>
>Subject: Re: PDP11/23+ goes on
> From: Al Kossow <aek at bitsavers.org>
> Date: Sun, 19 Jun 2005 18:51:57 -0700
> To: classiccmp at classiccmp.org
>
>The RQDxx controllers are a bit clunky to use with their
>breakout panel if you aren't putting it into a box designed
>for it.
Use the brakout designed for the BA123 to breakout a RQDX.
Just put the breakout card in some unused but powered slot
(last ore beyone the last) and it will be powered and
expand to multiple drives. I have my 11/73 with the 18/AB
backplane that way.
>I still say that the simplest thing to do is just to find
>a used Qbus MSCP SCSI conroller though.. If you can find
>someone scrapping Qbus microvaxen, they often have them
>hiding inside (that's where I got the ones that I have).
It is.
The problem with QBUS PDP-11 IDE is that I've not seen a
design that has a DMA implementation which make it slow.
Also drivers for common PDP-11 OSs other than unix are
roll your own, wich is possible IF you already have a
working system (with editors and asm).
What makes the Qbus interface for IDE nasty is the
PDP-11 read before write. That means you have to make
the read addresses seperate from the write addresses.
However, in the last few years there is the possibility
of a flash ram disk as devices in the multimegabyte range
are available. Qbus non rotating media, hummm. There
one for the builders to consider.
Allison
>
>Subject: Are you a specialist or a generalist?
> From: "river" <river at zip.com.au>
> Date: Mon, 20 Jun 2005 22:53:12 +1000
> To: <cctalk at classiccmp.org>
>
>Hi,
>
>Just wondering if you regard your classic computing skills as specialised or
>
>Do you have lots of various old systems with different processors and
>
>I regard myself as a generalist, mainly due to the various different
>
>rgds
>river
Line breaks?
I am a generalist, with focus. I have just about one of each micro
or more and a large assortment of systems. I focus on CP/M based,
Omnibus pdp-8e/f/m/a, Qbus PDP-11s and deskside/desktop VAXen (UvaxII,
uVAX2000 and 3100s). The focus is based on hardware size and OSs
represented. However, I do go off and dabble in 1802, TI9900, 8048/9
and other things that aren't comnputers.
Allison
>
>Subject: Re: VTSERVER booting
> From: "Fred N. van Kempen" <waltje at pdp11.nl>
> Date: Mon, 20 Jun 2005 10:34:03 +0200 (MEST)
> To: "General Discussion: On-Topic and Off-Topic Posts" <cctalk at classiccmp.org>
>
>On Mon, 20 Jun 2005, Allison wrote:
>
>> >I'm not 100% sure if it's your problem, but 2.11BSD doesn't work on the
>> >F11 chipset (11/23) because it doesn't have split I&D space. You'll
>> >need to use 2.9BSD or earlier on it.
>> >
>>
>> There are two versions of 2.11 if memory serves one with split
>> I&D and one without.
>No. 2.11 **only** runs on SPlitID systems. For non-ID systems,
>use 2.8, 2.9 or (perhaps) 2.10. Ultrix-11 runs on all of these,
>by the way. AND is fully supported by VTserver, as I maintain
>both :P
Ok. I only run V2.7 as unix without networking is limited
usefulness to me. I'd considered 2.11 on the 11/73 but never
got to it.
>> I know the Falcon, LSI-11/ and 11/2 did not even have MMU.
>Correct, all T11 based systems, plus the /20, /10, /03 and I
>believe the /04 dont have MMU, and are 64K systems.
That was a statement on my part as I've designed with the T-11 and
have the full internal DOC set for that. FYI: the T-11 implmentation
in the VT24x added a MMU like the 11/23 sans PDR/ACF.
Allison
>
>Subject: Re: DSD-440 and/or drive recommendation for PDP-8
> From: "vrs" <vrs at msn.com>
> Date: Sun, 19 Jun 2005 21:45:19 -0700
> To: "General Discussion: On-Topic and Off-Topic Posts" <cctalk at classiccmp.org>
>
>From: "Allison" <ajp166 at bellatlantic.net>
>> From: "vrs" <vrs at msn.com>
>> >(It's easily doable, but not trivial.)
>>
>> Heres the trick. Don't copy the parts, copy the logic and use
>> standard TTL. Most all the functions are easily available or
>> or sometimes available in denser (newer relatively) parts.
>> Trying to copy verbatim will be hard unless you have a lot of
>> DEC junker cards to strip. However most of the logical functions
>> are simple so substitutions are not as scary.
>
>Heh, I did that, but I shot myself in the foot by using the denser parts and
>not quite getting it right.
>
>Mine is actually a Posibus card (Ethan has the other prototype). There are
>other problems with Omnibus cards, too. You can get the bus interface close
>to spec., but not dead on, AFAIK with modern parts. Posibus is more
>forgiving. (Though you can get close enough for small Omnibus backplanes
>pretty easily.)
>
> Vince
I'll limit myself to Omnibus, experience there. Posibus and older
machines I've only done programming on.
It's not as dire as you'd expect. There are a few characteristicts that
must be met. Input current, lower is ok and newer devices are good
there. CMOS as inputs must be used with care as they have to switch
with a postive swing of under 3V(omnibus, any TTL/DTL interface.).
The next is output drive, most newer devices exceed those. Last is
output switching characteristics such as Vol at load current, again
thats easily met wit newer devices. What may be a bit worrisome is
that new devies sink harder on the bus and a small (10-50ohms)
series resistance will damp any ringing that results.
However, once your off the bus (core card logic) you can do as you
please as the system doesn't see it. Use care on decodes to avoid
spikes, timing races and other logic faults that can occur when
using mixed speed logic. Generally good care will avoid this.
The last area. Power and gound busses. Since there is a lower
likelyhood of having the DEC protocards any board constructed
should use good power distribution rules (better than PdP-8 days
as newer is faster!).
The rest of the details are in the DEC logic handbooks and PDP-8
handbooks. They are your friend, or try to be. ;)
Allison
Mike:
Thanks very much for the data sheet.
Rich
-----Original Message-----
From: cctalk-bounces at classiccmp.org [mailto:cctalk-bounces at classiccmp.org]
On Behalf Of Michael Holley
Sent: Sunday, June 19, 2005 4:24 PM
To: cctalk at classiccmp.org
Subject: Re: Looking for data sheet for AY5-2376
> Does anyone have a PDF copy of the data sheet for the General
> Instruments AY5-2376 ASCII keyboard encoder chip?
>
> Rich Cini
I have a 4 page version of the SMC KR-2376 on my SWTPC CT-1024 web page. It
was used in the SWTPC KBD-5 keyboard.
http://www.swtpc.com/mholley/CT_1024/KR2376.pdf
Michael Holley
www.swtpc.com/mholley
I managed to take some time off during my visit to Minneapolis to get to
Chippewa Falls Museum of Industry and Technology. It was an easy 2-hour
drive, and the museum opens at 1pm on weekdays. This trip had the air
of a pilgrimage for me, since I still have a Cray-1S architecture manual
I ordered from them c. 1981 and I've always wanted to see where they
came from.
Chippewa Falls is a tiny place, and I had no trouble finding the museum
(though the street it's on is currently closed off for work). I only
had an hour or so there, as I had to get back to my flight later that
afternoon.
The museum director, Yvette Viets Flaten, is very enthusiastic about her
Cray exhibits. It also turns out that her mother comes from
Northampton, not far from where I live (and Chippewa Falls also had a
show manufacturing industry, just like Northampton).
Complete CPUs they have there (AFAIK all except the X/MP are Cray
designs):
CDC 160A
CDC 1604
CDC 6500
CDC 7600
Cray 1
Cray X/MP
Cray 2
plus various other exhibits, including a ERA magnetic drum unit and
Seymour Cray's desk. You can walk around them all, and in many cases
there are samples of the internal boards removed for closer viewing.
There are some videos on show, but I didn't have time for those. I took
a few photos, but they haven't come out too well, so I suggest anyone
interested look at:
http://www.spikynorman.dsl.pipex.com/CrayWWWStuff/Cfaqmustop.htmlhttp://csumc.wisc.edu/gallery/album74
It really is worth a visit, if only to see a collection of the fastest
machines of their time. I think if you're travelling some distance you
might be able to ask nicely and visit out of hours. I'm hoping to get
back there in a month or two for a better look.
One thing Yvette is after is a circular sliderule (as Cray apparently
used), does anyone here have one they could spare for a good cause?
There is a site with details at http://www.cfmit.org/
--
Lawrence Wilkinson lawrence at ljw.me.uk
The IBM 360/30 page http://www.ljw.me.uk/ibm360
>
>Subject: Re: VTSERVER booting
> From: Patrick Finnegan <pat at computer-refuge.org>
> Date: Sun, 19 Jun 2005 20:06:17 -0500
> To: "General Discussion: On-Topic and Off-Topic Posts" <cctalk at classiccmp.org>
>
>I'm not 100% sure if it's your problem, but 2.11BSD doesn't work on the
>F11 chipset (11/23) because it doesn't have split I&D space. You'll
>need to use 2.9BSD or earlier on it.
>
There are two versions of 2.11 if memory serves one with split
I&D and one without.
I have to check but fuzzy memory says 11/23 had I&D or was that later
with the J11? I distinctly remember the MMU of the 11/23 as being
software compatable with 11/34, 11/69 and 11/70 (to the limits of
Q18 or Q22).
I know the Falcon, LSI-11/ and 11/2 did not even have MMU.
Allison
>
>Subject: Re: DSD-440 and/or drive recommendation for PDP-8
> From: "vrs" <vrs at msn.com>
> Date: Sun, 19 Jun 2005 21:19:21 -0700
> To: "General Discussion: On-Topic and Off-Topic Posts" <cctalk at classiccmp.org>
>
>> Even if you don't have an RX8E its a trivial board
>> easily copied.
>
>I don't know about trivial (I am trying to do one). There are a lot of
>unobtainium parts in an RX8E that have to be designed out of a modern clone.
>(It's easily doable, but not trivial.)
Heres the trick. Don't copy the parts, copy the logic and use
standard TTL. Most all the functions are easily available or
or sometimes available in denser (newer relatively) parts.
Trying to copy verbatim will be hard unless you have a lot of
DEC junker cards to strip. However most of the logical functions
are simple so substitutions are not as scary.
Allison
>
>Subject: DSD-440 and/or drive recommendation for PDP-8
> From: "Joe R." <rigdonj at cfl.rr.com>
> Date: Sun, 19 Jun 2005 21:24:04 -0400
> To: cctalk at classiccmp.org
>
> I'm looking for a drive for my PDP-8 but I don't want something the
>weight and size of a RK-05. I also have several DSD-440 drives available to
>me. Does anyone know if these can be used on the PDP-8 or have any
>experience with them? Does anyone know what kind of controller they need
>when used in the 8? Or does anyone have a suggestion about what other
>drive can be used with the -8? I realize that the DSD-440 or other drive
>may not be original for the -8 but I just don't want to deal with something
>the size, weight and unreliability of a RK-05.
>
> Joe
The DSD440 is unknown for PDP-8. I've always seen them on PDP-11s.
The best and easiest disk for PDP-8 is a RX01 (or RX02) Floppy drive
and RX8E board. Thats nice as it was supported and easy hook up
and repair. Even if you don't have an RX8E its a trivial board
easily copied.
Allison
>
>Subject: Re: Several comments on crashed drive in PDP-11/23
> From: "Jerome H. Fine" <jhfinexgs2 at compsys.to>
> Date: Sun, 19 Jun 2005 23:57:17 -0400
> To: General Discussion: On-Topic and Off-Topic Posts <cctalk at classiccmp.org>
>
> >Vintage Computer Festival wrote:
>
>>I downloaded and ran the TU58 simulator on my PC, then booted RT11 off
>>that. I then created a bootable RT11 system disk and used the floppy to
>>boot. Then I could mount blank images in the TU58 simulator and copy
>>files from the PDP11 to the PC. Then you can use the PUTR utility to
>>extract files from the images on the PC into standard PC files.
>>
>>
>Jerome Fine replies:
>
>If you have a 5 1/4" RX50 floppy drive on a real DEC PDP-11,
>then it is probably much faster and easier to prepare
>the floppy on the PC with PUTR in the first place. If
>the floppy is 8" and SSSD (RX01), then I understand that
>a PC can also have the same, but I don't know if PUTR
>can also handle 8" SSSD floppy media on the PC. In any
>case, I have no experience with 8" floppy media on a PC.
It's possible to do RX01, but RX02 format is beyond what
the 765 based PC clopped controller can do. RX02 is not
the common style of DD format. To do SD format you need
to do a fair amount of work and likely a older or even
XT PC to insure the FM mode at 8" data rates is available.
Easier shot is the RX50 (SSDD, 96tpi).
>I am certainly NOT saying that a TU-58 is a bad solution,
>only that other solutions are often better if they are
>possible. If the only way to start is with a TU-58, then
>it is certainly better than nothing.
Bob, of spare time gizmos did it right. 8051 with 2 256kB
rams to look like a TU58. Save for one difference. The
seek time is way faster!
Allison
