slightly to use 5.25? HD drives instead but
changed very little - to connect up to the main board they made a 34-50pin adapter. The
drives are YE-Data YD380 with a Shugart interface so without some sort of signal bender
like a DBit FDADAP I couldn?t just plug them into a PC.
However, the YE-Data YD380B has an IBM interface so we used one of those, same mechanism
so we figured it should work.
All the disk images for the Waveterm have been created using programs written by PPG
users. For some reason any disk we wrote with them wouldn?t read correctly in the machine
itself UNTIL we made a new image of that disk using Teledisk 2.15 then re-wrote it back to
the same floppy. My question is why should that make so much of a difference between
working and non-working disks?
Trivia: for anyone aware of the 80s UK Music industry this very machine was used for the
hit ?You spin me round (like a record)? by Dead Or Alive.
--
adrian/witchy
Owner of Binary Dinosaurs, the UK's biggest home computer collection?
t: @binarydinosaurs f: facebook.com/binarydinosaurs
w: www.binarydinosaurs.co.uk
23 Jun
23 Jun
4:37 p.m.
On 06/23/2018 02:17 PM, Adrian Graham via cctalk wrote:
All the disk images for the Waveterm have been created
using programs
written by PPG users. For some reason any disk we wrote with them
wouldn?t read correctly in the machine itself UNTIL we made a new
image of that disk using Teledisk 2.15 then re-wrote it back to the
same floppy. My question is why should that make so much of a
difference between working and non-working disks?
I'd love to be able to tell you, but I really haven't fooled with the
thing for something like 17 years.
--Chuck
4:53 p.m.
> All the disk images for the Waveterm have been
created using programs
> written by PPG users. For some reason any disk we wrote with them
> wouldn?t read correctly in the machine itself UNTIL we made a new
> image of that disk using Teledisk 2.15 then re-wrote it back to the
> same floppy. My question is why should that make so much of a
> difference between working and non-working disks?
On Sat, 23 Jun 2018, Chuck
Guzis via cctalk wrote:
I'd love to be able to tell you, but I really
haven't fooled with the
thing for something like 17 years.
I had hoped that Chuck would have an answer.
If you have two disks that are nominally the same, and one works and the
other doesn't, then there is obviously something different.
If the sector contents are the same, then the next step would be to
examine the headers, gaps, and addressmarks.
For example,
if the gap after index pulse, before the first sector ofeach track is too
short, then it may not be able to read the sector header of the first
sector. (A serious and common problem with NEC-style FDC, not usually a
prolem with WD-style FDC)
Do you have a way to examine the raw track encoding?
I used a track read with a WD 179x (slightly modified "Trakcess" on TRS80
Model 3), and "TE" with the Central Point Option Board.
--
Grumpy Ol' Fred cisin at xenosoft.com
5:20 p.m.
On Sat, Jun 23, 2018 at 5:53 PM, Fred Cisin via cctalk <
cctalk at classiccmp.org> wrote:
All the disk images for the Waveterm have been created
using programs
Also, sector ordering can be an issue with more obscure formats... Though
most drives can cope with non-consecutive ordering...
In some cases, track width may also matter, especially for 40 track formats.
written
by PPG users. For some reason any disk we wrote with them
wouldn?t read correctly in the machine itself UNTIL we made a new
image of that disk using Teledisk 2.15 then re-wrote it back to the
same floppy. My question is why should that make so much of a
difference between working and non-working disks?
On Sat, 23 Jun 2018, Chuck Guzis via cctalk wrote: I'd love to be able to tell you, but I really
haven't fooled with the
thing for something like 17 years.
I had hoped that Chuck would have an answer.
If you have two disks that are nominally the same, and one works and the
other doesn't, then there is obviously something different.
If the sector contents are the same, then the next step would be to
examine the headers, gaps, and addressmarks.
For example,
if the gap after index pulse, before the first sector ofeach track is too
short, then it may not be able to read the sector header of the first
sector. (A serious and common problem with NEC-style FDC, not usually a
prolem with WD-style FDC)
Do you have a way to examine the raw track encoding?
I used a track read with a WD 179x (slightly modified "Trakcess" on TRS80
Model 3), and "TE" with the Central Point Option Board.
I used to mess around with different drives on different machines. These
days I find that kyroflux just works (on the right drive) and I don't have
to mess with weird things. It has a small learning curve, but it allowed
me, through brute force, to read the Rainbow Venix disks that were
otherwise not readable... It was ~$100 well spent...
Warner
7:03 p.m.
What kind of drives does the Waveterm use?
What kind of disk controller?
For example:
96tpi (80 tracks per side) drives can read 48tpi (40 tracks per side)
disks, and can write them to virgin disks, but, when RE-writing them,
leave part of the old (wide) track alongside. Those RE-written disks are
readable by a 96tpi drive, but not necessarily by a 48tpi drive.
PC is NEC-style FDC.
WD-style disk controllers (such as 179x) can handle post index gap smaller
than NEC-style can. For reading with NEC controllers, that can often be
handled by masking the index pulse.
WD-style controllers can read sectors that have a WRONG number in the side
number field, but NEC can't ignore that field. The good news there is
on formats that use a wrong number in the side number field (such as
Kaypro DS), the WD controller doesn't MIND if it encounters the correct
number in that field.
SOME NEC-style FDCs can not handle 128 bytes per sector. Some can.
The FDADAP adapter's primary difference between a simple cable is support
of TG43, a signal to indicate that it is on an inner track, for
write-precompensation, etc.
24 Jun
24 Jun
2:15 a.m.
On 24 Jun 2018, at 03:03, Fred Cisin via cctalk
<cctalk at classiccmp.org> wrote:
What kind of drives does the Waveterm use?
I mentioned that in the original message, YE-Data YD380 HD with Shugart interface
What kind of disk controller?
WD FDC1793. The main board is an Eltec Eurocom II V7.
For example:
96tpi (80 tracks per side) drives can read 48tpi (40 tracks per side) disks, and can
write them to virgin disks, but, when RE-writing them, leave part of the old (wide) track
alongside. Those RE-written disks are readable by a 96tpi drive, but not necessarily by a
48tpi drive.
The PC?s drive is simply the IBM version of the same drive, it?s a YE-Data YD380B. We
format a disk using the PPG utilitiy (PPGA.EXE) then write a disk image to it, reread that
disk with Teledisk and re-rewrite it again with Teledisk back to the same disk.
PC is NEC-style FDC.
The motherboard in the PC we?re using is an Abit board containing a Winbond SuperIO chip,
I can?t remember which variant but it seems most Athlon-based boards have one of these
chips on, my own imaging machine is an Abit KV8Pro with a Winbond W83627HF. What that is
based on I have no idea.
WD-style disk controllers (such as 179x) can handle post index gap smaller than NEC-style
can. For reading with NEC controllers, that can often be handled by masking the index
pulse.
WD-style controllers can read sectors that have a WRONG number in the side number field,
but NEC can't ignore that field. The good news there is on formats that use a wrong
number in the side number field (such as Kaypro DS), the WD controller doesn't MIND if
it encounters the correct number in that field.
So you?re saying that the WD controller is a lot more easy-going for slightly awry disks
whereas the NEC one is more strict? That could make sense in this case were we going from
WD to NEC but we?re going the other way around.
The FDADAP adapter's primary difference between a
simple cable is support of TG43, a signal to indicate that it is on an inner track, for
write-precompensation, etc.
For 8? drives yes, which at some point I will be using assuming both my 8? drives work,
but it should also let me connect up a non-IBM drive to my PC if I read the description
correctly.
Cheers,
--
adrian/witchy
Owner of Binary Dinosaurs, the UK's biggest home computer collection?
t: @binarydinosaurs f: facebook.com/binarydinosaurs
w: www.binarydinosaurs.co.uk
23 Jun
23 Jun
11:16 p.m.
On 06/23/2018 04:53 PM, Fred Cisin via cctalk wrote:
I had hoped that Chuck would have an answer.
If you have two disks that are nominally the same, and one works and the
other doesn't, then there is obviously something different.
If the sector contents are the same, then the next step would be to
examine the headers, gaps, and addressmarks.
For example,
if the gap after index pulse, before the first sector ofeach track is
too short, then it may not be able to read the sector header of the
first sector.? (A serious and common problem with NEC-style FDC, not
usually a prolem with WD-style FDC)
Do you have a way to examine the raw track encoding?
I used a track read with a WD 179x (slightly modified "Trakcess" on
TRS80 Model 3), and "TE" with the Central Point Option Board.
Find a copy of Anadisk and see what it reports regarding sector
ordering. Also, you might try "sector" mode with the "Diagnostic
Read"
function, though that requires a bit of knowledge.
Another thing to try is to grab a copy of Dave Dunfield's ImageDisk
utility and see if it gets the copy right. You could also image both
disks and compare the images.
TeleDisk has a lot of "special case" code.
People still have diskette drives?
--Chuck
24 Jun
24 Jun
2:22 a.m.
Find a copy of Anadisk and see what it reports
regarding sector
ordering. Also, you might try "sector" mode with the "Diagnostic
Read"
function, though that requires a bit of knowledge.
I did have a look at that, now that the show we were getting ready for is finished I can
spend a bit more time analysing. I used Anadisk to get disk details for ImageDisk because
that?s usually the first utility I go for. It was just pure fluke that I suggested he
re-imaged the disk with Teledisk and was amazed that it worked.
Another thing to try is to grab a copy of Dave
Dunfield's ImageDisk
utility and see if it gets the copy right. You could also image both
disks and compare the images.
I?ll do that too. Hurray for ?special case? code though :)
People still have diskette drives?
Somebody has apparently got an HxC emulator running in a Waveterm but we didn?t have time
to investigate that route. Because the Waveterm is still thinking there?s 8? drives in
there it keeps the motor running constantly and only drops the head to read a sample.
Cheers,
--
adrian/witchy
Owner of Binary Dinosaurs, the UK's biggest home computer collection?
t: @binarydinosaurs f: facebook.com/binarydinosaurs
w: www.binarydinosaurs.co.uk
2373
days inactive
2374
days old
7 comments
4 participants
participants (4)
-
binarydinosaurs＠gmail.com -
cclist＠sydex.com -
cisin＠xenosoft.com -
imp＠bsdimp.com