17 Feb
2008
17 Feb
'08
2:32 p.m.
ok, been meaning to actually run something more than the little assembly
test apps on my HPs. How do folks bootstrap systems?
I've got 2 working HPs (HP-2108A and HP-2112A) and HS-TERMINAL (ie
HP-12531B) and I can get lights to light up with a 2400 baud connection
and this snippit running:
0001* set to TTY I/O number
0002 00011 SC EQU 11B
0004 00040 ORG 40B
0005 00040 106700 LOOP CLC 0 Turn off I/O interrupt system
0006 00041 060051 LDA TTYIO Load A-register with control word
0007 00042 102611 OTA SC Output control word
0008 00043 103711 STC SC,C Set control, clear flag
0009 00044 102311 WAIT SFS SC Wait for flag
0010 00045 024044 JMP WAIT Loop till ready
0011 00046 106511 LIB SC Input data word
0012 00047 106601 OTB 1 Output data word to switch register
0013 00050 024040 JMP LOOP Next char
0014 00051 160000 TTYIO OCT 160000 TTY input and print control word
LOOP 000040 01/0005 -- 01/0013
SC 000011 01/0002 -- 01/0007 01/0008 01/0009 01/0011
TTYIO 000051 01/0014 -- 01/0006
WAIT 000044 01/0009 -- 01/0010
4 symbols
14 lines assembled, 10 words generated
0 errors, 0 warnings
(also online as http://rikers.org/hp2100/ttyin.lst)
so what's next? Is there a tape loader for use over a 12531?
Also, I'm using a modified version of the perl assembler. I'd like to
actually use the native HP assembler under SIMH. The HP / CHM agreement
should allow for at least this much bootstrap information to be shared.
I'm happy to put up bootstrap instructions and tape images if someone
can provide them.
I don't have any other peripherals tested out yet. Not sure why my real
paper tape reader is not working yet. My paper tape punch is a long ways
from working and it's not HP equipment anyway.
I've got tape images that work in SIMH like the HP-Basic. Just want to
see these work on real hardware by loading them over a serial line.
There are quite a few HP collectors out there. How do you folks boot up
your hardware? ok, I could get an HP-7970 9 track, but I don't have that
either. It would seem useful to be able to bootstrap any given tape over
serial. I know it's only 2400 baud, but just about every HP collector
should have the needed hardware. :)
Help me out!
PS: Some nice list owner promised to get me a CD full of stuff to get
started, but alas, I'm afraid my process is never getting scheduled.
Perhaps I need more of a real-time OS.
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
18 Feb
18 Feb
1:53 a.m.
On Sun, 17 Feb 2008, Tim Riker wrote:
ok, been meaning to actually run something more than
the little assembly
test apps on my HPs. How do folks bootstrap systems?
I've got 2 working HPs (HP-2108A and HP-2112A) and HS-TERMINAL (ie HP-12531B)
and I can get lights to light up with a 2400 baud connection and this snippit
running:
There are up to four loader ROMs on the CPU board. You'll have to identify
which one you have. Normally there should always be the papertape loader
(used with the 8-BIT DUP REG card).
actually use the native HP assembler under SIMH. The
HP / CHM agreement
should allow for at least this much bootstrap information to be shared. I'm
happy to put up bootstrap instructions and tape images if someone can provide
them.
The bootstrap listings and manuals are all documented, have a look at the
manual numbered 12992-90001 (HP 12992 Loader ROMs).
I don't have any other peripherals tested out yet.
Not sure why my real paper
tape reader is not working yet. My paper tape punch is a long ways from
working and it's not HP equipment anyway.
I've got tape images that work in SIMH like the HP-Basic. Just want to see
these work on real hardware by loading them over a serial line.
The HP papertape reader is not serial. You may use one in RTE if you
generated the system appropriately.
Anyway you should get familiar with the bootstrapping procedure for the
HP1000. For example you need to preset the S register with the proper
values like ROM number and select code.
There are quite a few HP collectors out there. How do
you folks boot up your
hardware? ok, I could get an HP-7970 9 track, but I don't have that either.
I have that and I boot from that when I want to install RTE or run the
diagnostics. But usually I boot from the cartridge tape drive of the HP
2648 terminal (the "official" HP way when not using papertape).
It would seem useful to be able to bootstrap any given
tape over serial. I
know it's only 2400 baud, but just about every HP collector should have the
needed hardware. :)
Easy, use a BACI (the buffered serial interface card with the UART) and
emulate the binary load feature of the HP terminals; it's easy to
implement. In fact, I've written the counterpart. I am using the PERL
assembler to generate stand-alone programs and then transfer them to a
cartridge on one of our HP terminals. I then take the cartridge, put it
into the HP console terminal of the HP1000 and boot it.
This way I can boot from the HP7970 since I don't have the ROM to put into
the machine. I load the HP7970 loader from the HP2648 and then start that
loader.
Christian
6:40 a.m.
The bootstrap listings and manuals are all documented,
have a look at the
manual numbered 12992-90001 (HP 12992 Loader ROMs).
I don't have any other peripherals tested out yet.
Not sure why my real
paper
tape reader is not working yet. My paper tape punch is
a long ways from
working and it's not HP equipment anyway.
I've got tape images that work in SIMH like the HP-Basic. Just want to see
these work on real hardware by loading them over a serial line.
It would seem useful to be able to bootstrap any given
tape over serial. I
know it's only 2400 baud, but just about every HP collector should have
the
needed hardware. :)
I have a BOOT ROM that will work with a High Speed serial card. I'm not sure
the ROM was specifically written for that card but, it does work. I just
make a serial link to a PC and send the ABS image over the link.
Hmmmm... I think the ROM I use is the 12992C mini-cartridge tape loader.
The boot rom expects the image to be in HEX-ASCII format so, you'll have to
convert the ABS from binary to ASCII before sending it over the link. I
wrote a small Perl script to handle this. It works great!
I believe my HS card is configured for 1200 baud. At that speed it takes a
few seconds to upload a large image but, it's not that bad. IIRC < 1 minute
to upload HP BASIC.
Good luck,
SteveRob
steerex [at] ccvn [dot] com
19 Feb
19 Feb
12:49 a.m.
Main Account wrote:
I have a BOOT ROM that will work with a High Speed
serial card. I'm not sure
the ROM was specifically written for that card but, it does work. I just
make a serial link to a PC and send the ABS image over the link.
Hmmmm... I think the ROM I use is the 12992C mini-cartridge tape loader.
The boot rom expects the image to be in HEX-ASCII format so, you'll have to
convert the ABS from binary to ASCII before sending it over the link. I
wrote a small Perl script to handle this. It works great!
I believe my HS card is configured for 1200 baud. At that speed it takes a
few seconds to upload a large image but, it's not that bad. IIRC < 1 minute
to upload HP BASIC.
I'm interested to hear what rom this is. Can you give me an example of
what is expected in HEX-ASCII? HEX? not octal? :) Can I get a copy of
the Perl script?
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
10:42 a.m.
Steve Robertson
steerex [at] ccvn [dot] com
----- Original Message -----
From: "Tim Riker" <Tim at Rikers.org>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Tuesday, February 19, 2008 3:49 AM
Subject: Re: HP-21MX bootstrapping?
It's been a few years since I messed with this so, I hope I don't make any
errors here :-)
The problem is:
The machine is a 16-bit system. The compiled program listing is in OCT
(3-bit) format. The PC stores the binary ABS file as a byte (8-bits). The
serial loader expects a nibble (4-bits). Whew...
So, when you compile a prgram on the PC, the listing is generated as an
octal listing. IE:
...
102023 Blah
117703 Blah
...
Converted to hex:
102023 = 0x8413
117703=0x13C3
So the data in the PC would be stored in consecutive memory locations like
this: 84 13 13 C3
NOTE: The loader will only accept the ASCII characters "0-9" and "A-F"
over
the serial interface.
To send the data over the serial link, you have to send each nibble as an
ascci character::
"8","4","1","3","1","3","C","3"
When the serial loader gets the data, it converts the ASCII values to back
to HEX, then rotates the nibble left to create a 16-bit word out 4 nibbles
(4-bits each).
NOTE: The ABS file format also has "address" and "structure"
information
embedded in the file. So you can't just send that simple string. You would
have to include the preamble (address / byte count) and a checksum with the
data string.
Sorry, I can't find the perl script right now. I moved a couple of years ago
and can't find anything :-) . If you are keen on persuing this, let me know
and I'll try to hack something together for you.
The ROM's I have are the 12992C mini cartridge tape loader type. My
system(s) originally had a serial terminal attached. The terminal had a
minicartridge drive in it for mass storage. The operator would initiate the
ISL on the 1000, then send the data from the terminal to the 1000 thus
booting it. In my solution, the PC just emulates the terminal.
I may have a spare ROM if you beg a little :-)
BTW: Has anyone found a way to a copy the ROMs? I have tried numerous
equilivent devices but, haven't had any success replicating them.
See ya,
SteveRob
steerex [at] ccvn [dot] com
Main Account wrote:
> I have a BOOT ROM that will work with a High Speed serial card. I'm not
sure
> the ROM was specifically written for that card
but, it does work. I just
> make a serial link to a PC and send the ABS image over the link.
>
> Hmmmm... I think the ROM I use is the 12992C mini-cartridge tape loader.
>
> The boot rom expects the image to be in HEX-ASCII format so, you'll have
to
> convert the ABS from binary to ASCII before
sending it over the link. I
> wrote a small Perl script to handle this. It works great!
>
> I believe my HS card is configured for 1200 baud. At that speed it takes
a
> few seconds to upload a large image but, it's
not that bad. IIRC < 1
minute
to upload HP
BASIC.
I'm interested to hear what rom this is. Can you give me an example of
what is expected in HEX-ASCII? HEX? not octal? :) Can I get a copy of
the Perl script?
20 Feb
20 Feb
12:36 a.m.
Steve Robertson wrote:
[ how the loader works snipped ]
Sorry, I can't find the perl script right now. I moved a couple of years ago
and can't find anything :-) . If you are keen on pursuing this, let me know
and I'll try to hack something together for you.
Well, what I'm after is finding out if this is what an HP 264x terminal
sends. I guess I'll read the kernel source some.
There are no \n \r etc in there? just a continuous hex stream? That
should be simple enough to handle. I'll try digging out the loader.
The ROM's I have are the 12992C mini cartridge
tape loader type. My
system(s) originally had a serial terminal attached. The terminal had a
minicartridge drive in it for mass storage. The operator would initiate the
ISL on the 1000, then send the data from the terminal to the 1000 thus
booting it. In my solution, the PC just emulates the terminal.
I may have a spare ROM if you beg a little :-)
:) I may eventually. :)
BTW: Has anyone found a way to a copy the ROMs? I have
tried numerous
equivalent devices but, haven't had any success replicating them.
Jay West can do it. Not sure what he uses to copy them though.
Don't know if you saw in my other email that I was able to use the tape
loader rom to load hpbasic1.abs tonight! (well, last night as it's
morning now) It was just a timing issue. I wrote a little C app that
sends the file slowly while logging each block to stderr. I can likely
reduce the usleep() time and still have it work reliably. Fun times!
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
5:07 a.m.
New subject: Ebay grrrr #743
A nice-looking CBM Pet 2001 with chicklet keyboard (Item #300199788685),
described as:
"The computer powers up fine as you can see from the pictures, but it is
only half way through the booting up process. I know this because the
cassette motor is continuously turning. The screen is filled with little
patterns. This I hear is a simple problem to fix from what I have seen on
the net, memory chip problem etc. I am no expert on this machine so please
dont take me for gospel."
Erm, yes. Very simple to fix, if you happen to have a spare of the part (or
parts) that have failed. Otherwise, a complete bastard (pardon my Francais)
of a job.... and well beyond your average ePayer, I'd warrant.
And yet, from a "non expert" we're expected to believe this is an
"easy
repair"? Do people really fall for this sort of garbage.
Grrrrr.
(Cont. p93)
Cheers,
Ade.
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.516 / Virus Database: 269.20.8/1288 - Release Date: 19/02/2008
20:47
5:48 a.m.
New subject: Ebay grrrr #743
On Wed, Feb 20, 2008 at 02:07:53PM +0100, Ade Vickers wrote:
A nice-looking CBM Pet 2001 with chicklet keyboard
(Item #300199788685),
described as:
"The computer powers up fine as you can see from the pictures, but it is
only half way through the booting up process.... a simple problem to fix...
Erm, yes. Very simple to fix, if you happen to have a spare of the part (or
parts) that have failed. Otherwise, a complete bastard (pardon my Francais)
of a job.... and well beyond your average ePayer, I'd warrant.
Having fixed a number of PETs, the problem could be as simple as a failed
SRAM in low mem (you can swap with one from high mem to test), a failed
ROM (less likely), or most likely of all, cheap sockets. Power supply
problems can also be a possibility, but I'd think that's more likely
with newer PETs with DRAMs (you need three voltages to power 4116s).
Replacing sockets is not a difficult task if you are skilled with
a soldering iron. New sockets are not too expensive or difficult
to find (I use high-quality machine-pin sockets and get them by the
fistful at several to the dollar). I wouldn't recommend a novice start
a career in soldering by replacing two dozen sockets, though.
And yet, from a "non expert" we're
expected to believe this is an "easy
repair"? Do people really fall for this sort of garbage.
Well... clearly the seller is not an electronics expert, but if you
want to buy a 30-year-old computer of any variety and you don't know
the first thing about electronics, you'd better have a friend who does.
Given that it's a pickup offer, if I were in London, _I_ might
consider paying 50 GBP for a broken original PET (I sold a broken
one 10 years ago for $100 US). I don't know that I'd be happy to
pay much more than that, though. I think most ePay buyers would
prefer to have one that works out of the box.
I'll be curious to learn what this one goes for, eventually.
-ethan
--
Ethan Dicks, A-333-S Current South Pole Weather at 20-Feb-2008 at 13:40 Z
South Pole Station
PSC 468 Box 400 Temp -30.3 F (-34.6 C) Windchill -57.7 F (-49.8 C)
APO AP 96598 Wind 12.5 kts Grid 8 Barometer 682.0 mb (10553 ft)
Ethan.Dicks at usap.gov http://penguincentral.com/penguincentral.html
4:03 p.m.
New subject: Ebay grrrr #743
Well... clearly the seller is not an electronics
expert, but if you
want to buy a 30-year-old computer of any variety and you don't know
the first thing about electronics, you'd better have a friend who does.
When I make comments like that I get flamed alive. Oh well...
More seriously, if you want to run a real vintage computer (as opposed to
running emulators on modern hardware), then you have to rememebr that
you're running possibly 30 year old electronics. It's going to fail
sometime. It's going to need repair. So either you have to learn to do it
yourself (which is often not tht difficult, although sometimes you can
spend a long time trakcing doewn an oscure fault, or making a replacement
part, or...) Or you have to find (and be prepared to pay) a friend to do
the repair. And there aren't that many people who do classic computer
repairs.
-tony
11:15 p.m.
New subject: Ebay grrrr #743
Tony Duell Wrote:
Ah, I'm not in the office, so I can't flame the originator of that
comment... As I don't have the e-mail on my laptop here ;)
Still, to answer the question, I DO know the first thing about electronics -
it's the 2nd and 3rd things that I'm not so sure about.... Or, to put it
another way, I can do *basic* stuff, but I've not got the skills to do a
logic analysis, for example. In other words, I'd be reliant on chip-swapping
(and socket-swapping, assuming I had some compatible sockets, which I
don't).
Well...
clearly the seller is not an electronics expert, but if you
want to buy a 30-year-old computer of any variety and you don't know
the first thing about electronics, you'd better have a friend who does.
When I make comments like that I get flamed alive. Oh well... More seriously, if you want to run a real vintage
computer (as opposed to
running emulators on modern hardware), then you have to rememebr that
you're running possibly 30 year old electronics. It's going to fail
sometime. It's going to need repair. So either you have to learn to do it
yourself (which is often not tht difficult, although sometimes you can
spend a long time trakcing doewn an oscure fault, or making a replacement
part, or...) Or you have to find (and be prepared to pay) a friend to do
the repair. And there aren't that many people who do classic computer
repairs.
That's not a bad philosophy. I guess I've been lucky so far; the few
machines I have in my possession have been relatively healthy - the only
machine that's flaked out on me while in storage is an old Osborne luggable.
Oh, and I fried one of my QLs, but that was my own silly fault (switched the
monitor on after the QL = blown 8032)
Cheers,
Ade.
10:54 a.m.
New subject: Ebay grrrr #743
On 2/20/08, Ade Vickers <javickers at solutionengineers.com> wrote:
A nice-looking CBM Pet 2001 with chicklet keyboard
(Item #300199788685),
described as:
Hey, you could always go for this one:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=260213476011
Apparently it comes with an XBox, as well as unnamed products from
Apple, Atari and Sony!
3:58 p.m.
New subject: Ebay grrrr #743
A nice-looking CBM Pet 2001 with chicklet keyboard (Item #300199788685),
described as:
"The computer powers up fine as you can see from the pictures, but it is
only half way through the booting up process. I know this because the
cassette motor is continuously turning. The screen is filled with little
patterns. This I hear is a simple problem to fix from what I have seen on
the net, memory chip problem etc. I am no expert on this machine so please
dont take me for gospel."
Erm, yes. Very simple to fix, if you happen to have a spare of the part (or
parts) that have failed. Otherwise, a complete bastard (pardon my Francais)
of a job.... and well beyond your average ePayer, I'd warrant.
Actually, 99% of probelms in old PETs are due to the terrible IC sockets
used. I find it best to replace the lot with good turned-pin ones. That
may well cure this fault, if not, at least you know that's _not_ the
problem (nothing worst that trying to trace logic faults on a PCB with
bad connections!).
And yet, from a "non expert" we're expected to believe this is an
"easy
repair"? Do people really fall for this sort of garbage.
The only 'easy repair' is one that I've done. Even if I know what the
problem is, it may stil lnot be 'easy' to put right (the part maynot be
availale, it might be a right pain to fit, and so on).
When I buy on E-bay, I always assume the device will not be working (no
matter what the seller says, I rralise that I am buying 30 year old
computer hardware from a person who may not know that much about it), and
I asusme I am going to have to do some work on it. So far I've been
pleasantly supprised in most cases.
-tony
4:16 p.m.
New subject: Ebay grrrr #743
Actually, 99% of probelms in old PETs are due to the
terrible IC sockets
used. I find it best to replace the lot with good turned-pin ones. That
may well cure this fault, if not, at least you know that's _not_ the
problem (nothing worst that trying to trace logic faults on a PCB with
bad connections!).
Tony, believe in me: Turned pin sockets are an unnecessary spent money
The actual sockets are very good for anything. When the pet was made,
the socket technology was awful, but the nowadays sockets are good enough to
rival the turned pin "garry" sockets. Believe me!
21 Feb
21 Feb
10:55 a.m.
New subject: Ebay grrrr #743
Actually, 99% of probelms in old PETs are due to
the terrible IC sockets
used. I find it best to replace the lot with good turned-pin ones. That
may well cure this fault, if not, at least you know that's _not_ the
problem (nothing worst that trying to trace logic faults on a PCB with
bad connections!).
Tony, believe in me: Turned pin sockets are an unnecessary spent money
The actual sockets are very good for anything. When the pet was made,
the socket technology was awful, but the nowadays sockets are good enough to
rival the turned pin "garry" sockets. Believe me!
Maybe they've got better (although to be honest, I doubt it, most modern
stuff seems to be optimised for cheapness rather than quality). But the
difference in price is not that great, and keeping me from tearing my
hair out is worth something...
-tony
11:51 a.m.
New subject: Ebay grrrr #743
On Feb 21, 2008, at 1:55 PM, Tony Duell wrote:
I beg to disagree. I've had the odd bad contact in the formed-pin
sockets, I've never had a bad contact in turned-pin ones.
I also agree 100% with you here Tony. Machined-pin sockets are
worth every penny in my experience.
-Dave
--
Dave McGuire
Port Charlotte, FL
Actually, 99% of probelms in old PETs are due to the
terrible IC
sockets
used. I find it best to replace the lot with good turned-pin
ones. That
may well cure this fault, if not, at least you know that's _not_ the
problem (nothing worst that trying to trace logic faults on a PCB
with
bad connections!).
Tony, believe in me: Turned pin sockets are an unnecessary
spent money 
12:27 p.m.
New subject: Ebay grrrr #743
Tony Duell wrote:
I beg to disagree. I've had the odd bad contact in the formed-pin
sockets, I've never had a bad contact in turned-pin ones.
I suspect it's just that there's a huge variation in quality when it comes to
the formed-pin sockets - a turned-pin one is far more likely to work, but that
doesn't mean that there aren't some reliable formed-pin ones out there.
Whether it's possible to say that "manufacturer xyz is always good" is
another
matter...
Actually, 99% of probelms in old PETs are due to the
terrible IC sockets
used. I find it best to replace the lot with good turned-pin ones. That
may well cure this fault, if not, at least you know that's _not_ the
problem (nothing worst that trying to trace logic faults on a PCB with
bad connections!).
Tony, believe in me: Turned pin sockets are an unnecessary
spent money 
5:05 p.m.
New subject: Sockets (was: Ebay grrrr #743)
In article <47BDDEB7.70904 at gmail.com>,
Jules Richardson <jules.richardson99 at gmail.com> writes:
I suspect it's just that there's a huge
variation in quality when it comes to
the formed-pin sockets - [...]
What's the difference between these two? This conversation has been
going on and I confess I wouldn't be able to tell you which is which
when looking at the sockets.
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://www.xmission.com/~legalize/book/download/index.html>
Legalize Adulthood! <http://blogs.xmission.com/legalize/>
7:46 p.m.
New subject: Sockets (was: Ebay grrrr #743)
On 21 Feb 2008 at 18:05, Richard wrote:
> What's the difference between these two? This
conversation has
been
going on and I confess I wouldn't be able to tell
you which is which
when looking at the sockets.
Compare:
http://tinyurl.com/23b4x7
with
http://tinyurl.com/yt37e5
Note that the first type grasps the IC lead between two spring
leaves. The second (machine pin) type grasps the IC lead in a
"funnel". Generally, the second type grasps the IC pin more securely
with more contact surfaces.
I've been told that gold-plated sockets are a waste of money if what
you're going to be socketing is tinned-lead packages. Something
about gold-to-gold and tin-to-tin makes for the best results.
Does anyone have first-hand experience with this? Should I scrap all
of those gold-plated machine sockets that I have?
Cheers,
Chuck
8:06 p.m.
New subject: Sockets (was: Ebay grrrr #743)
In article <47BDD4FD.31419.2197E03 at cclist.sydex.com>,
"Chuck Guzis" <cclist at sydex.com> writes:
On 21 Feb 2008 at 18:05, Richard wrote:
Thanks, that explains it perfectly. Although I notice that neither of
those pages uses the terminology that's been used in this thread ;-).
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://www.xmission.com/~legalize/book/download/index.html>
Legalize Adulthood! <http://blogs.xmission.com/legalize/>
> What's the difference between these two?
This conversation has
been
going on and I confess I wouldn't be able to
tell you which is which
when looking at the sockets.
Compare:
http://tinyurl.com/23b4x7
with
http://tinyurl.com/yt37e5 
22 Feb
22 Feb
7:26 a.m.
New subject: Sockets (was: Ebay grrrr #743)
I've been told that gold-plated sockets are a
waste of money if what
you're going to be socketing is tinned-lead packages. Something
about gold-to-gold and tin-to-tin makes for the best results.
Well, it reduces corrosion. Whenever you run current through a
junction between metals of differing electronegativities, it promotes
corrosion at the junction. Thus, you want the chip pin surfaces and
the socket surfaces to be the same metal. (Gold is preferred over tin
in a different sense, in the sense that it corrodes less due to
environmental chemical effects. Which effect outweighs the other
depends on things like the local air pollutants, how much time the
metals spend in contact versus exposed to the environment, how much
current is run through the junctions....)
I'd say gold/gold is probably best, since it's the most
corrosion-resistant in both these senses. But whether I'd prefer
gold/tin or tin/tin depends on the tradeoffs.
/~\ The ASCII der Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents.montreal.qc.ca
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
21 Feb
21 Feb
10:58 a.m.
New subject: Ebay grrrr #743
Tony Duell wrote:
Actually, 99% of probelms in old PETs are due to the
terrible IC sockets
used. I find it best to replace the lot with good turned-pin ones. That
may well cure this fault, if not, at least you know that's _not_ the
problem (nothing worst that trying to trace logic faults on a PCB with
bad connections!).
I'll second that.
Another thing that fails is the power connector, where the wires from
the transformer and the smoothing capacitor arrive at the motherboard.
I ended up soldering the wires directly to the pins, and putting another
connector in the middle of the cable so I could get the board out.
Philip.
20 Feb
20 Feb
1:05 p.m.
Someone else pointed out to me that there are ROM images here:
http://www.bitsavers.org/bits/HP/1000_firmware/HP1000BootProms/
for some reason the roms only use the lower 4 bits of each byte. so each
word is in the lower half of four consecutive bytes in the ROM. I don't
see offhand how the IBL knows where to update slot information in the
ROMs. It also presumably needs to update address offsets as the code is
relocated? This part should be able to use relative addressing. Anyone
know how slot update works?
Any pointers on why IBL won't work on my HP-2108A ? :)
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
1:28 p.m.
On Feb 20, 2008 1:05 PM, Tim Riker <Tim at rikers.org> wrote:
Someone else pointed out to me that there are ROM
images here:
http://www.bitsavers.org/bits/HP/1000_firmware/HP1000BootProms/
for some reason the roms only use the lower 4 bits of each byte. so each
word is in the lower half of four consecutive bytes in the ROM. I don't
see offhand how the IBL knows where to update slot information in the
ROMs. It also presumably needs to update address offsets as the code is
relocated? This part should be able to use relative addressing. Anyone
know how slot update works?
HP 1000 E-Series and F-Series
Computer Microprogramming
Reference Manual
http://www.bitsavers.org/pdf/hp/1000/2109-90004_1000EF_uPrgRef_Apr80
(.pdf extension missing on this file? rename as .pdf)
Appendix G, Page 8, INITIAL BINARY LOADER firmware listing.
From a quick scan of the comments it appears that the
IBL firmware
handler recognizes I/O instructions as it unpacks them from the loader
PROM and fixes them up on the fly if the select code is less than 10.
-Glen
3:45 p.m.
On 20 Feb 2008 at 13:28, Glen Slick wrote:
From a quick scan of the comments it appears that the
IBL firmware
handler recognizes I/O instructions as it unpacks them from the loader
PROM and fixes them up on the fly if the select code is less than 10.
Greater than or equal to 10 (it leaves select codes less than 10, e.g.,
DCPC instructions, alone). It actually adds SC-10 as derived from the S
register to the I/O instructions. This is how it can configure both select
codes of a two-card device (e.g., 7900 disc) from the single select code
entered via the front panel.
On 20 Feb 2008 at 14:05, Tim Riker wrote:
It also presumably needs to update address offsets as
the code is
relocated?
IBL MRG instructions reference the current page, i.e., contain only the
offset from the current page, so they are relocatable to any page. In
practice, the microcode copies the ROM into the last page of physical
memory <= 32K.
Any pointers on why IBL won't work on my HP-2108A
? :)
Bad ROM? I presume you've tried storing values into the last 64 locations
in memory via the front panel to confirm that writing memory works. You
might try a scope on the ROM outputs to see if they move when you press the
IBL button.
-- Dave
21 Feb
21 Feb
9:21 a.m.
J. David Bryan wrote:
Thanx for the help! :)
It has the soldered in default rom (1816-0420), as well as another that
I have not yet identified (1816-0639). I have tried setting memory for
the address range 177700-177777 and I can set/reset values everywhere
I've tried. Preset clears overflow, and hitting IBL (with S as 001100)
sets it, so it looks like the microcode does run. Nothing changes in
memory. Hmm.. I take that back, I think that 177777 may change and
nothing else. I'll have to try again tonight.
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
Any pointers
on why IBL won't work on my HP-2108A ? :)
Bad ROM? I presume you've tried storing values into the last 64 locations
in memory via the front panel to confirm that writing memory works. You
might try a scope on the ROM outputs to see if they move when you press the
IBL button.
10:47 a.m.
On 21 Feb 2008 at 10:21, Tim Riker wrote:
Thanx for the help! :)
You're welcome.
It has the soldered in default rom (1816-0420), as
well as another that I
have not yet identified (1816-0639).
The latter is an early 7900/7905 disc boot ROM. It was superceded because
it will not boot a multiple-drive MAC (7905/06/20/25) system.
I have tried setting memory for the address range
177700-177777 and I
can set/reset values everywhere I've tried.
Note that memory ranges go up to 077777 octal (32K words). Bit 15 is
reserved for indirection. So the upper range would be 077700-077777.
Preset clears overflow, and hitting IBL (with S as
001100) sets it....
The IBL microcode sets overflow as an error indication. This occurs in two
cases: either it cannot find any read/write location in memory, or the
select code entered in the S register is less than 10 octal. (11 octal
should be OK. I presume that you're hitting STORE and that if you select
another register and then S again, the correct value returns?)
The memory check starts at 077700 (32K - 64) and decreases by 4K each time
until the read data matches the written data (the data used is the two's
complement of the address). How much memory is in the machine, is the
address range configured correctly on each memory card, is the memory
controller good, and is there a MEM card in place or not?
If overflow is setting, the ROM isn't being read, because the microcode is
bailing out before that.
-- Dave
11:44 a.m.
J. David Bryan wrote:
Thanx!
sorry, yes. I meant 077700-077777.
exactly
It has the
soldered in default rom (1816-0420), as well as another that I
have not yet identified (1816-0639).
The latter is an early 7900/7905 disc boot ROM. It was superceded because
it will not boot a multiple-drive MAC (7905/06/20/25) system. I have tried
setting memory for the address range 177700-177777 and I
can set/reset values everywhere I've tried.
Note that memory ranges go up to 077777 octal (32K words). Bit 15 is
reserved for indirection. So the upper range would be 077700-077777. Preset clears
overflow, and hitting IBL (with S as 001100) sets it....
The IBL microcode sets overflow as an error indication. This occurs in two
cases: either it cannot find any read/write location in memory, or the
select code entered in the S register is less than 10 octal. (11 octal
should be OK. I presume that you're hitting STORE and that if you select
another register and then S again, the correct value returns?) The memory check starts at 077700 (32K - 64) and
decreases by 4K each time
until the read data matches the written data (the data used is the two's
complement of the address). How much memory is in the machine, is the
address range configured correctly on each memory card, is the memory
controller good, and is there a MEM card in place or not?
32kw installed. I have tried some simple write/read tests all the way
through memory and they worked. I have not touched the range settings on
the card, but it sounds like a good thing to try is just remove 3 of the
4 cards and see if that changes anything.
If overflow is setting, the ROM isn't being read,
because the microcode is
bailing out before that.
Agreed. The question is why. :)
FYI: I'm started taking notes here:
http://rikers.org/wiki/HardwareNotes
I note the original installed boards there. At present just the labels,
I'll add part numbers soon. I have moved the IO cards around too.
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
2:59 p.m.
On 21 Feb 2008 at 12:44, Tim Riker wrote:
Agreed. The question is why. :)
The microcode sets overflow on at the start of the loader routine and turns
it off after verifying the starting memory location and the select code.
These are only these two jumps out of the routine with overflow set.
If P is set as expected by IBL (i.e., 077700 for 32K), then the select code
= 10 check is failing. Perhaps the ribbon cable from
the front panel to
the CPU is bad.
If P is set to 177700 (bit 15 on), then the memory check is failing.
FYI: I'm started taking notes here:
http://rikers.org/wiki/HardwareNotes
Your list shows slot 118 open. The HP installation manual recommends that
the memory controller go in 118 and the memory cards follow above in the
cage.
You might try reseating the ribbon cable that connects the left-side edge
connectors of the memory cards and the controller card. (The right side
connectors are open.) Also, the ribbon cable that attaches to the DCPC
card.
Note that the memory controller must be strapped for an M-Series; if it's
strapped for an E-Series, the timing will be off.
-- Dave
22 Feb
22 Feb
11:29 p.m.
J. David Bryan wrote:
You might try reseating the ribbon cable that connects
the left-side edge
connectors of the memory cards and the controller card. (The right side
connectors are open.) Also, the ribbon cable that attaches to the DCPC
card.
Well, you were close. :) The card edge connector from the cable on the
front panel has some pins that barely touch. To confirm I swapped (shh!)
the front panel from the 2112 with the 2108. Sure enough, now the 2108
will IBL, and the 2112 will not.
Anyone recommend a place to order a replacement the 50 pin card edge
connector?
I suspect Crisis will sell me a replacement front panel (part
02108-60024) and it does have one broken switch plate on it. Don't know
for how much. :/
The switches work ok, so I think all I need to do is replace the edge
connector. I suppose I could cut one off the end of the memory card
cable, but that sounds like a Bad Idea, as I should be able to find one
someplace.
Thanx for the help!
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
23 Feb
23 Feb
8:17 a.m.
On Fri, Feb 22, 2008 at 11:29 PM, Tim Riker <Tim at rikers.org> wrote:
Well, you were close. :) The card edge connector from the cable on the
front panel has some pins that barely touch. To confirm I swapped (shh!)
the front panel from the 2112 with the 2108. Sure enough, now the 2108
will IBL, and the 2112 will not.
Anyone recommend a place to order a replacement the 50 pin card edge
connector?
What is on the other end of the cable with the 50-pin card edge
connector? Is that somewhat permanently attached to the front panel?
I just took a quick look at a 2113B front panel and without
dissasembling it or taking a close look it appears it might be
permanently attached. If you can't easily remove the whole cable and
make an equivalent new cable, there might be enough extra length in
the cable to cut of the bad connector and press on a new 50-pin card
edge connector.
If you haven't built any IDC cables I wouldn't recommend trying it on
the permanently attached front panel cable the first time you try
until you know you'll get good results. The first couple of times I
tried building some IDC cables without proper tools I ruined a couple
of connectors and a couple of inches of cable. Then I acquired some
proper IDC press tools and built a couple of sets of Pertec tape drive
cables with 50 conductor ribbon cable and IDC headers on one end and
card edge connectors on the other without too much trouble.
I bought some part# 15-CE50 at www.emtel.com and I think they were
cheaper a few months ago there than they are today. Now they list
them as $26.90 for q10.
Still cheaper than Digi-Key
Digi-Key: CW CCE50S-ND $5.89, 3M MCE50K-ND $11.67
3:30 p.m.
On 23 Feb 2008 at 0:29, Tim Riker wrote:
Well, you were close. :) The card edge connector from
the cable on the
front panel has some pins that barely touch.
I thought I was closer than that -- I had written, "Perhaps the ribbon
cable from the front panel to the CPU is bad." :-)
Anyone recommend a place to order a replacement the 50
pin card edge
connector?
Digi-Key may have a complete replacement cable; I'll have to look at my
2108B on Monday when I get to work. But, from memory, the cable has a 50
pin IDC edge connector on one end and a four-row IDC "transition" connector
on the other that's soldered into the board. Part number C3PET-5006G-ND
might be it.
But you'd have to unsolder the 50-pin transition connector from the front
panel PCB. It'd be easier to cut off and replace the edge connector, I
believe.
The OEM part for the edge connector is 3M part number 3415-0001.
I suppose I could cut one off the end of the memory
card cable....
Very unlikely to work, I'm afraid. Those are pretty much one-time assembly
connectors, and I've had very poor luck in separating the two parts of the
connector, removing the old cable bit, and reusing it. Nine times out of
ten, I'd wreck the connector in getting it apart (most often, I'd wind up
pulling one of the IDC pins out with the cable bit). Much better to try a
new one.
On 23 Feb 2008 at 8:17, Glen Slick wrote:
...there might be enough extra length in the cable to
cut of the bad
connector and press on a new 50-pin card edge connector.
I think that's the way to go here.
The first couple of times I tried building some IDC
cables without
proper tools I ruined a couple of connectors and a couple of inches of
cable.
I'll second that. With an arbor press and mandrels designed for to receive
the connector halves, it goes together easily. If you have to use Vise-
Grips, it's quite difficult. A bench vise with smooth jaw faces is a bit
better if you do things slowly and carefully and ensure that the parts and
the cable line up correctly as you squeeze them together. It takes quite a
bit of force to punch through 50 conductors simultaneously.
-- Dave
5:14 p.m.
From: jdbryan at acm.org
snip
I'll second that. With an arbor press and mandrels designed for to receive
the connector halves, it goes together easily. If you have to use Vise-
Grips, it's quite difficult. A bench vise with smooth jaw faces is a bit
better if you do things slowly and carefully and ensure that the parts and
the cable line up correctly as you squeeze them together. It takes quite a
bit of force to punch through 50 conductors simultaneously.
-- Dave
Hi
It can be done with visegrips but you need to support the entire length
of the connector. Two Iron flat stocks should work.
I like to us a bench vise with a couple pieces of flat stock if I
don't have smooth jaws on the vise.
Dwight
_________________________________________________________________
Helping your favorite cause is as easy as instant messaging.?You IM, we give.
http://im.live.com/Messenger/IM/Home/?source=text_hotmail_join
5:45 p.m.
On 23 Feb 2008 at 18:30, J. David Bryan wrote:
I'll second that. With an arbor press and
mandrels designed for to receive
the connector halves, it goes together easily. If you have to use Vise-
Grips, it's quite difficult. A bench vise with smooth jaw faces is a bit
better if you do things slowly and carefully and ensure that the parts and
the cable line up correctly as you squeeze them together. It takes quite a
bit of force to punch through 50 conductors simultaneously.
Even a cheapie hand press, such as Jameco 73252 (Hanlong IDC Crimping
Tool) at $16 will pay for itself in connections not botched. The
problem with a Vise-Grip or any sort of pliers is that the jaws move
in an arc, meaning that one side gets more pressure than the other.
Unless you're very careful, you can end up bending some of the pins.
A bench or woodworker's vise with both jaws faced by wood or plastic
(say, ABS or HDPE) will work in a pinch, particularly if you can
rabbet a slot in one side to hold the connector. Most bench vises
have patterned hard steel jaws which, if not faced with softer
material, will leave an impression in the connector body and could
possibly deform the pin holes. A good size woodworker's clamp might
also do the trick (I haven't tried it, having a hand press in my
toolkit).
Leave a little excess cable poking through the connector and flush-
trim with a fresh utility- or X-acto knife for a nice-looking job.
That's the way I do it, anyway.
Cheers,
Chuck
6:17 p.m.
On Saturday 23 February 2008 20:45, Chuck Guzis wrote:
On 23 Feb 2008 at 18:30, J. David Bryan wrote:
I have a pair here, I just looked at 'em, and the only markings I see that
would indicate a model or part number say "6SP" -- and the nifty thing is
that they have pads on the ends of the jaws that swivel, so that with proper
adjustment they'd do exactly the job needed. :-)
--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, ?a critter that can
be killed but can't be tamed. ?--Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
I'll second that. With an arbor press and
mandrels designed for to
receive the connector halves, it goes together easily. If you have to
use Vise- Grips, it's quite difficult. A bench vise with smooth jaw
faces is a bit better if you do things slowly and carefully and ensure
that the parts and the cable line up correctly as you squeeze them
together. It takes quite a bit of force to punch through 50 conductors
simultaneously.
Even a cheapie hand press, such as Jameco 73252 (Hanlong IDC Crimping
Tool) at $16 will pay for itself in connections not botched. The
problem with a Vise-Grip or any sort of pliers is that the jaws move
in an arc, meaning that one side gets more pressure than the other.
Unless you're very careful, you can end up bending some of the pins.
6:39 p.m.
On 23 Feb 2008 at 21:17, Roy J. Tellason wrote:
I have a pair here, I just looked at 'em, and
the only markings I see that
would indicate a model or part number say "6SP" -- and the nifty thing is
that they have pads on the ends of the jaws that swivel, so that with proper
adjustment they'd do exactly the job needed. :-)
The only one I've ever seen in nearly 20 years is the Hanlong HT-214,
even with GC or Eagle branding on it. See:
http://www.hanlong.com.tw/largeimages/16/16a.htm
Prices all over the place--Jameco at $15.95 is the cheapest that I'm
able to find.
Cheers,
Chuck
7:03 p.m.
On Saturday 23 February 2008 21:39, Chuck Guzis wrote:
On 23 Feb 2008 at 21:17, Roy J. Tellason wrote:
Looks good. But that's not what I have on hand here, which looks more like
this:
http://ecx.images-amazon.com/images/I/11z1zBFKbBL._AA160_.jpg
--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, ?a critter that can
be killed but can't be tamed. ?--Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
I have a pair here, I just looked at 'em,
and the only markings I see
that would indicate a model or part number say "6SP" -- and the nifty
thing is that they have pads on the ends of the jaws that swivel, so
that with proper adjustment they'd do exactly the job needed. :-)
The only one I've ever seen in nearly 20 years is the Hanlong HT-214,
even with GC or Eagle branding on it. See:
http://www.hanlong.com.tw/largeimages/16/16a.htm
Prices all over the place--Jameco at $15.95 is the cheapest that I'm
able to find. 
7:29 p.m.
On 23 Feb 2008 at 22:03, Roy J. Tellason wrote:
Looks good. But that's not what I have on hand
here, which looks more like
this:
http://ecx.images-amazon.com/images/I/11z1zBFKbBL._AA160_.jpg
Ah, the Vise-grip C-clamp. There's also a Pony clamp that's a work-
alike. If you wanted to go a notch up, I recall seeing this thing at
my favorite Cheap Chinese Tool Store:
http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=3552
With some modification, that'd work pretty well, it seems to me.
Might be useful for other things as well, depending on your
inclinations. Make one heckuva walnut cracker...
Cheers,
Chuck
7:40 p.m.
On Saturday 23 February 2008 22:29, Chuck Guzis wrote:
On 23 Feb 2008 at 22:03, Roy J. Tellason wrote:
Yeah it would. Dunno if I'd wanna clean up the mess afterwards, though. :-)
The other thing I've used is a small vise, with some padding on the jaws, as
others have mentioned. I don't do a *lot* of those connectors, though some
years back I got a box of 36-pin "centronics" type and some ribbon cable. I
don't think I used 'em all up, though.
--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, ?a critter that can
be killed but can't be tamed. ?--Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
Looks good. But that's not what I have on
hand here, which looks more
like this:
http://ecx.images-amazon.com/images/I/11z1zBFKbBL._AA160_.jpg
Ah, the Vise-grip C-clamp. There's also a Pony clamp that's a work-
alike. If you wanted to go a notch up, I recall seeing this thing at
my favorite Cheap Chinese Tool Store:
http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=3552
With some modification, that'd work pretty well, it seems to me.
Might be useful for other things as well, depending on your
inclinations. Make one heckuva walnut cracker... 
8:09 p.m.
Chuck Guzis wrote:
On 23 Feb 2008 at 22:03, Roy J. Tellason wrote:
Maybe for English walnuts but it wouldn't stand up to real (black)
walnuts. Might not even last until the water got hot.
Looks good. But that's not what I have on
hand here, which looks more like
this:
http://ecx.images-amazon.com/images/I/11z1zBFKbBL._AA160_.jpg
Ah, the Vise-grip C-clamp. There's also a Pony clamp that's a work-
alike. If you wanted to go a notch up, I recall seeing this thing at
my favorite Cheap Chinese Tool Store:
http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=3552
With some modification, that'd work pretty well, it seems to me.
Might be useful for other things as well, depending on your
inclinations. Make one heckuva walnut cracker...
Cheers,
Chuck
7:58 p.m.
On Saturday 23 February 2008 20:45, Chuck Guzis wrote:
On 23 Feb 2008 at 18:30, J. David Bryan wrote:
After buying a proper IDC crimper, I couldn't figure out why I would
have ever tried to use something else to crimp the connectors,
especially considering how inexpensive it was, and how much easier it
was to crimp connectors without fouling them up.
Pat
--
Purdue University ITAP/RCAC --- http://www.rcac.purdue.edu/
The Computer Refuge --- http://computer-refuge.org
I'll second that. With an arbor press and
mandrels designed for to
receive the connector halves, it goes together easily. If you have
to use Vise- Grips, it's quite difficult. A bench vise with smooth
jaw faces is a bit better if you do things slowly and carefully and
ensure that the parts and the cable line up correctly as you
squeeze them together. It takes quite a bit of force to punch
through 50 conductors simultaneously.
Even a cheapie hand press, such as Jameco 73252 (Hanlong IDC Crimping
Tool) at $16 will pay for itself in connections not botched. The
problem with a Vise-Grip or any sort of pliers is that the jaws move
in an arc, meaning that one side gets more pressure than the other.
Unless you're very careful, you can end up bending some of the pins.
24 Feb
24 Feb
11:11 a.m.
After buying a proper IDC crimper, I couldn't
figure out why I would
have ever tried to use something else to crimp the connectors,
especially considering how inexpensive it was, and how much easier it
was to crimp connectors without fouling them up.
I have a small smooth-jaw vice on my electronics workbench which I've
found very useful for holding cables when I'm soldering and crimping IDC
connectors. After a little practice, I've never mangled a connector when
crimping it. The 'practice' taught me how much force on the vice screw
handle is necessary to crimp the connector, if things don't seem to be
going smoothily, I don't just apply more force, I stop and find out
what's misaligned.
The only connectors I have problems with (obviously) are the 'transition'
type with the solder pins sticking out. For the 2-row ones, I put the
pins into one of those solderless plugblock breadboards (about the only
use for such things :-)) and stick that in the vice. I don't use the
4-row transition connectors, if I did, I'd probably drill a metal block
to thake the pins.
On thhe other hand, in a moment of weakness about 15 years ago I bought
the cheap IDC crimp pliers that Radio Shack sold. I never got on with
them. I ruined more connectores using those that I ever did with the vice.
-tony
2:17 p.m.
On 24 Feb 2008 at 19:11, Tony Duell wrote:
I have a small smooth-jaw vice on my electronics
workbench which I've
found very useful for holding cables when I'm soldering and crimping IDC
connectors. After a little practice, I've never mangled a connector when
crimping it. The 'practice' taught me how much force on the vice screw
handle is necessary to crimp the connector, if things don't seem to be
going smoothily, I don't just apply more force, I stop and find out
what's misaligned.
An inexpensive imported drill press vise works well if you don't need
to crimp the IDC in the middle of a cable:
http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=30999
At $10, it's probably one of the cheapest alternatives to a real IDC
crimp tool.
Cheers,
Chuck
25 Feb
25 Feb
3:05 p.m.
I had written:
I'll have to look at my 2108B on Monday when I get
to work. But, from
memory, the cable has a 50 pin IDC edge connector on one end and a
four-row IDC "transition" connector on the other that's soldered into
the board.
That's exactly the case (I've looked :-). I think that replacing the card-
edge IDC connector would be a lot easier than unsoldering the transition
connector and replacing with a pre-assembled cable.
-- Dave
12 Mar
12 Mar
1:16 p.m.
J. David Bryan wrote:
Digi-Key may have a complete replacement cable;
I'll have to look at my
2108B on Monday when I get to work. But, from memory, the cable has a 50
pin IDC edge connector on one end and a four-row IDC "transition" connector
on the other that's soldered into the board. Part number C3PET-5006G-ND
might be it.
But you'd have to unsolder the 50-pin transition connector from the front
panel PCB. It'd be easier to cut off and replace the edge connector, I
believe.
The OEM part for the edge connector is 3M part number 3415-0001.
The cable assembly was cheaper than the connector itself, though not the
factory gray. I ordered a C3PES-5006G-ND and put it in last night. Seems
to be working fine now. IBL now loads code into memory.
Thanx again!
I wish I had a full diagnostics virtual tape image to more fully check
out on all my systems.
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
7:52 p.m.
Its easy to bootstrap HP-IPL/OS, which has decent diagnostic capability to
check
out 21MX machines. Look at:
http://www.infionline.net/~wtnewton/oldcomp/hp2100/
----- Original Message -----
From: "Tim Riker" <Tim at rikers.org>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Cc: "On-Topic Posts Only" <cctech at classiccmp.org>
Sent: Wednesday, March 12, 2008 4:16 PM
Subject: Re: HP-21MX bootstrapping?
J. David Bryan wrote:
Digi-Key may have a complete replacement cable;
I'll have to look at my
2108B on Monday when I get to work. But, from memory, the cable has a 50
pin IDC edge connector on one end and a four-row IDC "transition"
connector
on the other that's soldered into the board. Part number C3PET-5006G-ND
might be it.
But you'd have to unsolder the 50-pin transition connector from the front
panel PCB. It'd be easier to cut off and replace the edge connector, I
believe.
The OEM part for the edge connector is 3M part number 3415-0001.
The cable assembly was cheaper than the connector itself, though not the
factory gray. I ordered a C3PES-5006G-ND and put it in last night. Seems
to be working fine now. IBL now loads code into memory.
Thanx again!
I wish I had a full diagnostics virtual tape image to more fully check
out on all my systems.
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
20 Feb
20 Feb
3:47 p.m.
Glen Slick wrote:
HP 1000 E-Series and F-Series
Computer Microprogramming
Reference Manual
http://www.bitsavers.org/pdf/hp/1000/2109-90004_1000EF_uPrgRef_Apr80
(.pdf extension missing on this file? rename as .pdf)
Appendix G, Page 8, INITIAL BINARY LOADER firmware listing.
So much good info to read. I had pulled down this file before, but not
gotten to reading it. Very interesting stuff. Thanx!
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
From a quick scan of the comments it appears that
the IBL firmware
handler recognizes I/O instructions as it unpacks them from the
loader
PROM and fixes them up on the fly if the select code is less than 10. 
6:47 p.m.
New subject: HP 1000 F Series 21MX
Hi All,
With all this talk about the HP - 1000 I forgot I have one of
these bad boys to sell.
Check it out if you get a chance on ebay #300200776101
Rob
Rob Borsuk
email: rborsuk at colourfull.com
Colourfull Creations
Web: http://www.colourfull.com
19 Feb
19 Feb
10:49 a.m.
Yo,
Anyone know where I can get a copy of hpasm.c? Jeffs site seems to be down.
Steve Robertson
steerex [at] ccvn [dot] com
20 Feb
20 Feb
12:28 a.m.
Steve Robertson wrote:
Anyone know where I can get a copy of hpasm.c? Jeffs
site seems to be down.
I've had much more luck with the perl script. Sources here:
http://rikers.org/hp2100/asm21-pl.txt
I've added some features to it like other characters in the mnemonics
and *-1 etc support in OCT.
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
21 Feb
21 Feb
9:34 a.m.
Steve Robertson wrote:
Anyone know where I can get a copy of hpasm.c? Jeffs
site seems to be down.
I emailed Jeff and found that his email is down too, and hosted
elsewhere. Anyone know where Jeff is?
As for the file... archive.org is your fiend:
http://web.archive.org/web/20020206230023/oscar.taurus.com/~jeff/2100/emula…
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
19 Feb
19 Feb
12:47 a.m.
Christian Corti wrote:
There are up to four loader ROMs on the CPU board.
You'll have to
identify which one you have. Normally there should always be the
papertape loader (used with the 8-BIT DUP REG card).
I think the early MX models and later M-series have the paper tape in
00. E/F Series modles should have 264x loader in 00?
One of my machines does (HP-2108A) does not seem to be able to load any
roms using the IBL. Not sure why. I'm currently using an HP-2112A which
does have a paper tape loader in 00.
The bootstrap listings and manuals are all documented,
have a look at
the manual numbered 12992-90001 (HP 12992 Loader ROMs).
yep, been there. :)
The HP papertape reader is not serial. You may use one
in RTE if you
generated the system appropriately.
Anyway you should get familiar with the bootstrapping procedure for the
HP1000. For example you need to preset the S register with the proper
values like ROM number and select code.
I'm doing this. My paper tape reader does not reply. The paper tape boot
code will drive a 12521 except for writing the control word. I do this
first and then call the loader.
I have that and I boot from that when I want to
install RTE or run the
diagnostics. But usually I boot from the cartridge tape drive of the HP
2648 terminal (the "official" HP way when not using papertape).
I don't think I have a 264x loader rom anywhere. If anyone has a spare,
I'd love to get one. I'd really like to see the 264x protocol duplicated
such that I can use a terminal type application and send file images
over as if they were being read from a 254x terminal.
I have a 2645 with dual drives, but I have no cartridges for it. You
have tape cartridges and drives that work? Sweet.
Easy, use a BACI (the buffered serial interface card
with the UART) and
emulate the binary load feature of the HP terminals; it's easy to
implement. In fact, I've written the counterpart. I am using the PERL
assembler to generate stand-alone programs and then transfer them to a
cartridge on one of our HP terminals. I then take the cartridge, put it
into the HP console terminal of the HP1000 and boot it.
This way I can boot from the HP7970 since I don't have the ROM to put
into the machine. I load the HP7970 loader from the HP2648 and then
start that loader.
do you have anything that emulates the binary load feature of the
terminals? What about a pointer to documentation on how this works?
When using my paper-tape loader and talking to a 12531, I can upload
simple tape images and run them. When loading the basic1.abs file I
have, it dies when it hits the first 0x0a in the file. I suspect that
the linux terminal code is getting in my way. I've tried this:
stty -F /dev/ttyUSB0 -ixon -ixoff crtscts cstopb raw -hup intr '' quit
'' erase '' kill '' eof '' eol '' start ''
stop '' susp '' rprnt ''
werase '' lnext '' flush '' 2400
and then I start paper-tape loader on the box and run:
$ cat basic1.abs > /dev/ttyUSB0
again, this works just fine for many abs files I've tried. None of those
appear to have 0x0a's in them.
Anyone have a tape image of the 264x loader? I could recreate it fromthe
listing, but an image would help. Now if it just does not have any
0x0a's in it, I'm one step closer.
http://www.cs.ubc.ca/~hilpert/e/HP21xx/software.html
has an "ASCII Binary Loader" That might be useful, but alas, I have not
yet gotten it to assemble. Working on it some. :)
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
2:06 a.m.
Tim Riker wrote:
http://www.cs.ubc.ca/~hilpert/e/HP21xx/software.html
has an "ASCII Binary Loader" That might be useful, but alas, I have not
yet gotten it to assemble. Working on it some. :)
Just a comment on that, it was written for a requirement similar to yours: to
download into the HP machine over a serial line from a modern machine
running a terminal program. Please note that the protocol/load format is no
relation to the HP Basic-Binary-Loader format: if you wish to use it you'll
have to convert your binaries to it's language (tis pretty simple but more work
none-the-less). The advantage of it is everything pertinent goes over the
serial line in printable ASCII, and it ignores everything else.
(It works fine for me but I created all my own support stuff (cross-assembler &
simulator) to work with it. As it was written to assemble under my own
assembler, there may/will be some minor diffs from the HP asm standards, but
the machine code is there in the listing to toggle it in without having to assemble
it.)
Based on other's responses I suspect you'll find a solution with an original HP
loader.
Another possibility might be to modify the old HP teleprinter loader:
http://www.cs.ubc.ca/~hilpert/e/HP21xx/software/bblteleprinter.html
It uses the HP BBL load format.
It was for loading over a serial line from the paper tape reader on a teletype,
but uses a 12531A IO interface, which was a bit-bang interface. It could be
converted for the 12531B/C by, in short, replacing the ReadByte routine with
instructions appropriate for the 12531B/C.
(Also note that my experience in this is limited to playing with a 2116 system).
2:19 a.m.
Brent Hilpert wrote:
(It works fine for me but I created all my own support
stuff (cross-assembler &
simulator) to work with it. As it was written to assemble under my own
assembler, there may/will be some minor diffs from the HP asm standards, but
the machine code is there in the listing to toggle it in without having to assemble
it.)
Heya Brent! :)
As you likely saw, I posted a source version of your ascii loader. What
differences does your assembler have?
The loader has no license info. I presume it's ok for me to post my source?
Based on other's responses I suspect you'll
find a solution with an original HP loader.
Another possibility might be to modify the old HP teleprinter loader:
http://www.cs.ubc.ca/~hilpert/e/HP21xx/software/bblteleprinter.html
I'll look into this as well. Thanx!
It uses the HP BBL load format.
Not sure what this is. Can you point me to any docs?
It was for loading over a serial line from the paper
tape reader on a teletype,
but uses a 12531A IO interface, which was a bit-bang interface. It could be
converted for the 12531B/C by, in short, replacing the ReadByte routine with
instructions appropriate for the 12531B/C.
ok
(Also note that my experience in this is limited to
playing with a 2116 system).
lucky dog. :)
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
11:21 p.m.
Tim Riker wrote:
The format is documented (or at least my understanding of it) at the
beginning of the above-linked file, based on my disassembly of the
machine code in the HP manual. Look at the other two period BBLs on my
site for comparison.
Actually, to be a little more accurate in terminology: the BBL (Basic Binary
Loader) bootstrap programs read data in ABS (ABSolute-address load record)
format. BBL & ABS date from the original 2116 systems. Originally the BBLs were
manually toggled into core at the top of memory.
ABS format (.abs) does require a fully-transparent, 8-bit, binary datapath.
There was also a relocatable record format usable at later stages in the bootstrap.
I believe the ABS format, with new BBLs for new devices, continued to be used with
the later 2100 systems but others can comment on that better than I can.
I will hazard a guess that if an HP serial-line loader for the 2100 shows
up it may look a lot like what I'm describing in modifying the old teleprinter
loader.
Didn't you have a 2116B with a broken aluminum frame, or was that someone else?
As you likely saw, I posted a source version of your ascii loader. What
differences does your assembler have?
The loader has no license info. I presume it's ok for me to post my source?
I would just suggest that you add a log/history comment with date/name under my
creation date/name comment to indicate changes and to distinguish it from my
original.
Another
possibility might be to modify the old HP teleprinter loader:
http://www.cs.ubc.ca/~hilpert/e/HP21xx/software/bblteleprinter.html
I'll look into this as well. Thanx!
It uses the HP BBL load format.
Not sure what this is. Can you point me to any docs? It was for
loading over a serial line from the paper tape reader on a teletype,
but uses a 12531A IO interface, which was a bit-bang interface. It could be
converted for the 12531B/C by, in short, replacing the ReadByte routine with
instructions appropriate for the 12531B/C.
ok (Also note
that my experience in this is limited to playing with a 2116 system).
lucky dog. :)
20 Feb
20 Feb
12:18 a.m.
Brent Hilpert wrote:
Tim Riker wrote:
I would just suggest that you add a log/history comment with date/name under my
creation date/name comment to indicate changes and to distinguish it from my
original.
done. Also added a converted bbltelprinter.asm lst and abs oh, and
uploaded the 12992K (paper tape BBL) source, list and tape image too.
http://rikers.org/hp2100/
I believe the ABS format, with new BBLs for new
devices, continued to be used with
the later 2100 systems but others can comment on that better than I can.
Just a brain fart. I was thinking ABS and seeing BBL and went "huh?" I
got it now. :)
I will hazard a guess that if an HP serial-line loader
for the 2100 shows
up it may look a lot like what I'm describing in modifying the old teleprinter
loader.
the BACI card is different to code to, and Jay says that's what the 264x
loader uses, not looked at that yet.
Didn't you have a 2116B with a broken aluminum
frame, or was that someone else?
Yes to the frame, but it's a 2116A. I don't know anyone else with an A.
Mine is still having power supply issues, I've just not gotten back to
it. If anyone else has an HP-2116A, I wanna hear from you! Still looking
for Volume2 of the docs to cover the different core layout on this machine.
By the way.... The issue with the paper-tape loader was just a timing
issue. There apparently is no flow control on my serial line, so I had
to slow down the upload. I've just loaded HP-Basic (hpbasic1.abs) over
the serial line using the paper-tape IBL. woot! First I've gotten basic
up and running on this HP-2112A. :)
The issue I was seeing with 0x0A was just the default Linux flushing.
adding a fflush() call in my test app got me past that. :)
Thanx for the help! I'll post the steps online someplace soon.
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
19 Feb
19 Feb
2:11 a.m.
Tim Riker wrote:
http://www.cs.ubc.ca/~hilpert/e/HP21xx/software.html
has an "ASCII Binary Loader" That might be useful, but alas, I have not
yet gotten it to assemble. Working on it some. :)
Updated the perl assembler and made some code changes to the listing. I
think I have it assembling now. Feel free to grab a copy if you like:
http://www.cs.ubc.ca/~hilpert/e/HP21xx/software/abl.html - original
http://rikers.org/hp2100/asm21-pl.txt - updated assembler
http://rikers.org/hp2100/abl.asm - source
http://rikers.org/hp2100/abl.lst - listing
http://rikers.org/hp2100/abl.abs - tape image
the perl script adds 20 trailing nulls. How may are really needed?
my basic1.abs has 0x80 as the last byte but it's after nulls. I assumed
that was not needed and removed it.
What tape image manipulation tools do we have?
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
21 Feb
21 Feb
9:50 a.m.
Replying to Terry who uses the a pass-through paper tape emulator and a
qbasic app to bootstrap. Copying the list...
Thanx for the comments. :)
Does the QBasic app do anything more than just copy the bytes? Ie: is
the PIC expecting a new format?
The method I use at the moment is that I have a 12531 set to 2400 baud
in slot 11. I'm also using this as a console once I have hp basic or
similar running.
I'm running Linux so then I run stty thusly:
# stty -F /dev/ttyUSB0
0:0:800000fb:0:0:0:0:0:0:0:1:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0
which sets raw, clearing everything else and setting 2400 8n2
I set S to 001100, preset, ibl, run and the I use:
http://rikers.org/hp2100/abscopy.c
which shows some useful (useless?) information on stderr and dumps char
by char to stdout with a usleep() in between to get around the lack of
flow control. Thusly:
$ ./abscopy < basic1.abs > /dev/ttyUSB0
and everything loads up fine.
The paper tape bootloader does not expect to talk to a 12531, but works
as the 12531 inits to a state where it'll read the bytes. ie: the
software interface looks the same.
The downside of this is that I need to change the port setup in order to
talk to basic as it does not expect 8n2.
I may well look into using the BACI and the 264x loader, but I don't
(yet?) have a rom for that, and I expect it will require 2 serial ports.
Not a big deal I suppose. :)
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
22 Feb
22 Feb
1:05 a.m.
Tim Riker wrote:
An 'A'! .. even better. Found your photos again, see you've effectively fixed
the frame .. thank goodness for symmetry. I made schematics for 2116C power supply,
should be largely applicable to the A, at least to get an idea of the organisation,
if of any help.
The method I use at the moment is that I have a 12531
set to 2400 baud
in slot 11. I'm also using this as a console once I have hp basic or
similar running.
I'm running Linux so then I run stty thusly:
..
which sets raw, clearing everything else and setting
2400 8n2
..
The paper tape bootloader does not expect to talk to a
12531, but works
as the 12531 inits to a state where it'll read the bytes. ie: the
software interface looks the same.
Glad you figured out the paper tape BBL would work for the HST, I was
forgetting earlier that the byte-read interface was so similar.
I had a timing problem with the HST initially, it was resolved by sending 2
stop bits and now runs at 19200 bps. There's some explanation on:
http://www.cs.ubc.ca/~hilpert/e/HP21xx/io12531BCD.html
if you haven't seen it already. I take it you're using a 12531C (HST) if
you're
running at 2400. If you're already sending 2 stop bits though .. wonder if the
oscillator could have drifted far enough out to be a problem.
It's been interesting to hear about this IBL stuff added in the later machines.
Didn't you
have a 2116B with a broken aluminum frame, or was that someone else?
Yes to the frame, but it's a 2116A. I don't know anyone else with an A.
Mine is still having power supply issues, I've just not gotten back to
it. If anyone else has an HP-2116A, I wanna hear from you! Still looking
for Volume2 of the docs to cover the different core layout on this machine.
1:16 a.m.
Brent Hilpert wrote:
http://www.cs.ubc.ca/~hilpert/e/HP21xx/io12531BCD.html
if you haven't seen it already. I take it you're using a 12531C (HST) if
you're
Whoops, --> change that to 12531D
(HST)
(the C is a 2nd version of the B "Bufferred Teleprinter" for 110bps.)
4:07 p.m.
Brent Hilpert wrote:
I had a timing problem with the HST initially, it was
resolved by sending 2
stop bits and now runs at 19200 bps. There's some explanation on:
http://www.cs.ubc.ca/~hilpert/e/HP21xx/io12531BCD.html
if you haven't seen it already. I take it you're using a 12531C (HST) if
you're
running at 2400. If you're already sending 2 stop bits though .. wonder if the
oscillator could have drifted far enough out to be a problem.
Yes, I saw this before. I don't have my cable wired the same as yours. I
don't think my CSR/DSR/DCD go anyplace useful. Does your cable actually
get you valid CSR? or are you just forcing it? ie: you connect to "U -
+12V via 1k". I'll check the schematics, just have not yet. :)
An 'A'! .. even better. Found your photos
again, see you've effectively fixed
the frame .. thank goodness for symmetry. I made schematics for 2116C power supply,
should be largely applicable to the A, at least to get an idea of the organization,
if of any help.
There are schematics for the B power supply on page 5-91 of:
http://www.bitsavers.org/pdf/hp/2116/02116-9153_2116B_Vol2_Oct70.pdf
You've seen that I presume? That's what I'm working off. Al was trying
to track down an A version of Vol2 that he though he had, I've not heard
back on it. I could really use this if someone out there has it. :)
I was lead to believe that the C (perhaps only later C's?) used a
switching power supply. I assumed, perhaps incorrectly, that there are
different voltage outputs required for the A, B and C due to the
different core memory designs. From what I understand:
The A has 1 or 2 vertical 4kw stacks
The B has 1 or 2 horizontal 8kw stacks
The C has no stacks, but uses HP-2100[AS] compatible core cards
Both the A and B can use a 2nd chassis to add 1 or 2 more stacks. The C
can fit enough cards to take it to 32kw internally.
Brent Hilpert wrote:
Brent Hilpert wrote:
Mine is labeled "HS TERMINAL" so I presume it's a D. I'll double check
to be sure. I'll add more info on my notes page:
http://rikers.org/wiki/HardwareNotes
Thanx again for the help!
--
Tim Riker - http://Rikers.org/ - TimR at Debian.org
Embedded Linux Technologist - http://eLinux.org/
BZFlag maintainer - http://BZFlag.org/ - for fun!
http://www.cs.ubc.ca/~hilpert/e/HP21xx/io12531BCD.html
if you haven't seen it already. I take it you're using a 12531C (HST) if
you're
Whoops, --> change that to 12531D
(HST)
(the C is a 2nd version of the B "Buffered Teleprinter" for 110bps.)
23 Feb
23 Feb
12:58 p.m.
Tim Riker wrote:
Yes, I used that in part as I RE'd the C power supply.
I was in a similar situation to you, I didn't have the manuals specifically
for the C but (eventually) found the B manuals on bitsavers.
(I'll claim my schematic is layed out more sensibly, but that's
just my opinion .. depends on one's preferences.)
Brent Hilpert wrote:
The resistor to +12 is provided on the 12531 board, I just made the connections
to force CTS/DSR/DCD high in case one's terminal needs such.
No, there's nothing in the way of flow-control going on.
--
I had a timing problem with the HST initially, it
was resolved by sending 2
stop bits and now runs at 19200 bps. There's some explanation on:
http://www.cs.ubc.ca/~hilpert/e/HP21xx/io12531BCD.html
if you haven't seen it already. I take it you're using a 12531C (HST) if
you're
running at 2400. If you're already sending 2 stop bits though .. wonder if the
oscillator could have drifted far enough out to be a problem.
Yes, I saw this before. I don't have my cable wired the same as yours. I
don't think my CSR/DSR/DCD go anyplace useful. Does your cable actually
get you valid CSR? or are you just forcing it? ie: you connect to "U -
+12V via 1k". I'll check the schematics, just have not yet. :) An
'A'! .. even better. Found your photos again, see you've effectively fixed
the frame .. thank goodness for symmetry. I made schematics for 2116C power supply,
should be largely applicable to the A, at least to get an idea of the organization,
if of any help.
There are schematics for the B power supply on page 5-91 of:
http://www.bitsavers.org/pdf/hp/2116/02116-9153_2116B_Vol2_Oct70.pdf
You've seen that I presume? That's what I'm working off. Al was trying
to track down an A version of Vol2 that he though he had, I've not heard
back on it. I could really use this if someone out there has it. :)
I was lead to believe that the C (perhaps only later C's?) used a
switching power supply. I assumed, perhaps incorrectly, that there are
different voltage outputs required for the A, B and C due to the
different core memory designs. From what I understand:
The A has 1 or 2 vertical 4kw stacks
The B has 1 or 2 horizontal 8kw stacks
The C has no stacks, but uses HP-2100[AS] compatible core cards
Both the A and B can use a 2nd chassis to add 1 or 2 more stacks. The C
can fit enough cards to take it to 32kw internally.
The C still used linear but - as you suggest - with some voltage mods for the
changed core memory. The 2100's were the first to use switchers, TMK.
The B used anti-coincident addressing with the core, and I believe this is how
it achieved the doubling of memory capacity over the A. There may have
been little or no change to the power supply from the A to the B, as
anti-coincident addressing just requires (in principle) a little twiddling with
the address decoding.
If you're working on your A again sometime, I would be interested in some
information regarding board part numbers to fill out some of the question marks on
http://www.cs.ubc.ca/~hilpert/e/HP21xx/HP2116.html
regarding the 'A' model, for the sake of comprehensiveness.
Correlating some of these part numbers from your A with those of the B might
help resolve whether or not there are any differences in the power supplies
between the two. (The above page has a list of board part numbers.) Specifically:
Power Supply board A301
Power Supply board A302
Sense Amplifier
Inhibit Driver
Address Driver
Too bad about the missing blinkenlights cover .. need to find a real junker to
get one from, or, in the realm of plastics and silk-screening, it
might not be too difficult to recreate one.
6128
days inactive
6153
days old
60 comments
23 participants
participants (23)
-
alexandre-listas＠e-secure.com.br -
ard＠p850ug1.demon.co.uk -
bshannon＠tiac.net -
cc＠informatik.uni-stuttgart.de -
cclist＠sydex.com -
dkelvey＠hotmail.com -
ethan.dicks＠usap.gov -
glen.slick＠gmail.com -
hilpert＠cs.ubc.ca -
javickers＠solutionengineers.com -
jdbryan＠acm.org -
john＠guntersville.net -
jules.richardson99＠gmail.com -
legalize＠xmission.com -
mcguire＠neurotica.com -
mouse＠Rodents.Montreal.QC.CA -
pat＠computer-refuge.org -
philip＠axeside.co.uk -
rborsuk＠colourfull.com -
rtellason＠verizon.net -
silent700＠gmail.com -
steerex＠ccvn.com -
Tim＠Rikers.org