When you are cleaning up and restoring a piece of classic computer
hardware, would you correct any minor mistakes made in the original
production?
I have a particular exmaple in mind. In bits on myu bench at the moment
is an HP2623 graphics terminal This was HP's response to the well-knwon
Tektronix terminals, it is, however a raster scanned device. But it has a
little state machine (or parhapes you could call it a very simple
procesosr) to speed up line drawing.
But wanyway...
On the video monitor PCB (HP call it the 'Sweep' PCB), there is a 74LS221
dual monostable chip in the common HP horixontal drive circuit, The first
half of the chip delays the sync pulse (and the delay time is controlled
by the horizontal centering control), the second half is triggered from
that and provides a pulse of known width to the horizontal drive transistor.
But I digress. You don't need to know any of that...
WHile cleaning the dirt off this PCB (HV attracts dust electrostatcally,
of couse), I noticed that the ground pin of this IC never made it through
the board. It's folded under the package, but has been effectively
surface-mount soldered to the pad when the board was wave-soldered. It
works, but I can't beeleive it's as good as a proper through-hole
connection, and it's certian;y not what was intended.
I could trivially desolder the IC, straighten the pin, and solder it back
properly. But should I? What would others do?
-tony
This is off-tipic for 2 main reasons : (1) the device in question is not
even 1 year old yet and (2) it's not a computer. But I suspect many
people here rememebr the MB Big Trak programmable toy tank from the early
1980s, and may even rememebr the Circuit Cellar article that involved
modifying one. So anyway...
You can buy a modrn version of the Big Trak in the UK now. To the user
it's much the same as the origianl version (I will mention differences)
later. Inside it's very different. The cheapest I've found it in a shop
is \pounds 34.00 in Hawkin's Bazaar. I've sene it for a few 10's of pence
cheapr on the web, but then youy have to add shipping.
Anyway, I am sure you know what I did to mine first, even before puttign
the batteryies in and trying it out.Yes, I took it apart. Given that many
people here are like me, I;ll explain how to do that.
1) Remove the battery cover (2 screws) and batteris if you have fitted
them.
2) Next take off the plastic strips joining the front and rear axles.
There are held on by 2 scres each (self-tapping screws with integral
washers). If you have put the optional stickers on, you have to peel them
off to get to these screws.. Then slide off the front and rear wheels.
the centre 9driving) wheels are held on to the gearbox output shaftes hy
a cetnre screw. Take these off next.
3) The next bit is non-obvious and had me puzzles for a minute or so. At
he back of the lower body thre's a gey plastic part the only purpose of
which is to hide a coupleof fixing screws. It's held in place by 4
plastic 'barbs', one each side and 2 along the rear endge. Carefully free
it and slide it out downards.
4) Ondo the tiny screw thus revealed. Ths holds the kybaord bezel in
plaec on top. Carefully unclip the bezel. On the underside of it is a
little pCB carrying the power switch and the IR (?) LED for the trailer
interfae. Undo the fixing screw, free the PCB amd set the bezel aside.
5) Back on the uinderside undo the dozen or so screws that hold the upper
body in palce. Put the Big Trak the right way up and carfully lift off
the upper bod. It's still linked by the flexiprint from the keybaord.
This is not pluged in, it's clamped agianst the control PCB by a plastic
clamp. Undo the 2 screws and lift of the clamp, then lift the kayboard
tail over the pillers. The upper body is not free,
6) Undo the 2 screws holdign the LED holder for the 'photon cannon' in
palce. This alos releases the loudspeaker which is lightly glued in place
under it. Free that too.
7) I should have deoslderesd the battery wires next, otherwise you hve
the lower body hanging on them, which is a bit of a pain. But I didn't.
Anyway, undo th 4 screws on the bottom of the gearbox. Remove the gearbox
complete with the control PCB downwards. The front axle is now free,
don't loose it.
8) It appears that the PCB can be removed by desoldering the motor
connectios (ML+, ML-, MR+. MR-) and then taking out the 4 corner screws.
I have not done this.
9) The gearbox is held together by 5 screws on top (and can be dismantled
with the PCB in palce). Take these out, turn the gearbox over, andlift
off the cover. You can now see all the gearing.
When reasbmbling note that there are several sies of selt-tapping screwn.
Don't get them in the wrong holes...
OK, diffeernces from the origianl version
It runs everyting (logic and motors) off a single 4.5V supply (3 D cells in
series) ratehr than +/-3V (4 D cells) and 9V (pp3) of the original version
The 'photon cannon' is now an LED (blue I susepct) not a filament lamp
The interface to the trailer is another LED (IR?) not a jack socket.
The cotnroller bvoard is totrally redesigned. There are 2 ICs. These are
epxoy-capped chip-on-board devices iwth a difference. They are not on the
main PCB, they are on little daugherboards which are then soldered to the
main PCB. The only think on the daughterboard is the epoxy-capped chip.
So in theory they can be replaced. The larger IC is clearly a
microcontroller (but not the TMS1000 of the origjnal one) in a package
which looks to have the smae footprintas a PLCC device, I am not sure what
the smaller IC is yet. Also on the PCB are a fair number of transsitors,
R's and Cs, etc. Mostly SMD devices.
The gearbox is redeisgned too. It's a lot easier to take apart and
reassmeble than the orignal one.
The magnetic clutch between the 2 motors is still there.
The optical feedback from the gearbox is still there too. In fact there
are now 2 IR light bariers, one for each motor, senmsing slots in the
next-to-last gear in the train. I beelive the origianl version only had
one light barrier.
For a fairly cheap toy it's well made. It can be taken apart and put
together again withut damage. Most of it is held together by self-tapping
screws, but the battery cover (which the average owner has to remvoe and
rrfit serveral times( is held on by machien screws going into tapped
inserts in the lower body. Nice!.
I may well do a Circuit Cellar style modification to mine (that is, a
serial interfce thart simulates keypresses. Or maybe repaec the
microcontroller totally.
-tony
Hello all,
I would like to play^H^H^H^H use some of my vintage portables the way
they were intended. For example, the Radio Shack Model 100 and the HP
110 Portable have 300bps internal modems. I have several other modems
thru which these machines could talk to a *NIX but I do not want to pay
for multiple phone lines in my home. How can I simulate a POTS
connection between these modems? I don't need a dial tone or ringing,
etc., as I can just tell one machine to connect and the other to answer.
Way back I heard that a few modems would work if the tip and ring wires
of the modems were crossed and connected, but most need something more
than that. Is there some simple circuit I could build? Surely I'm not
the first to want to avoid phone bills for such play^H^H^H^H testing.
Jim
Hi.
This week I had the pleasure to visit a Zuse Z23 on display in my city.
The Z23 is a first generation, discrete transistor and diode logic, drum
based machine with a small "cache" of core memory. It is hosted at the
Fraunhofer Institute for Experimental Software Engineering. Here is the
Press release regarding the Zuse machine:
http://www.iese.fraunhofer.de/de/presse/current_releases/PM-2010-16-210610z…
Ohh, if you can't read german (at all we are talking about a german
computer ;-) ) here is the english version:
http://www.iese.fraunhofer.de/press/current_releases/PM-2010-16-210610z23.j…
I took some pictures:
http://www.unixag-kl.fh-kl.de/~jkunz/bilder/Zuse_Z23/
The machine is in very good, but not working condition. I suspect it
would be a huge task to make it operable again, like the PDP-1
restoration. But it could be done. There is a complete set of
documentation, schematics, some spare parts, ... A former user of that
machine from the local university explained the machine and its inner
workings in detail to us. In one word: Fascinating! :-)
BTW: There is a tube based Zuse Z22 in working condition on display in
the ZKM museum in Karlsruhe (Germany). The Z23 is kind of a transistor
reimplementation of the Z22:
http://www.zkm.de/algorithmische-revolution/index.php?module=pagemaster&PAG…
Pictures only in the german version:
http://www.zkm.de/algorithmische-revolution/index.php?module=pagemaster&PAG…
--
\end{Jochen}
\ref{http://www.unixag-kl.fh-kl.de/~jkunz/}
Help...
I thought it would be "easy", just hook it up and let it go. I thought I had the proper load file (xpcsrv20.sys) in the /tftpboot directory. TCPDumps seemed to indicate that I should have a proper file with the last 6 digits of the MAC address. When I did that, I got the dump to indicate that something was transfering, but it looks like to petered out a ways through. It then goes into a silly arp sequence for a router in the 172.31... network while my network here is 192.168.0..
Questions:
1) What files should be in the /tftpboot directory?
2) Can I get access to the configuration through one of the serial ports (which one)?
3) Does anyone have one of these beasts "working" with a Linux host? I like to get it up and running.
I seem to have LOTS of documents on this, but none seem to be of the "step-by step" variety using a Unix (Linux) as the RARP/TFTP host.
I would prefer responses of list, but will read then here if you like.
Yes, these things are a bit old. Nobody really uses terminal servers these days, and while 40 ports is a bit much, I do have a bunch of serial devices I'd like to connect to. It looked like this would be a nice thing to get working.
Thanks,
...Tom Watson
I look back with great fondness in how Stan was such a help in digging up
historical ?facts? that I used in my book?He will truly be missed. All my
blessings.
Murray :)
(I just mentioned this in the thread about Burroughs part numbers, but
am putting it forth in a separate question in case it gets missed by
some readers.)
I have a core memory module from some Burroughs machine. It is a large
plug-in PCB about 16 inches square containing address & inhibit drivers
& such. Mounted to this is a near-same-size daughter-board containing
the sense amplifiers and the folded planar array of cores underneath a
metal shield.
It is organised as 32 KWords, 20 bits wide.
Any suggestions as to what machine this may have come from?
The 20 bits is the raw-memory-word width, one could speculate that some
of those bits were used for ECC and the machine-word-width is smaller,
although I'm not sure off the top of my head whether 4 bits is enough
for ECC on 16 bits (isn't it >log2(n) for 1-bit ECC on n bits?)
Hi,
I seem to remember that it was possible to save a FORTRAN IV program together with the FRTS runtime
system as a runnable .SV core image file. It seems that ADVENT does so...
Any idea?
(minor question: How does something like "R PROG ARGS" work?)
Regards,
Philipp
--
http://www.hachti.de
----------------------Original Message:
Date: Fri, 27 Aug 2010 00:30:24 -0700
From: Paxton Hoag <innfoclassics at gmail.com>
Subject: Re: Burroughs core memory module <-> machine
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Message-ID:
<AANLkTik=GrMyH34cMQahxmPcSYoraFQ5+7bS=e2a=7dD at mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
> I have a core memory module from some Burroughs machine. It is a large
> plug-in PCB about 16 inches square containing address & inhibit drivers &
> such. Mounted to this is a near-same-size daughter-board containing the
> sense amplifiers and the folded planar array of cores underneath a metal
> shield.
>
> It is organised as 32 KWords, 20 bits wide.
>
> Any suggestions as to what machine this may have come from?
>
Burroughs made several successful solid state accounting machines
before branching to full computers. Some of these had core memory and
were made in the 1970s.
...
Paxton
------------------------Reply:
Indeed, specifically a few models in the E series, but those memories were
*much* smaller; I think I still have one somewhere.
In the transition from mechanical base10 systems to solid state binary
systems some early E models used an interesting device called a "core
counter" which was effectively a non-volatile IC-sized solid state magnetic
divide-by-ten counter that emitted a pulse for every ten pulses in, a
logically direct replacement for the mechanical cams and 10-position gears
that most adding and accounting machines had been built around up till then.
32K x 20 sounds like it'd be out of a Medium or Large system (B2xxx or
greater).