This is only a preview of the April 1993 issue of Silicon Chip. You can view 48 of the 96 pages in the full issue, including the advertisments. For full access, purchase the issue for $10.00 or subscribe for access to the latest issues. Articles in this series:
Items relevant to "Build An Audio Power Meter":
Items relevant to "Three-Function Home Weather Station":
Articles in this series:
Items relevant to "12VDC To 70VDC Step-Up Voltage Converter":
Articles in this series:
Items relevant to "A Digital Clock With Battery Back-Up":
|
COMPUTER BITS
BY JOE ELKHORNE
Upgrading to a 386 - now I
know what a "kludge" is!
Is your computer due for upgrading but you
have not known where to start? The process is
fraught with pitfalls as Joe Elkhorne found to
his frustration & cost.
For the past couple of years, I've
been perfectly happy running a modest XT compatible, an Amstrad PC20. It is perfectly adequate for simple
word processing and you can't get
much simpler than PC-Write 2.41.
Recently, I found it necessary to
provide a state-of-the-art WP document to a local book publisher. With-
self-contained unit in a low-profile,
plastic case.
So I had the heart of a super new
system and an uphill battle ahead of
me. The first problem was not difficult to solve. I'd let another friend
have an oscilloscope I wasn't using,
again on a "pay me when you can
basis". When he subsequently ac-
"Right at this point, I ran into the first pitfall.
Having worked with dodgy Taiwanese clones
in the past, it didn't even occur to me that this
board might be labelled."
electronics shop in a nearby suburb.
"Hi there - I need the ribbon cable
that goes from a controller card to the
floppy drive". The salesperson shakes
her head in disgust. "They come with
the controller cards". "Ifl had one, I'd
not be asking for one. The thing is,
I've got a secondhand card that wasn't
provided with cables".
"Oh yeah ... "
Silence. Finally, I explicate the obvious, "So I need one".
"We don't sell them". Scratching
my head in perplexity, I inquire if
they know who might. "You can make
one yourself". "I'm game, but I need
the bits. I've already got a hammer".
"We sell the bits". Good, at last we're
getting somewhere.
At home with a bag of bits, I carefully inspect, align and assemble: ribbon cable, one, with IDC connectors,
for connecting floppy drive. I do believe this technology was designed by
Mr Murphy but more on this subject
later.
Getting it together
out a platform to work on, I solicited
the help of a friend with a splendid
386 system. He'd been through the
learning curve with earlier versions
of Windows and Word for Windows.
This was an opportunity for him to
show off his equipment and I was
very impressed with the power of the
system.
When Graeme upgraded to a 486
shortly after that, he made me an offer
I couldn't refuse: accept his 386
motherboard and pay him when I was
abk Bonzer! There's just one trouble
with upgrading the PC-20. It can't really be done. The physical architecture
is not like a standard clone - it's a
64
SILICON CHIP
quired a case, power supply, a CGA
card and a keyboard from another person, I agreed to take these bits and
pieces in lieu of cash. I think that the
cashless society is already here!
So, let's see where we stand. I have
a 386 DX/25 motherboard with 8Mb
of RAM, a "Baby" tower case with a
200W power supply which almost fits,
a 360Kb 5¼-inch floppy drive, a 101key keyboard, multi-I/O card, CGA
card and a PC-20 CGA monitor. Well,
that's not a bad start - all the big
pieces. But like any enterprise of this
nature, it's always the fiddly bits that
are a hassle.
Fortunately, there was a discount
Now my mate Rob had not only
brought one case and power supply,
which he was kind enough to fit, but
had also mounted an old XT board in
it expressly for the purpose of holding the plug-in cards in place. Right at
this point, I ran into the first pitfall.
Having worked with dodgy Taiwanese clones in the past, it didn't even
occur to me that this board might be
labelled.
I looked at the mass of front panel
wiring and pulled all the little connectors off without a second thought.
Getting the XT board out involved
compressing the nylon support posts
- and holding your jaw right. Only
after the board was out, in the clear
light of day, did I notice that all the
appropriate points were clearly stencilled. Twenty-twenty hindsight is a
wondrous thing.
Fitting the 386 motherboard into
the case was not difficult- a couple of
the nylon support posts had to be
moved. But there weren't quite enough
of them and some plastic from the
junkbox provided additional support
at a non-critical point.
Nor was the power connection a
problem. There's only a couple of possibilities because they're keyed and
the recommended end-to-end position had already been marked with a
felt pen. At this point, I truly regretted that the documentation for the
386 motherboard hadn't turned up.
Oh, well, who needs a power LED
anyway? None of the other front panel
functions are critical to a fundamental test. I then planked the floppy
drive on a magazine on top of the
floppy mounting bracket for ease of
testing. The multi-I/0 card was
plugged into a convenient slot and
my homemade ribbon cable joined
up . The CGA card went alongside and
the Amstrad monitor was set beside
this "test jig", on the kitchen table. I'd
already made a 5¼-inch boot disk. All
that remained was throwing the power
switch - and praying; not necessarily
in that order!
The smoke test
Rob and his friend Liz turned up
just at that point. If they expected a
mushroom cloud and some gratuitous
language, they were disappointed.
=ftfflll/JIIJ=
=U!l!l!-
There was movement at the station.
It's true the system didn't boot up
initially but there did seem to be the
proper start-up sequence.
"Let's make sure the CMOS setup
knows about the A: drive", said Rob.
We did·this and a keyboard Ctrl-AltDel proved the system was responding properly. Unfortunately, though
the floppy drive's LED came on, nothing else happened.
We inspected the drive to make sure
nothing had been whacked out of
As well, I'd done a diagnostic/configuration dump, so I knew the port
addresses (serial, printer) and interrupt vectors which had been successfully working.
Again we fired up the unit and
forced the system into the CMOS setup procedure. Golly, I wonder which
of the 47 types I should call this beast?
We exited the CMOS setup to have a
think about things. Despite the obvious report that no C or D hard drive
existed, the system booted up prop-
"Rob & his friend Liz turned up at that point. If
they expected a mushroom cloud and some
gratuitous language, they were disappointed.
There was movement at the station."
shape. Everything looked all right but
we wondered about the mysterious
little jumpers. Just maybe, we decided,
this drive was old and needed to be
told, some way or another, that it was
in an AT-style machine. Meanwhile,
it was very frustrating being this close
without actual operation. We decided
to be daring and put the 40Mb
hardcard into the system.
I'd backed up all the files that morning, using the ARJ utility, to the existing 720Kb drive of the PC-20. Regardless of"finger trouble" - short of something really catastrophic like dropping the hardcard - I should be able
to resume operation totally on the old
machine.
erly! Obviously, the system bus recognises the existence of the hardcard
controller and acts appropriately.
Well, this was exciting. At that point,
we added the serial port cable and
modem and called our favourite bulletin board.
That effort was successful and we
called it a night. No sense in pushing
our luck, we decided. The following
morning, just to be thorough, I proved
that the parallel port worked. Now all
I needed was to get a floppy working.
Because of my BBS activities and the
small hard drive setup, I do a lot of
archiving on floppy disks. I proceeded
to do more kitchen table tests - unsuccessfully.
Position Vacant - Technical Writer
SILICON CHIP has a position for an enthusiastic technical writer. This person will already be a frequent reader of this
and other technical magazines and will probably have a number of hobby interests. Technical qualifications are not
mandatory although a good background in electronics is necessary. Good writing skills are essential for this position
and some familiarity with computers would be an advantage.
Do you think you could do the job? You will mainly be involved in the writing and preparation of articles for this
~agazine. For example, you could be required to write the constructional article on the ~lectric fence project in this
issue or the short feature article introducing the Digital Cassette Deck, after attendance at the recent Australian
release.
You will also have the opportunity to build simple electronic projects under supervision and you will be involved in
proof reading and other activities involved in the day to day production of the magazine.
The salary package is negotiable.
If you think you could actively contribute to the improvement and growth of SILICON CHIP, send your application in
writing to The Publisher, PO Box 139, Collaroy Plateau , NSW 2097. Applications close 21 st April, 1993.
APR,L
1993
65
COMPUTER BITS - Upgrading to a 386
I tried the technical support telephone number for the drive manufacturer, only to learn that their local
office was closed. It's amazing how
many operations here in Melbourne
have shut, with everything allegedly
handled from Sydney. And not all of
the companies have 008 numbers,
unfortunately.
supports 360Kb drives. This snippet
of information was verified in the discount dealer's catalog. Off to the local
bloke again.
"Hi, it's me again. I need a high
density drive, a three-and-a-half,
please". This time, there's a bloke at
the counter. He rummages around
looking for said unit while I go over to
the display and find the catalog
An old adage
number for the cheap controller. HavThen I recalled the old adage, "when ing perused the catalog thoroughly
all else fails, read the book" and re- over morning coffee, I know there are
membered that I'd seen a book on PC several to choose from. All I need, I've
architecture and hardwar~ technical told myself, is one floppy and this
details. I made a special trip to the controller is linkable for two drives of
city. In fact, I found several compre- the same type, from 360Kb to 1.44Mb.
"That's not the right controller",
hensive reference books to choose
from. Selecting one, I flipped to the , says the clerk. "I know that there's no
table of contents and went straight to
doco with it. I can see through the
the chapter on floppy drives.
plastic wrap. But see these links? And
Right. One of the lurks is the old note the description in your catalog?
cable twist - oops! - and the link for Now, can I be wrong? I've been wrong
DR1 on the floppy drive itself. Fair before".
He decides that maybe I'm not
enough, my mistake: the straightthrough connection as I'd made the wrong and rings up the sale. I decide
cable is actually the B: drive position. I'm on to a good thing because there's
Hey, suppose all this fiddling has two drive cables in the packet. Wow!
"The young woman takes the drive straight to
one of the demo machines, kills the power,
pops the top, pulls the cables and swaps it into
the system It fires up, and works beautifully."
done something to the drive itself?
Ah, substitution test. No, I don't have
a second floppy but I can put it back
on the PC-20. Yes, it works. Well, the
old process of elimination is one of
the tried-and-true ways of troubleshooting. Now, at least, I have one
less variable to worry about.
Funny thing though, when I put it
back in the 386, not even the LED
would come on. Fiddle, wiggle, oops,
the motor started. Could it be? Now
I'm suspecting my cable but I don't
have a spare.
At this point, I'm telling myself that
a 360Kb drive is old-hat anyway! I
quickly justify the need for a 1.44Mb
drive. Bewdy! Oh, and a controller of
course, since the multi-1/O card only
66
SILICON CHIP
And I've got the mounting adaptor
bracket kit so I can install this beauty
permanently.
Racing home, I pessimistically decide to test it externally before I bolt it
in place. I swap controllers, checking
the stencilling on the new and deciding that yes, it does say it's set for
1.44Mb. I swap controllers, attach one
of the new cables to the high-density
drive and turn it on. It doesn't go.
Oh, yes - I've got a formatted
1.44Mb disc in place. Well, at least
the packet said they were. The LED
was on, though. Unfortunately, it's
always on. I wonder what will happen on the PC-20? I look at its manual.
Yes ... DRIVEPARM should work.
It's kitchen table time again. I'm
really wondering if I could borrow a
second monitor from somewhere; I'm
tired of carrying the one-and-only back
and forth. Never mind. Hey, power up
and look at that: the system recognises there's a controller card in the
slot and even knows what size it handles. But, oh dear. Why doesn't the
drive work? Oh dear, what's happened
now? Why is my boot disk in the A:
drive corrupt?
I go through every permutation
that's possible. The mounting adaptor kit came with a 3½ to 5¼ ribbon
cable adaptor and a power adaptor to
match the usual power supply connector. By flicking a switch in the PC20, I can tell the system that the outboard drive is the A: drive. The power
adaptor lets me try the edge connector on the cable, rather than the dualrow miniature connector which would
be more commonly used.
Aha! As I'm about to put cable to
connector, something rather suspicious catches my eye. I look closely
and discover, deep inside the 5¼ edge
connector, a smashed contact shorting
across from one side to the other. Why
hadn't I seen this before? Because I
was using the smaller and more appropriate connector.
Score: cables 2; Joseph 0.
I throw the thing aside with several
rude words and find the other new
cable. Double checking the "repair
and upgrade" book, I confirm my understanding of the A position and B
position. Now I go back to the 386
and try it. It still doesn't work!
As soon as I walk into the discount
electronics store, I get the impression
that I've worn out my welcome. Even
the laying on of plastic money might
not help. I state the problem: can't get
the little drive to work in two machines!
The young woman is knowledgeable - she takes the drive straight to
one of the demo machines, kills the
power, pops the top, pulls the cables,
and swaps it into the system. It fires
up and works beautifully.
Fair enough. Whatever is wrong is
my problem, not theirs. Having seen
the little drive work, I believe I can
trust it. Unfortunately, back home, it
still doesn't want to respond. All I'm
getting is "Not Ready Error Reading
Drive A (Abort, Retry, Fail)".
Can it perhaps be the controller?
Hey, if I'd bought the next model up,
I'd be able to mix and match, put the
360Kb onto it to prove the system to
that point, without any hassle. Why
don't we take another drive today?
They're definitely not overjoyed to
see me again so soon but are more
than happy to swap/upgrade. A little
more money for them and maybe it
will get me out of their hair.
This new board is switch-selectable
for four drives, independently. How
can I go wrong? I start with a known the 360Kb drive - and prove that the
controller and the system are happy
to work together. I swap over to the
1.44Mb drive but no go. How weird
can you get?
Yet another foul-up
Nonetheless, I'm happier with this
co·n troller since there is some documentation. In fact, just as a second
test, I even drop it into the expansion
slot on the PC-20 and the system boots
up and recognises the various switch
configurations. Incidentally, in the
process of all this drive testing, I've
been experimenting with other facets
of operation. I discover "Yet Another
Foul Up" - the serial port sometimes
locks up when using Telix.
I spend an hour or so with a breakout box, trying to figure out what and
why. Finally, I disable the XON/XOF
(software) flow control, dropping back
to DTR (hardware) control, and that
seems to help. The original RS-232
port was a true 25-pin connection.
The flying lead from the card to the
backplane connector is skimpier. This
needs further investigation at a later
stage.
Meanwhile, it's back to the drive
anomalies. I'm starting to feel paranoid and begin with what should be a
known working situation: the 360Kb
as B-drive on multi I/O. Yes, that's
OK. Now I fit the 4-way controller, set
the DIP switches appropriately, reboot,
and the system CMOS recognises the
existence of the 360Kb drive without
intervention. I do DIR B: and that's
fine. Everything working.
Though this might seem laborious,
eliminating variables and starting with
a known fully working situation at
the immediate moment is a good, if
pessimistic, plan.
I add A: (1.44) without power, of
course, and re-start the system. B: now
does not work- Not Ready Error Reading Drive (Abort, Retry, Fail). Hmm,
let's disconnect the ribbon cable from
B and try A only, as A. Reset, CMOS
COMING NEXT MONTH
Remote Volume Control
This project will let you adjust the volume & balance of your hifi stereo
system from the comfort of your lounge chair. It's based on a dedicated
microprocessor chip to simplify the circuitry and to make construction as
easy as possible. It won't degrade the sound quality either, thanks to
excellent specifications. Full details in the June 1993 issue.
Woofer Stopper (or Dog De-Barker)
Do you have trouble with barking dogs? If so, this project could be the
answer to your prayers. It produces a retaliatory high-energy supersonic
tone which discomforts the dogs and discourages further barking. After a
few weeks of use, the dogs will be miraculously quiet.
check, OK, reboot, Not Ready Error
Reading Drive A.
Well, all I know for certain is that
the 360Kb drive on its own works.
Can there be something incompatible
about the 1.44Mb drive and both of
my computers? Later, a call to Amstrad's technical support section elucidates the information that the PC20 will not support a high density
drive, even with an external controller. "Why is the information in the
manual so ambiguous?" "It's there for
completeness only ... " "Groan".
This was the day I went back to the
PC92 show for the second time. I
bought Windows 3 .1 for $95 (on 1. 2Mb
which is all they had) and leaving
there, went to "the local blokes" and
bought a 1.2Mb drive.
I raced home with all the goodies.
Using the new 1.2Mb drive only and
the 4-way controller set as A=l.2 and
B=360, at last I had a working highdensity drive! Finally, I reset the controller for B=l.44, cobbled the thing
into place and was rewarded with a
proper start-up sequence. The A: drive
was still perfectly happy. Asking the
B: drive anything, however, still gave
me a DOS error - but the LED was
coming on and going off.
What the heck: Format B: /s. Oh,
wow! It's doing something. Presumably, somewhere along the line, the
actual disk I'd been using was corrupted. I spend the afternoon and
evening doing real computer things
with two working high-density drives.
Joy!
The following morning, I turn the
computer on -and the B: drive refuses
to work! The LED comes on and goes
off, but again I'm getting DOS error
messages. The final lurk turns out to
be a dicky power adaptor for the
1.44Mb drive and the forceful application of a pair of visegrip pliers ensures that the intermittent connection will never happen again.
Everyt,hing now gets bolted back
into place and re-tested. Two months
later, the fundamental hardware is still
reliable - at least at the DOS level.
I've now embarked on a learning curve
for Windows 3.1, DOS 5.0, Word for
Windows 2 and the mysteries and
frustrations of printer drivers or the
lack thereof. But that's a story for next
month's issue.
SC
APRIL
1993
67
|