>
>Subject: Re: US Sources for old ICs
> From: Paul Koning <pkoning at equallogic.com>
> Date: Wed, 26 Oct 2005 11:31:08 -0400
> To: cctalk at classiccmp.org
>
>>>>>> "Chuck" == Chuck Guzis <cclist at sydex.com> writes:
>
> Chuck> The chat about the FDC reminded me of a data acquisition
> Chuck> project I've had partly completed for a couple of years.
>
> Chuck> Where is a good source for purchasing small-quantity ICs in
> Chuck> the US? In particular, I'm looking for a few 64Kx9 async
> Chuck> FIFOs (e.g. IDT7208 or CY7C466) in DIP packages (I can find
> Chuck> the Cypress part at DigiKey, but only in QFP).
>
>Jameco tends to have the older stuff in the older packages. They
>still sell 4116s, for example, and 7400 series (plain 7400, not the
>74ABCDEF00 newfangled variants).
>
> paul
JDRmicrdevices get the catalog. Oddball stuff like the CY466 won't
be there but TTL and CMOS generic logic is. For the oddballs maybe
>from Digikey or Mouser.
Allison
I recently stumbled across an odd set of emulations available in the
controller for a printer I have. This system was introduced in the early
80's, and was only recently EOL'd by xerox (the 4090LPS).
The default terminal emulation is ADM3, which works well with the Link MC5
terminals. ADM11 is also available, along with VI-90 and VG920, which I've
never heard of. To make life a little bit easier while playing with this
beast, I would like to tie it to a pc (linux or windows). Changing the
system directly is not really an option, as the whole thing is a very
bizarre setup, based loosely on DEC architecture.
The question is, does anyone know anything about the VI-90 or VG920 terminal
emulations? And, does anyone know of a way to directly emulate an ADM3 or
ADM11 terminal under linux or windows? (via serial, of course)
The system has many options available, including: a scsi tape drive option
(I have the equipment, but it's another xerox oddity to run), a shared-disk
interface option, ethernet (early implementation for thicknet), a 9-track
tape drive (I have, and use, this one), an "online" (bus & tag
channel-attach) interface, which I also have. It's an odd, and interesting,
system.
>From what I've been told, the entire thing was written in RAD50. It can
carry, total, about 512M of mem, spread across the various subsystems (Main,
Font, Graphics), all on individual chips, nothing socketed, lots of
blinkenlights.
Anyone wanting more info (or even an entire controller!) just ask!
--Shaun
--
"If you live to be a hundred, I want to live to be a hundred minus one day,
so I never have to live without you." -- Winnie The Pooh
http://www.lungs4amber.org
Eric J Korpela wrote:
> Xilinx used to have software that would run under Solaris...
woodelf <bfranchuk at jetnet.ab.ca> wrote:
: They could indeed. All I know is the free?/low cost software is
: windows based
: and the hi-priced software is windows/other. I suspect they think if
: you are not
: using a Pee Sea you have $$$ to shell out for software.
Xilinx' free download is available for Winblows and for Linux. I have
downloaded the Linux version, but ran into a brick wall installing it.
The downloadable package is actually a .sh file that constitutes a
/bin/sh script with binary data following it (I know, a totally crazy
arrangement, but I didn't look more closely into how it actually works
due to lack of interest in this manure). When you run it, it unpacks
into a bunch of other files. Those consist of an installer executable
and a bunch of files for it to unpack. When I saw that the latter were
ZIPs, I thought I would just extract them manually, but not so fast -
they are encrypted!!!
The fuckers are forcing the use of their assinine installer, which is of
course not runnable. Attempting to run it produces a complaint about
missing libXm.so.3, i.e., it's some fucking GUI thing. I didn't even
bother finding, downloading and compiling whatever package would provide
that libXm.so.3, since the next thing it would complain about of course
would be that I have no X display. And if I turned inside out to give
it the goddamn X display it wants (that would require making a trip to
my old facility, which would in turn require borrowing a friend for the
duration of the entire procedure since I don't feel safe going there
alone [it's a terrible neighbourhood], and none of my friends have that
kind of spare time), I'm sure there would be something else that would
make it unhappy and it still won't run far enough to unpack the fucking
encrypted archive.
In short, Xilinx has done more than enough to deserve the death penalty
under Galactic law. They are badly in need of a date with Madame La
Guillotine.
Fortunately, I may have found an alternative solution using Altera. Not
that Altera is any better, but the company for which I'm currently doing
a consulting project involving FPGAs uses Altera, and I've got Quartus
(Altera's SW) for Linux through them. I haven't seen it yet actually,
I've just got an E-mail from the boss saying that the CD is on my desk
for me to play with tomorrow. So I'll look at it tomorrow. I'm hoping
that I can install it on my Linux box at the company and then copy the
installed files to my own one (i.e., pirate it).
: Back to FPGA software for the VAX, I don't see the hardware a problem
: but the software could be still be one. The small PDP'S has hoards of
: paper tape
: software, the big ones all need a OS.
The OS is not a problem - I am the maintainer of the one true OS for the
VAX (which is of course 4.3BSD-Quasijarus).
MS
>
>Subject: Re: Looking for an 8 bit FDC...
> From: Dave Dunfield <dave04a at dunfield.com>
> Date: Tue, 25 Oct 2005 23:28:30 -0400
> To: "General Discussion: On-Topic and Off-Topic Posts" <cctalk at classiccmp.org>
>
>Have you tried just using the diskette controller in a PC? - I've
>found that almost all of my P1ish machines can read/write FM, and
>I've used ImageDisk to backup and restore almost all of the soft-
>sector formats that I have here (although some are better than others.
>I've found that Intel mainboards work very well, although the ones
>I have only support one drive select).
Again an older 486 and an AT (older) FDC is a hard solution to beat
for the majority.
Amazing how an relatively easy 80% solution covers 97% of the need. A
PC controller isn't a complete solution but for a huge amount of day
to day effort it's a usable solution. Those disks that cant be read
with that solution are already known. Those bad actors are either
hard sectord where each is likely a special case or sufficiently odd
enough to warrent keeping a sample system fully operational.
The alternate is some system that has a bus so you can insert
controllers to fit the oddball situations and just plug them in.
I use S100 for that, right now the only formats that crate can't handle
is the GCR formats (no systems here use it) and hard sector other than NS*
though I do have an Altair hard sector system.
The only problem that isn't solved however is they guy on the other
side of the planet with the only other oddball system like it and
needs a boot disk.
I've have had this problem since I put a NS* conttroller in my first
S100 system in '77. That problem is what I want is on brand Q media
and format and I have brand N that doesn't do that.
Allison
>
>Subject: Re: Looking for an 8 bit FDC...
> From: "Chuck Guzis" <cclist at sydex.com>
> Date: Tue, 25 Oct 2005 11:22:55 -0700
> To: cctalk at classiccmp.org
>
>On 10/25/2005 at 1:59 PM Allison wrote:
>
>>The 765A was least integrated but with the right external support more
>>flexible than the 37C65 and later.
>
>You still have to put up with the "blindness' during the IAM interval and
>the lack of support for MFM 128-byte sectors--and the lack of support for
>non-standard address marks (not at all uncommon in old formats). At least
>with the WD chips, you can read the raw track data.
Ah huh, what? Who does 128byte MF sectors??? If you really have to it's
possible to tell a 765 that it's a short sector in mfm. As to non standard
address marks that was a tandyism and 765 "read deleted" works.
>>>The data seperator was a fairly simple circuit and could be done with
>>a counter or for improved results with a latch and prom.
>
>Or just find a nice little 8-pin WD 9216 data separator. If you wanted
>more than that, a PLL might be the ticket.
If you can find one 9216 is good or the 9229. IF not the latch/prom is
a synthetic PLL (which is the logic in the 9216 etal) and very good with
less headaches than a real PLL.
In general a 765A on a late XT or early AT class FDC is the trick
more possibilities than not. Those that it can't be require extraordinary
effort.
There are few reasons that seem to abound for do alls. Those be "archives"
and I want to run a zzxxyy sim on a PC. Those that most often archive
have a host that can read it. The other case is obvious. However, the case
I run into the most is I wish to run ABC on Northstar* under CP/M and need
to get it on to the AmproLB. For that the Ampro connected to the serial port
of the NS* and MDM740 neatly sidesteps the "format problem" and once on the
AmproLB I can run DOS and write it to a 720K 3.5" floppy under
(DOS/WIN/NT/Linux) and run it under MyZ80 or whatever.
Allison
Hi Jules,
I replied yesterday evening, but somehow my e-mail did not show up (at
least here).
So, I resend it. Sorry if it hits the list a second time.
I have a design that worked - 20 years ago, with a 1793. It has a
PLL-based data
separator, using a 72LS629 and 2 4-bit counters, IIRC. I need to look it
up.
It was interfaced to 6809 bus signals, so changing it for 6502 is
straight forward.
I recently bought a few 1793 chips from BGMICRO, and then decided to
make my design
a lot simpler by using a 2793 as that FDC has a data separator
integrated.
With the correct clock crystal selection you can use the hardware for
3", 3.5", 5.25"
_and_ 8" drives (under software control)!
I am working on a 2793 which is piggy-back connected to the
"Blinkenlight" Core Board
which is based on the 6809E and has more than enough RAM and EPROM on
board.
But that's just a matter of how you write the software how to read the
floppy.
For instance, sector-wise or track-wise as the 1793 and 2793 have
read-track commands.
I would not use DRAM chips, but takje a single static 32kx8. Saves work
on soldering
and room on the board.
I started a webpage on this new project, see www.pdp-11.nl/ in the "my
projects" folder.
Vince still has Core Board kits (all parts included for $70), so the job
is limited to
the FDC part (and the software ...) But the Core Board is a 6809E
instead of the 6502,
I regard that's a plus !!!
Handling double density will become tricky with the 6502, and even with
the 6809,
because of timing requirements (polling for DRQ is asserted). Single
density can be
handled by polling, but double density requires the use of IRQ or FIRQ
or the use the
SYNC instruction (need connect DRQ* to IRQ or FIRQ).
I am waiting on the arrival of my 2nd BGMICRO order that has the 2793
...
gd luck,
- Henk, PA8PDP
This message and attachment(s) are intended solely for the use of the addressee and may contain information that is privileged, confidential or otherwise exempt from disclosure under applicable law.
If you are not the intended recipient or agent thereof responsible for delivering this message to the intended recipient, you are hereby notified that any dissemination, distribution, or copying of this communication is strictly prohibited.
If you have received this communication in error, please notify the sender immediately by telephone and with a "reply" message.
Thank you for your cooperation.
9000 VAX <vax9000 at gmail.com> wrote:
> Altera supplies free "web" edition of its quartus II software too. I
> am using that and I think it is as good as the linux one.
But it's for Winblows only, isn't it? Dunno about you, but I'm not
bringing a Winblows machine into the house, that's absolutely out of the
question! Even Linux is right on the limit of what I can tolerate.
MS
woodelf <bfranchuk at jetnet.ab.ca> wrote:
> From what I remember of Altera software , it brands itself to the
> HD id or the network ID.
I don't see how they can successfully do this under Linux, or any kind
of UNIX. I guess they could try, but this "protection" would be
laughable - a baby can get around this kind of "copyprotection". It
ain't DOS or Winblows, it's Linux! I can intercept all system calls it
makes, I can hack the kernel, I can do whatever I want to make it
believe it's running on the machine it's "licensed" to.
> Also I think you also may be stuck
> with the GUI they use too.
No, all the work is done by command line tools which the GUI calls, and
the command line tools are documented. I'll use them as a backend for
Icarus Verilog.
MS
>OK, I'm wanting to build a board with an 8 bit CPU (probably Z80,
>possibly 6502) and a floppy controller IC on board with the intention of
>hanging it off my PC (via serial or parallel, undecided yet) and
>allowing me to read and write *most* formats from various 1980's 8 bit
>micros...
Funny, when I asked a few months ago if anyone was interested in
collaborating on doing exactly this, I was practically run off the list
with comments about how impractical it was, nobody would build it,
and wouldn't a catweasel be ever so much better... So I let the idea
drop (still might build one privately however).
>Intel's 8271 looks like a possibility at the moment, but I thought I'd
>poll the list for alternative ideas too. FM support is of course
>critical - MFM is less of an issue as the host PC can handle that.
>
>I've never built any kind of computer from scratch, so it'll be a useful
>experience. I figure on putting just enough code in ROM to support
>downloading of actual firmware to the device over whatever the link is
>to the PC, as that should save a lot of headache!
>
>Hopefully RAM requirements will be low enough that I can go the SRAM
>route and avoid messing around with DRAM refresh (although IIRC the Z80
>has much of the necessary stuff built in...)
I think the 8271 is pretty limiting ... I'd vote for a 1793 (or 2793
which has a bulld in data-separator) - it is used in a lot of classic
systems, can handle a wide variety of formats, and can do raw track
reads.
A single 32k SRAM should be plenty to buffer a few tracks, which is
really all you need.
Have you tried just using the diskette controller in a PC? - I've
found that almost all of my P1ish machines can read/write FM, and
I've used ImageDisk to backup and restore almost all of the soft-
sector formats that I have here (although some are better than others.
I've found that Intel mainboards work very well, although the ones
I have only support one drive select).
Regards,
Dave
--
dave04a (at) Dave Dunfield
dunfield (dot) Firmware development services & tools: www.dunfield.com
com Collector of vintage computing equipment:
http://www.parse.com/~ddunfield/museum/index.html
>
>Subject: RE: What is DECdatasystem?
> From: "Julian Wolfe" <fireflyst at earthlink.net>
> Date: Tue, 25 Oct 2005 22:02:48 -0500
> To: "'General Discussion: On-Topic and Off-Topic Posts'" <cctalk at classiccmp.org>
>
>So were these prebuilt systems versus built-to-configure systems?
Mostly, more like limited menu systems. Aka if you need to do this
than buy that.
>In that case, that would make sense, just like the edusystem -8s.
Very similar. Both were packaged with some options.
Allison
>
>-----Original Message-----
>From: cctalk-bounces at classiccmp.org [mailto:cctalk-bounces at classiccmp.org]
>On Behalf Of Allison
>Sent: Tuesday, October 25, 2005 9:28 PM
>To: cctalk at classiccmp.org
>Subject: Re: What is DECdatasystem?
>
>>
>>Subject: What is DECdatasystem?
>> From: "Wolfe, Julian " <ISC277 at CLCILLINOIS.EDU>
>> Date: Tue, 25 Oct 2005 21:16:07 -0500
>> To: "'General Discussion: On-Topic and Off-Topic Posts'"
><cctalk at classiccmp.org>
>>
>>Okay, looking over the web, I've seen several PDP11 machines marked
>>"DECdatasystem"... They don't seem to have anything in common...it's like a
>>random marking DEC decided to put on machines. Anyone know what it means?
>
>The common thread of 'datasystem is they are bundled systems sold as
>application solutions rather than "computers".
>
>It's part of DEC's history that Digital didn't sell computers, they sold
>programmed data processors. that and quirky marketing.
>
>Allison
>Ex Digit.
>
>
