17 Jan
2003
17 Jan
'03
4:15 a.m.
I dug the 2001 out of the cupboard the other day - it's reorganise time,
try to get *even more* stuff into exactly the same amount of space as less
(but still lots of) stuff occupied. It never works, but it does give me an
excellent excuse to have a play around with some of them.
Unfortunately, the PET seems to have developed an odd fault: It won't take
a BASIC program, and some keywords seem to be knackered...
It reports 7167 bytes free, as per spec. But if I type:
10 REM blah
or
10 ?"HELLO"
it hangs. If I load a program from tape, it does one of two things:
a) Loads but won't run, typing LIST produces something like this:
10
2
READY.
b) It loads & hangs immediately after the READY. prompt comes up.
I tried to put together a FOR loop (the idea being I could use it to
display chunks of memory at a time, e.g.
FORI=0TO10:?I;:NEXT
Instead, the machine goes into an infinte loop displaying zero on a new
line each time (it's ignored the I variable, and the semicolon after the
PRINT command.
If I replace NEXT with NEXTI, it fails reporting NEXT without FOR.
So.... I figure the BASIC ROM has become slightly corrupted, OR I've got a
flakey memory chip which reads OK but doesn't write properly. The question
is, how to find out?
1) Can the BASIC ROM be swapped with one from, say, a 3032 or 4016? In
fact, which one IS the BASIC ROM?
2) If it's a dodgy memory chip, what's the best way of isolating it? I have
an oscilloscope, but nada skill in this sort of thing.
3) If, as my money is on, it's the BASIC ROM, can it be replaced with an
EPROM - if so, there's a whole gamut of additional questions to follow....
Thanks in advance!
--
Cheers, Ade.
Be where it's at, B-Racing!
http://b-racing.com
17 Jan
17 Jan
12:25 p.m.
New subject: SWTPC was(Re: PET 2001 oddity)
Adrian,
I used to sell the old PET 2001 from CBM complete with 'chiclet' keyboard.
Adrian Vickers wrote:
I dug the 2001 out of the cupboard the other day -
it's reorganise time,
try to get *even more* stuff into exactly the same amount of space as less
(but still lots of) stuff occupied. It never works, but it does give me an
excellent excuse to have a play around with some of them.
Unfortunately, the PET seems to have developed an odd fault: It won't take
a BASIC program, and some keywords seem to be knackered...
It reports 7167 bytes free, as per spec. But if I type:
10 REM blah
or
10 ?"HELLO"
it hangs. If I load a program from tape, it does one of two things:
a) Loads but won't run, typing LIST produces something like this:
10
2
READY.
b) It loads & hangs immediately after the READY. prompt comes up.
I tried to put together a FOR loop (the idea being I could use it to
display chunks of memory at a time, e.g.
FORI=0TO10:?I;:NEXT
Instead, the machine goes into an infinte loop displaying zero on a new
line each time (it's ignored the I variable, and the semicolon after the
PRINT command.
If I replace NEXT with NEXTI, it fails reporting NEXT without FOR.
So.... I figure the BASIC ROM has become slightly corrupted, OR I've got a
flakey memory chip which reads OK but doesn't write properly. The question
is, how to find out?
I would try the RAM first. Swap all the 8K RAM chips about. That used to do
wonders. It should be your first line of defense.
1) Can the BASIC ROM be swapped with one from, say, a 3032 or 4016? In
fact, which one IS the BASIC ROM?
2) If it's a dodgy memory chip, what's the best way of isolating it? I have
an oscilloscope, but nada skill in this sort of thing.
Swap the memory chips and keep track of what goes where. If a certain chip is
flakey it will show up.
3) If, as my money is on, it's the BASIC ROM, can
it be replaced with an
EPROM - if so, there's a whole gamut of additional questions to follow....
Proprietary ROMs have been upgraded using modern chips. See Mike Holley's
Southwest Technical Products site offering a ROM monitor upgrade:
http://home.attbi.com/~swtpc6800/ROM_Emulator/ROM_Index.htm
Eric
Thanks in advance!
--
Cheers, Ade.
Be where it's at, B-Racing!
http://b-racing.com
4:38 p.m.
--- Adrian Vickers <avickers(a)solutionengineers.com> wrote:
I dug the 2001 out of the cupboard the other day...
Can you give more information about *which* 2001 you have? There
are at least two major motherboard revisions for the static-RAM PETs.
They use different ROMs and RAMs.
Also, as I have one, there are also models of 2001 that have _dynamic_
RAMs. My 2001-N has 32K worth of 4116 chips (from the factory) and a
full-sized graphic keyboard. If it were for sale in Europe, it would
most likely been labelled a 3032, for comparison, but here in the States,
it says "2001" on the front.
I bring it up because I _think_ you are talking about the original
chicklet-keyboard SRAM PET, but you didn't come right out and say so.
Unfortunately, the PET seems to have developed an odd
fault: It won't
take a BASIC program, and some keywords seem to be knackered...
.
.
.
So.... I figure the BASIC ROM has become slightly
corrupted, OR I've got
a flakey memory chip...
1) Can the BASIC ROM be swapped with one from, say, a 3032 or 4016?
Most likely not. If your 2001 has 2114 SRAMs, perhaps. There are
upgrade ROMs (BASIC 2.0, anyway) _for_ the 2001
In fact, which one IS the BASIC ROM?
BASIC lives in several ROMs, the Kernel lives in several ROMs. BASICs
prior to 4.0 occupied $C000-$DFFF, and the Kernel occupied $E000-$FFFF,
minus the PIAs and VIAs, etc., at $E800. BASIC 4.0 starts at $B000
to add room for diskette commands.
In my PET, $9000, $A000 and $B000 are filled with user-supplied firmware
(better machine-language monitor, BASIC extensions and a tape speeder).
These sorts of things, plus ROMs for word processors and other applications
were somewhat typical amongst serious PET users. Dunno if yours has
anything like that, but if you find ROMs where you expect none to be,
that's probably why.
2) If it's a dodgy memory chip, what's the
best way of isolating it? I
have an oscilloscope, but nada skill in this sort of thing.
The typical way of testing RAMs in a C= service center was to rotate
a pair of them from bank to bank and see if the symptoms shift. It
was especially handy for detecting zero page problems.
Not sure how you'd use an oscilloscope to detect transient memory
errors not caused by design problems (like an undershoot problem).
A logic analyzer might help... you could set it up to trigger when
things don't happen when you expect, for example.
3) If, as my money is on, it's the BASIC ROM, can
it be replaced with an
EPROM - if so, there's a whole gamut of additional questions to
follow....
Yes, with the caveat that you might not find an EPROM with an exact
pinout to match the masked programmed ROMs you'd be removing. In the
case of _my_ 2001-N, it takes 2332 masked programmed ROMs, or 2532
EPROMs, *not* 2732 EPROMs. If your board is older, I do not know
what parts exist that would drop in place. You can make/buy ROM
pin swabbers to use any 2K or 4K EPROM in that socket.
There is also a fellow on the cbm-hackers list who has a board
that plugs into the CPU socket and provides a full boat of RAM
and ROM. Very handy for sweeping away problems with 8K static
PETs. Might even be cheaper than a box full of pin swabbers.
Now... I _have_ seen RAM and ROM chip fail in PETs. What is much
more likely, however, is to have those cheap-ass crappy sockets
oxidize. Reseating and/or removing and reinstalling all the socketed
parts is the first order of business. Check the pins of all the
ICs you remove. Some chips from that vintage have enough silver
in the plating to tarnish to black. Check the classiccmp archives
for the "best" way to remove it. Be sure not to bend any pins
while you are at it (easier than you might think - just fixed
another machine where a corner pin was folded under and looked OK
from 2 angles)
So... if you are attempting to do a minimal repair that also looks
much like the original, you might have to resort to pin swabbers
if it's a flakey ROM chip. If you just want it to work and aren't
hung up on "original" parts, you might want to contemplate a
complete ROM/RAM transplant. ISTR the boards are around $35USD,
but that's only a ballpark estimate.
Good luck,
-ethan
P.S. - if you don't know about http://www.funet.fi/pub/cbm/ it's a
great site with schematics and firmware.
Thanks in advance!
--
Cheers, Ade.
Be where it's at, B-Racing!
http://b-racing.com
__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com
10:53 p.m.
At 06:41 18/01/2003, Ethan Dicks wrote:
Personally, I'd rather keep this one as original as possible, i.e. keep the
original (buggy) ROMs. However, if it means keeping it working, then I'm
prepared to substitute the MOS ROMs with a board/set of boards containing
more modern EPROMs wired appropriately containing the original images. Not
that I've got a clue how to go about doing that, mind...
Well, there are no unpolulated sockets, so perhaps it's a fair bet that
there's some extra stuff in there. Goodness knows what, though, all the
ROMs are MOS 6540's.
Figures :( I've had to do this once already to locate a broken 2114 (when
I got the machine, it claimed only 1600 bytes of memory were free).
Luckily, I have one more spare 2114 (I got two from someone on this list, I
forget who [for which I apologise] and thank them (again) profusely, as if
it is a broken 2114, there's a chance the second spare might be essential).
Anyway, if it *is* a failing RAM chip, my guess is it's the one which sits
near the bottom of BASIC memory, since the first line number causes a
crash. Does anyone have a map showing the correlation between memory
addresses & specific chips?
If it's on the schematics, I'll be looking there next (so no need to answer
that question).
<snip the rest for brevity>
Ta for that, all good info!
Thus, the order of the day is:
1) Try to determine which 2114 chip might have gone bad, either by
sequenced swapping or by trying to be clever with the schematics.
2) If that doesn't fix it, re-seat all ROMs.
3) If that doesn't fix it, go to plan C - which doesn't exist yet...
Also, I know this has been asked & answered before, but what's the best way
of cleaning the board? It has more dust on it than I am properly
comfortable with, and besides it looks horrible.
I'm guessing that *ideally* it should be washed in distilled water with a
very soft brush, then left in a warm place to dry thoroughly. OTOH, maybe
alcohol would be a better idea? I have some disc head cleaning alcohol
which would suffice - recommendations please!
--- Adrian Vickers
<avickers(a)solutionengineers.com> wrote:
OK, I can but try....
According to the sticker on the back, this is a 2001-8BS. The motherboard
layout appears to be the same as the one pictured on www.zimmers.net on the
4k/8k layout. I've taken some photos of my motherboard, which can now be
found here:
http://helmies.org.uk/images/cbm2001/mobo_front.jpg
http://helmies.org.uk/images/cbm2001/mobo_back.jpg
http://helmies.org.uk/images/cbm2001/mobo_side.jpg
NOTE: Each picture is approx 449K = long download over a modem!
I dug the 2001 out of the cupboard the other
day...
Can you give more information about *which* 2001 you have? There
are at least two major motherboard revisions for the static-RAM PETs.
They use different ROMs and RAMs. 
Also, as I have one, there are also models of 2001 that
have _dynamic_
RAMs. My 2001-N has 32K worth of 4116 chips (from the factory) and a
full-sized graphic keyboard. If it were for sale in Europe, it would
most likely been labelled a 3032, for comparison, but here in the States,
it says "2001" on the front.
I bring it up because I _think_ you are talking about the original
chicklet-keyboard SRAM PET, but you didn't come right out and say so.
Apologies, I didn't; I keep forgetting that the 2001 went through more
development than perhaps any of the other PETs. Yes, it is a chicklet
keyboard version, with static RAM (2114's) and original 28-pin MOS ROMs.
Unfortunately, the PET seems to have developed an
odd fault: It won't
take a BASIC program, and some keywords seem to be knackered...
.
.
.
So.... I figure the BASIC ROM has become slightly
corrupted, OR I've got
a flakey memory chip...
1) Can the BASIC ROM be swapped with one from, say, a 3032 or 4016?
Most likely not. If your 2001 has 2114 SRAMs, perhaps. There are
upgrade ROMs (BASIC 2.0, anyway) _for_ the 2001 In fact, which
one IS the BASIC ROM?
BASIC lives in several ROMs, the Kernel lives in several ROMs. BASICs
prior to 4.0 occupied $C000-$DFFF, and the Kernel occupied $E000-$FFFF,
minus the PIAs and VIAs, etc., at $E800. BASIC 4.0 starts at $B000
to add room for diskette commands.
In my PET, $9000, $A000 and $B000 are filled with user-supplied firmware
(better machine-language monitor, BASIC extensions and a tape speeder).
These sorts of things, plus ROMs for word processors and other applications
were somewhat typical amongst serious PET users. Dunno if yours has
anything like that, but if you find ROMs where you expect none to be,
that's probably why. 2) If it's
a dodgy memory chip, what's the best way of isolating it? I
have an oscilloscope, but nada skill in this sort of thing.
The typical way of testing RAMs in a C= service center was to rotate
a pair of them from bank to bank and see if the symptoms shift. It
was especially handy for detecting zero page problems. Good luck,
Thanks - I'm going to need that (in place of specific skills :)
P.S. - if you don't know about
http://www.funet.fi/pub/cbm/ it's a
great site with schematics and firmware.
I already knew of it, but had forgotten how useful it might be. Will check
that out next. Ta!
--
Cheers, Ade.
Be where it's at, B-Racing!
http://b-racing.com
18 Jan
18 Jan
9:43 a.m.
A non-text attachment was scrubbed...
Name: not available
Type: text
Size: 389 bytes
Desc: not available
Url :
http://www.classiccmp.org/pipermail/cctalk/attachments/20030118/7b72efbd/at…
17 Jan
17 Jan
8:37 p.m.
At 18:18 17/01/03 +0000, Adrian Vickers wrote:
So.... I figure the BASIC ROM has become slightly
corrupted, OR I've got a
flakey memory chip which reads OK but doesn't write properly. The question
is, how to find out?
1) Can the BASIC ROM be swapped with one from, say, a 3032 or 4016? In
fact, which one IS the BASIC ROM?
3) If, as my money is on, it's the BASIC ROM, can it be replaced with an
EPROM - if so, there's a whole gamut of additional questions to follow....
Sorry, not knowing much about this machine in particular, can't suggest too
much, but...
2) If it's a dodgy memory chip, what's the best
way of isolating it? I have
an oscilloscope, but nada skill in this sort of thing.
... at least perhaps you can rule out the RAM this way. Assuming memory is
socketed, what I'd do is swap it between two banks. Just take all the ICs
in one bank, and swap them with the another. If it's a memory fault, your
sypmtoms will change (probably quite a lot). If it's not a memory fault,
chances are the symptoms will remain the same.
This method depends on a couple of assumptions (that there's only one
faulty memory IC, for instance), but generally it's pretty safe to make
those assumptions, and it's saved me lots of time in the past.
The thought of trying to track down a RAM fault using an oscilloscope
sounds pretty scary to me, but perhaps there's a way of doing it.
For what it's worth though, this sounds more like a faulty BASIC ROM then a
RAM problem, as you suggest.
Mike.
8013
days inactive
8014
days old
5 comments
5 participants
participants (5)
-
ard＠p850ug1.demon.co.uk -
avickers＠solutionengineers.com -
erd_6502＠yahoo.com -
mike＠ambientdesign.com -
vze2wsvr＠verizon.net