Re: HP 7970E - interest to split?

[Jon Elson via cctalk]

On 09/30/2017 10:31 PM, Chuck Guzis via cctalk wrote:

On 09/30/2017 06:29 PM, Jon Elson wrote:


Ahh, but the genius is built into the heads!  Note that nobody
recorded 800 BPI on 7-track tapes. I think they supported 200 and 556
BPI, only.  There were some low-density 9-track tapes, but generally
most 9-tk tapes were 800 BPI.

Say what?   We certainly did a lot of tape at CDC at 800 NRZI on the
607s.657 and 667.   Most IBM 729s, but for the very early models could
support 800 NRZI.

800 NRZI for 9 track was common only until about 1970; 1600 PE was
available on the 2400 series.   By 1975, it had pretty much become the 9
track interchange standard.  In spite of packing more data on a piece of
tape, 1600 PE is actually more reliable than 800, as a dropped bit
transition doesn't corrupt data downstream.


OK, I never used 7 track tapes, but I thought that they were 
recorded at lower density than 9 track.

Yes, I know about 1600 PE, 3200 PE and 6250 GCR.

Jon

Re: HP 7970E - interest to split?

[Chuck Guzis via cctalk]

On 09/30/2017 06:29 PM, Jon Elson wrote:

> Ahh, but the genius is built into the heads!  Note that nobody
> recorded 800 BPI on 7-track tapes. I think they supported 200 and 556
> BPI, only.  There were some low-density 9-track tapes, but generally
> most 9-tk tapes were 800 BPI.
Say what?   We certainly did a lot of tape at CDC at 800 NRZI on the
607s.657 and 667.   Most IBM 729s, but for the very early models could
support 800 NRZI.

800 NRZI for 9 track was common only until about 1970; 1600 PE was
available on the 2400 series.   By 1975, it had pretty much become the 9
track interchange standard.  In spite of packing more data on a piece of
tape, 1600 PE is actually more reliable than 800, as a dropped bit
transition doesn't corrupt data downstream.

--Chuck

Re: HP 7970E - interest to split?

[Jon Elson via cctalk]

On 09/30/2017 12:47 PM, Chuck Guzis via cctalk wrote:

What I found a little surprising is that on a 7970 head assembly, the DC
resistance of the 9 track coils is about 76 ohms, but on the 7 track
side, the resistance measures out to about 27 ohms.

Yet, both feed into the same read amplifiers.  On the standard 7970
dual-mode drives, the switching between 7 and 9 track, is performed with
a JFET on the input.  I use small-signal DIP relays.

I would have expected a big difference in the signal levels between the two.


Ahh, but the genius is built into the heads!  Note that 
nobody recorded 800 BPI on 7-track tapes.
I think they supported 200 and 556 BPI, only.  There were 
some low-density 9-track tapes, but
generally most 9-tk tapes were 800 BPI.  So, this means that 
a transition spans less longitudinal distance along the tape.
So, you get less output from the head for an 800-BPI 
transition. For 800 BPI, they might make the head poles a 
bit shorter, but they HAD to make the gap narrower!  The gap 
must be no more than half the length of a bit, or the
signal output rapidly drops.  I'm guessing the smaller head 
poles and narrower gap forced them to use more turns of much 
smaller wire to get the signal back up to what the read amps 
need.


Jon

Re: formatting MFM drives on a IBM PC

[Chuck Guzis via cctalk]

On 09/30/2017 04:12 PM, Tom Gardner via cctalk wrote:
> I think Chuck has it backwards, AT Attachment as defined by the ANSI
> committee publically predates IDE.  Although IDE was used internally at WD
> it did not surface publically until well after the ANSI committee adopted AT
> Attachment, abbreviated ATA.  The AT in AT Attachment or ATA has never stood
> for "Advanced Technology" although many presume so.

"IDE" was a Western Digital term for drives used in the Compaq PC,
dating from 1986.   I've probably got documents from about that time
talking about IDE, if I look.   Because of Compaq's introduction of the
thing early on, that's what we called it then.

The "ATA Standard" began its work in 1988 by the Common Access Method
committee of ANSI X3T10 and eventually came out with a standard in 1994,
but that was long after "IDE" was the lingua franca term for these
drives.  The ANSI document:

https://ecse.rpi.edu/courses/S15/ECSE-4780/Labs/IDE/IDE_SPEC.PDF

In, you'll read:

"The application environment for the AT Attachment Interface is any
computer which uses an AT Bus or 40-pin ATA interface. The PC AT Bus is
a widely used and implemented interface for which a variety of
peripherals have been manufactured.  As a means of reducing size and
cost, a class of products has emerged which embed the controller
functionality in the drive.  These new products utilize the AT Bus fixed
disk interface protocol, and a subset of the AT bus.  Because of their
compatibility with existing AT hardware and software this interface
quickly became a de facto industry standard."

So, even ANSI X3 talks about the PC AT bus.  And yes, "AT", according to
IBM, stands for "Advanced Technology"

Pretty much, all you need to connect an ATA-1 drive to the 5170 bus is a
couple of transceivers and an address decoder.

--Chuck

RE: formatting MFM drives on a IBM PC

[Fred Cisin via cctalk]

On Sat, 30 Sep 2017, Tom Gardner via cctalk wrote:

I think Chuck has it backwards, AT Attachment as defined by the ANSI
committee publically predates IDE.  Although IDE was used internally at WD
it did not surface publically until well after the ANSI committee adopted AT
Attachment, abbreviated ATA.


Do you, by any chance have the dates?

A casual look seems to show the WD IDE ("Integrated Drive Electronics") 
being used by Compaq in 1986?, although not necessarily NAMED "IDE", yet.


A casual look seems to show ANSI ATA as being early 1990s.



The AT in AT Attachment or ATA has never stood
for "Advanced Technology" although many presume so.


DOES it stand for something else?
It is generally assumed that it refers to the IBM PC/AT (aka 5170), where 
AT DID, indeed, stand for "Advanced Technology"

RE: formatting MFM drives on a IBM PC

[Tom Gardner via cctalk]

I think Chuck has it backwards, AT Attachment as defined by the ANSI
committee publically predates IDE.  Although IDE was used internally at WD
it did not surface publically until well after the ANSI committee adopted AT
Attachment, abbreviated ATA.  The AT in AT Attachment or ATA has never stood
for "Advanced Technology" although many presume so.

A (P)ATA drive never directly connected to an AT bus, nor for that matter
any other bus but always required some form of adapter, albeit very simple
in the case of the 5170 type bus.  

The AT Attachment compatibility is with the Task File register set which can
be sent over any interface parallel, serial, carrier pigeon, whatever.  A
SATA drive could be connected to a 5170 with an 5170 bus to SATA bridge or
more like a PATA to SATA bridge (with a  PATA HBA in the 5170) - PATA to
SATA bridges did exist and u might find one on eBay.  Booting would be a
problem and so would capacity but it should talk.

So Serial ATA makes sense to me

Tom

-Original Message-
From: Chuck Guzis [mailto:ccl...@sydex.com] 
Sent: Thursday, September 28, 2017 5:37 PM
To: cctalk@classiccmp.org
Subject: Re: formatting MFM drives on a IBM PC

On 09/28/2017 05:12 PM, Jules Richardson via cctalk wrote:
> On 09/27/2017 09:59 AM, Ethan via cctalk wrote:
>> The idea of IDE, as my understanding, is the controller that existed 
>> as an ISA card was moved onto the actual drive, and then what became 
>> the controller was mostly just extending the ISA bus over to the 
>> drive.
> 
> I actually have an IDE "controller" somewhere which is just a tiny PCB 
> with an ISA connector on one side and a 40 pin IDE connector on the 
> other, along with a couple of ICs (presumably buffers/latches, but I 
> don't know without finding it).  It's somewhat unusual, given that IDE 
> ports were normally included as part of multi-I/O boards, or (a little
> later) often incorporated into the motherboard.

IDE used to be called "ATA" - "AT Attachment"; i.e. something tailored to
the PC AT (5170) 16-bit ISA bus.

What I find perplexing is the acronym "SATA" for "Serial ATA".  The name
would imply that a drive can be connected to a 5170, but I'm not aware of
any SATA adapters for the 5170 PC/AT.

--Chuck

Norton Utilities on 8"?

[Richard Cini via cctalk]

All –

 

    I’m in the process or restoring a Seattle Gazelle for the 
Vintage Computer Federation and one of the disks that it came with had “Norton 
Utilities” written on it. I’ve imaged the disk but I’m not entirely sure what 
system it was for, and the directory seems to be scrambled, and the disk 
geometry doesn’t seem to fit with the system (16 sectors, 512b DSDD) versus the 
normal (8 x 1024b DSDD).

 

    A quick search shows that v1.0, and maybe 2.0, were around at 
the time…no word on what distribution media, though.

 

    Is anyone aware of there being a version available on 8” disks?

 

Thanks!

 

Rich

 

--

Rich Cini

http://www.classiccmp.org/cini

http://www.classiccmp.org/altair32

 

Re: HP 7970E - interest to split?

[Paul Koning via cctalk]

> On Sep 30, 2017, at 11:50 AM, Stefan Skoglund via cctalk 
>  wrote:
> 
> fre 2017-09-29 klockan 12:52 -0400 skrev Paul Koning via cctalk: 
>> ...
>> The approach would be somewhat similar to the disk support in SIMH, which 
>> does already have "RAW" as one of the formats, meaning direct access to a 
>> real disk.
> 
> RAW disk does that also includes using a zfs/fc/iscsi volume as disk ?
> It should considering they appear as sd{a-z} on the host.

I would assume yes, for the reason you mentioned.  But I haven't tried it.

paul

Baltimore Area Collector Liquidating

[Fran Smith via cctalk]

Hiya,

Ok its time to get rid of more stuff from my "Stash"

A mostly complete Decmate with the RX02 pedestal
A complete Next slab system with display and laser printer.
A complete IBM Aptiva with original boxes
2 Vaxstation 4000 Model 60's
A huge pile of Dec Professional 380's
A huge pile of Vaxstation 3100's
Two Televideo Dumb terminals
And piles and piles more stuff that can be picked through.

I did pay real money for all of this and have stored these treasures for 
many years in my finished basement. To be honest I paid too much but 
that is how collecting goes :)


I am only liquidating because honestly this stuff needs to be used , 
displayed and monkey'd about with rather than languish here.


My health also took a downturn and I can't deal with heavy items anymore.

The prices are cheap (Well Well Below That "Site") so contact me to get 
some idea.


If you buy more I will throw in all sorts of free or near free stuff.

All items must be picked up at my Baltimore , Maryland area home. I will 
not ship because I really don't want to moosing around heavy stuff.


Thanks!!

Fran

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

Re: HP 7970E - interest to split?

[Chuck Guzis via cctalk]

What I found a little surprising is that on a 7970 head assembly, the DC
resistance of the 9 track coils is about 76 ohms, but on the 7 track
side, the resistance measures out to about 27 ohms.

Yet, both feed into the same read amplifiers.  On the standard 7970
dual-mode drives, the switching between 7 and 9 track, is performed with
a JFET on the input.  I use small-signal DIP relays.

I would have expected a big difference in the signal levels between the two.

--Chuck

Re: HP 7970E - interest to split?

[Jon Elson via cctalk]

On 09/30/2017 06:58 AM, Tony Duell via cctalk wrote:


One of my tape drives has a set of equal-value resistors connected to
each of the read channels, the others ends of the resistors being brought
out to a testpoint. You connect a 'scope there and read a skewmaster
tape. The idea is that if the read transitions all occur at the same time
(no skew) you get a pefect step on the 'scope, if some are later than
others you get a staircase. You adjust the heads for the smallest
width of the stairs.


Yes, this is the classic NRZI skew adjust testpoint.  The 
resistors sum up the read data latches, that are set by 
seeing a transition of either polarity on a data channel, 
and reset by a one-shot that is set to about half a 
character time.
You run a skew tape or a tape written with all ones.  If you 
see a square wave, all is good.  If you see a slope on the 
wave, adjust the skew adjust screws on the head mount to 
make it as straight as possible.


Jon

Re: HP 7970E - interest to split?

[Stefan Skoglund via cctalk]

fre 2017-09-29 klockan 12:52 -0400 skrev Paul Koning via cctalk: 
> > On Sep 29, 2017, at 12:33 PM, Jon Elson via cctalk  
> > wrote:
> > 
> > On 09/29/2017 10:56 AM, Henk Gooijen via cctalk wrote:
> >> 
> >> 
> >> 
> >> On a related note my plan is to make a USB-based, Pertec-compatible
> >> controller for it. Not sure how SimH connects with peripherals so I'm
> >> /very/ eager to talk with someone familiar with its workings. I'll also
> >> release all the board files and firmware as open-source. Timeline as always
> >> is completely unknown, though I do have a now-vested interest in making it
> >> work.
> 
> It looks like it only handles tape container files (.TAP files) right now.  
> But the machinery basically has a tape emulation module (sim_tape.c) which 
> has a number of functions roughly corresponding to what a tape drive does 
> (read forward, write tape mark, rewind, etc.).  It doesn't look like a 
> terribly hard task to support a real tape drive as another "container file 
> format".  One question I can think of is how OS-dependent raw SCSI access is.
> 
> The approach would be somewhat similar to the disk support in SIMH, which 
> does already have "RAW" as one of the formats, meaning direct access to a 
> real disk.

RAW disk does that also includes using a zfs/fc/iscsi volume as disk ?
It should considering they appear as sd{a-z} on the host.

Re: formatting MFM drives on a IBM PC

[allison via cctalk]

On 09/29/2017 06:42 PM, Peter Corlett via cctalk wrote:
> On Fri, Sep 29, 2017 at 01:08:24PM -0700, Fred Cisin via cctalk wrote:
>>> On 09/29/2017 11:20 AM, Jon Elson via cctalk wrote:
> [...]
>>> Older BIOS firmware provided no means for the user to define the geometry of
>>> a connected drive - just a list of predefined types, and those often maxed
>>> out at far less than any 512MB limit. There were various software solutions
>>> to get around it, though.
>>> Of course operating systems had various limits on the maximum size of a
>>> partition on top of that - e.g. I think it was 32MB in earlier versions of
>>> MS-DOS.
>> What are the current drive size limits?
> The latest relevant standards seem to be ATA-6 for ATA, and SBC-3 for SCSI,
> which have 48- and 64-bit logical block addresses (LBAs) respectively.
>
> ATA LBAs are *always* 512 bytes no matter what the physical sector size is, so
> this sets a hard limit of approximately 144EB. This limit also implies that
> sectors must be a multiple of 512 bytes (e.g. modern 4kiB-sectored disks) and
> unusual sector sizes are not supported.
>
> SBC-3 does not specify the size of a block, but it notes that most disks
> support 512 bytes with some also supporting 520 or 4096. Assuming 512-byte
> logical blocks, the limit is 9.44ZB.
>
> Since the hard disk equivalent of Moore's Law seems to be running out of steam
> at mere tens of terabytes, we probably won't need to raise either limit for a
> while yet :)

What not mentioned is that LBA addressed drives have cache.  With that
sector size is meaningless
as the track (most likely) or several are read into the cache.  There is
LRU applied to keep the media
up to date and also insure the cache is not stale.  With cache it makes
doing 512byte blocks a trivial
issue.

Now we have SSDs...  Whole new game same old protocols.

Allison

Re: HP 7970E - interest to split?

[Anders Nelson via cctalk]

Hi Chuck,

How did you program compile if you were missing a semicolon? I'm assuming
you're programming in C and not scripting it somehow... Good news on that
front!

Was it typical to read a record(s) in reverse if an error was encountered?
Sounds like that might require buffering the whole record. Easy for us now
certainly.

=]

--
Anders Nelson

+1 (517) 775-6129

www.erogear.com

On Sat, Sep 30, 2017 at 1:37 AM, Chuck Guzis via cctalk <
cctalk@classiccmp.org> wrote:

> Hi Anders,
>
> I found the reason for my SD card problems--a missing semicolon.   There
> was supposed to be a stall until SDIO transfer complete; instead of a
> stall, the next statement was being executed.
>
> Argh.  It's those little things that really trip you up.
>
> At any rate, I've even got read-reverse working, so I can read a block,
> parity check it and retry automatically if there's an error.
>
> Not too shabby for a $10 MCU board.
>
> --Chuck
>
>

Tandy Assembly

[Peter Cetinski via cctalk]

Only one week to go until the inaugural Tandy Assembly event in Chillicothe, 
OH.  We’ve been planning this for over a year and it’s exciting to finally see 
it coming together.  We’ll have exhibits and presentations covering all of the 
Tandy computers, from the TRS-80 Model I to the Tandy 6000.  The pockets, 
portables and PCs too.  If you can’t join us in Ohio, we plan on live streaming 
many of the presentations and events on YouTube.  

http://www.tandyassembly.com 

Thanks,

Pete

Re: HP 7970E - interest to split?

[Tony Duell via cctalk]

On Sat, Sep 30, 2017 at 6:32 AM, Chuck Guzis via cctalk
 wrote:

> Still a few too many read errors for my taste, so this weekend is going
> to be a session with a 'scope and alignment tape.   I think I've got a
> way to adjust skew--just hook a logic analyzer probe to each channel and
> set the LA to trigger on state changes.   The HP box I'm using is good
> to about 25 nsec.

One of my tape drives has a set of equal-value resistors connected to
each of the read channels, the others ends of the resistors being brought
out to a testpoint. You connect a 'scope there and read a skewmaster
tape. The idea is that if the read transitions all occur at the same time
(no skew) you get a pefect step on the 'scope, if some are later than
others you get a staircase. You adjust the heads for the smallest
width of the stairs.

-tony

Re: formatting MFM drives on a IBM PC

[Peter Corlett via cctalk]

On Fri, Sep 29, 2017 at 01:08:24PM -0700, Fred Cisin via cctalk wrote:
>> On 09/29/2017 11:20 AM, Jon Elson via cctalk wrote:
[...]
>> Older BIOS firmware provided no means for the user to define the geometry of
>> a connected drive - just a list of predefined types, and those often maxed
>> out at far less than any 512MB limit. There were various software solutions
>> to get around it, though.

>> Of course operating systems had various limits on the maximum size of a
>> partition on top of that - e.g. I think it was 32MB in earlier versions of
>> MS-DOS.
> What are the current drive size limits?

The latest relevant standards seem to be ATA-6 for ATA, and SBC-3 for SCSI,
which have 48- and 64-bit logical block addresses (LBAs) respectively.

ATA LBAs are *always* 512 bytes no matter what the physical sector size is, so
this sets a hard limit of approximately 144EB. This limit also implies that
sectors must be a multiple of 512 bytes (e.g. modern 4kiB-sectored disks) and
unusual sector sizes are not supported.

SBC-3 does not specify the size of a block, but it notes that most disks
support 512 bytes with some also supporting 520 or 4096. Assuming 512-byte
logical blocks, the limit is 9.44ZB.

Since the hard disk equivalent of Moore's Law seems to be running out of steam
at mere tens of terabytes, we probably won't need to raise either limit for a
while yet :)