This is only a preview of the November 1989 issue of Silicon Chip. You can view 46 of the 112 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. Items relevant to "The Radfax Decoder":
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
|
You've got to be Sharp in this gaille
This month's main story has many claims to fame
- if 'fame' is the right word. It set something of a
record in time taken to obtain a spare part and
also a record for patience on the part of the
customer. And on the bench it made another
record bid; no less than four faults when only one
was suspected.
The story started when I was
called to the local school to do a
routine repair job on the PA system.
It was the commencement of a term
and I encountered a new principal
who had just been transferred from
a school in the far south of the
state. The full significance of this
last point was not appreciated until
much later.
After I fixed the PA system, we
began chatting and he soon raised
the matter of VCRs. Did I service
these devices? When I said I did he
explained that he had a VCR which
was faulty and that it had had a
checkered history of service.
In fact, he had some very nasty
things to say about the serviceman
concerned which I would not dare
repeat in print. But briefly, it appeared that it had been serviced on
two or three occasions for various
faults and after the last service, the
present fault had appeared.
So could I fix it? That was
something of a leading question but
sticking my neck out, I said yes assuming it was capable of being
fixed economically. And feeling
that the industry owed him something, I emphasised that if the
repair appeared not to be economical, there would be no charge for
the advice.
He accepted my offer and added
that he didn't care how long it took
to fix the thing - just so long as it
was genuinely fixed in the end. I
doubt if either of us guessed how
long it would really take but he was
48
SILICON CHIP
as good as his word; not once did he
pressure me.
Naturally, I needed to know the
nature of the complaint and as
much of its history as possible. He
was a bit vague about this. Basically, it was a failure to reproduce colour but more importantly, it was intermittent. I tried to determine
whether this was a recording fault
- was it only on tapes he recorded
or also on pre-recorded tapes? but it seemed that he had not noted
this distinction.
Anyway, he duly delivered the
machine to the shop. It turned out to
be a Sharp VC-483X, a model of
about four years old and one for
which I had a manual. At the first
opportunity I set it up, pushed in a
tape and made a recording. There
was no colour on playback and
none when the tape was played in
another machine. On the other
hand, the machine would play a
pre-recorded tape in full colour.
OK, so we had a fault somewhere in
the chroma circuitry.
My next step was to open the
machine and check for anything obvious. My first stop was the Y/C
(luminance/chrominance) board
which sits above the drum and
transport mechanism. It is mounted
component side down and is
covered with a clear plastic sheet.
Bodgie pot
And did I find something "obvious"? One glance was enough;
someone, presumably the previous
N\"( FIRST s-rz::>p Wf=I\S
11-tE Y/C (LUM\NANc.£.
CHROMINA~C£)SOAR'D
WI-'\~ Sl"t'S A'SOV~ ~
1'RUIV\ic-~,
~Ar-J\~N'\, MOUl\r\""ED
COIY\f>~ S\De; t>OWN
&GO~Wt~A
a...€A~ 1)f.A~C 'SH"1'"
serviceman, had attacked the copper side of the board and made an
unholy mess of it. There was also a
small trimpot which had been added and which obviously didn't
belong there.
Backtracking through the circuit
and wiring patterns provided some
clarification. The bodgie pot was a
replacement for R508, a lkn pot used to set the choma record level. In
its original form, it is a much
smaller unit (about 5mm in diameter) and is normally mounted on
the component side of the board.
But it wasn't just the bodgie pot
that shocked me. It was obvious
that a number of other components
in this part of the circuit had been
removed and replaced with new
components, and equally obvious
that it had been done by someone
who had never come to terms with
proper soldering techniques in
modern equipment.
In short, it was a mess, with
great blobs of solder around the
various joints, plus a fair share of
surplus blobs and slivers scattered
around the board. The wonder was
that the consequences had not been
more spectacular.
As to the reason behind this
mess, one can only speculate. My
guess is that it had started with the
loss-of-colour fault, possibly intermittent, and that the serviceman
had changed or tested all these
components, including the pot, in an
effort to track it down. That's fair
enough as far as it goes. My real
gripe concerns the shockingly poor
manner in which it was done.
Cleaning up
The first thing to do was to get
stuck into the board and clean it up.
This I did and even if I do say it
myself, I was able to get it back
very close to its original condition.
The only sour note was the bodgie
pot for which I had no replacement
and which had to be ordered. But at
least I had eliminated any potential
shorts and dry joints.
Then I tried the machine again.
And lo and behold, it worked; a
perfect picture in full colour. Of
course I didn't trust it; it was known
to be intermittent and while I
wanted to believe that I might have
corrected something in the cleanup, that seemed like a long shot. So I
simply kept on playing it at odd intervals while I waited for the
replacement pot.
I rang the owner and explained
the situation. He wasn't at all con-
cerned and in fact, I gained the impression that he was so cheesed off
with the machine that he had mentally written it off. I felt I would
have only had to suggest that he
scrap it and he would have done so.
Of course, I had no intention of
making any such suggestion.
The lack of the replacement pot
meant the machine was now more
or less pushed to one side. I tried it
on odd occasions and it always
worked but after a few weeks,
pressure of other work prevailed
and I abandoned even this ritual. I
did keep needling the supplier
about the pot though, but always
without result.
This went on for nearly six months - just as well the customer
wasn't in hurry. Then one day a
package of other spares arrived
and there among them was the pot.
"Aha!", I thought. "Now to get this
one off the shelf".
Normal Rate -
I opened the machine again, fitted the new pot, tidied things up
where the old one had been, and
tried running a tape. No joy. The
cassette carrier accepted the tape
and put it on the deck but after that
nothing happened. Fault number
two.
It wasn't a hard one to find. The
idler wheel which drives the supply
and takeup reels had packed it in. It
was probably about due for
replacement anyway and six months sitting unused on the shelf was
the last straw. Anyway, it wouldn't
drive the takeup reel and the protective circuit shut the system
down.
That didn't seem to be any pro-
ly, if it runs at all. Then C813, a
1 O!,tF 35V electro, bypasses the
same Vee rail and if it is open or
low value, the result is weaving
verticals.
TETIA TV Tip is supplied by the
Tasmanian branch of The Electronic Technicians' Institute of
Australia. Contact Jim Lawler, 16
Adina St, Geilston Bay, Tasmania
7015.
$35 per hour.
No charge for kits that can't be
repaired.
3 month Warranty on repairs .
Construction - fixed or hourly
1 2 Month Warranty on
Manufactured Kits.
Custom Designing,
Manufacturing, Large or Small
Quantitites.
HYCAL
INSTRUMENTS
Design, Manufacture, Repair of
Electronic Equipment.
(02) 633 5897
Unit 4,
62 Great Western Highway,
Parramatta, NSW 2150.
Idle idler
TETIA TV TIP
National TC2202 (M8 chassis)
Symptom: Set reluctant to turn on.
Sometimes starts with a squeak,
sometimes only squeaks. When it
does get started it shows severe
hum bars or horizontal wriggle .
Cure: There are actually two different faults, closely associated
with each other. R804, a 4. 7k0
SW resistor, feeds Vee to the
horizontal oscillator. If this is dry
jointed the oscillator runs erratical-
FIX-A-KIT
Kit Repairs - $15 per hour.
Trading hours:
8am to 3pm Monday to Friday.
D.DAUNER
ELECTRONIC
COMPONENTS
WE STOCK A WIDE RANGE
OF ELEC TRONIC PARTS
•
for
Development • Repair
• Radio Amateur
• Industrial Electronic
• Analog and Digital
WH ILE STOCKS LAST
NEOSID HELIX FILTER
for UHF
now available
COAXIAL RELAY 28VDC $32.00
TRANSMITTER VALVE 5786 $18.00
Come and see.
Showroom:
51 Georges Crescent,
Georges Hall, NSW 2198
(Behind Caltex Service Station In Blrdwood Road)
Phone 724 6982
TRADING HOURS:
Monday to Friday 9.00 a.m. to 4.00 p.m.
Saturday from 9.00 a.m. to 12.00 noon.
N OVEMBER 1989
49
SERVICEMAN'S LOG -CTD
blem. I had a spare idler on hand
and it was a relatively simple job to
fit it. Then I made another attempt
to run a tape, only to realise that
the machine was now completely
dead - there wasn't even a clock
function and when there is no clock
it is something pretty fundamental.
This was fault number three and
it was a beauty. I went immediately
to the mains fuse, F901, which is
about as fundamental as you can
get. But it was intact and that's
when the fun started; a real chaseme-Charley effort.
There is no way that I could
describe the whole operation in the
space available; nor could the
reader follow it without a manual.
Instead, I will content myself with a
broad picture, with a few component ident codes tossed in for the
benefit of my colleagues who do
have a manual.
IN ~HO'ln",
The power supply (PWB-P)
generates several voltages - including two 9V and two 13V supplies - which appear at plug PA. I
checked all these voltages, hoping
to find at least one that had failed.
But no luck; they were all as
prescribed which was a bit of a setback.
One of these supplies - the 13V
rail on pin PA4 - goes to board
PWB-1 (IF Tuning Circuit) and appears at plug IEl. I followed this
because the 13V supplies a package
(RUNTKO157GE) on this board
which is a power supply circuit in
its own right. It delivers several
voltages, some positive, . some
negative and some AC.
Again I drew a blank; all voltages
were as marked. At this point
pressure of more urgent jobs forced
me to put the machine back of the
shelf. But I did contact the owner
rr w~~
F-\ N\E.sS,
wrn-\ G~EA, '8LO~~
OF SOLDE.R":'; ..
50
SILICON CHIP
and bring him up to date. He waved
off my explanation: "No worries,
mate - when you're ready". Patience indeed.
It was several weeks before I
was able to tackle the machine
again. Still convinced that it was a
voltage supply problem somewhere,
I set about tracing and checking
each supply rail in turn to its
ultimate destination.
Blind alleys
I will spare the reader the agony
of the blind alleys I followed; suffice it to say that there were
several. But I eventually began
tracing the 9V rail which commences at pin 9 of the power supply
plug PA. This is shown as going to
EB4. (Where the heck is EB4 ?)
Well it turned out to be on board
PWB-E, the mechanical control circuit, portrayed on a 3-section
foldout sheet. The 9V rail comes into it on plug EB (pin 4) and rather interestingly, had now increased to
9.2V (good news for the perpetual
motion inventors).
According to the circuit, this rail
goes straight from the plug to a
device marked 'PR' but with no indication as to what this is. From
there it runs all over the place,
feeding a variety of components,
and also exits from the board via at
least five other plugs. So it is obviously a very vital rail.
Having established this much on
the circuit, I began checking this
voltage on the board. Fortunately, I
was able to pinpoint a couple of
places where it should have been,
only to find that it was missing.
Well that was something.
Now I was looking for the device
'PR' which, while shown hard
against the plug on the circuit, was
nowhere near it in real life. I finished up back-tracking along the copper pattern until I finally found it.
It turned out to be a safety
resistor, with voltage on one side
and none on the other. So that was
it - except that there was nothing
on the device, the circuit, or in the
parts list to indicate its value. Fortunately, most safety resistors fall
into a fairly narrow category between 0.50 and 20 and with a
nominal 0.5W rating. So, as a temporary measure, I fished out a
similar device for another machine,
fitted it, and we were back in
business.
Which was a considerable relief
after all that effort. But why had
the resistor failed? It had obviously
been overloaded, rather than failed
spontaneously, yet there was no
sign of distress in the replacement.
Naturally I was concerned that
there was another intermittent
somewhere but there wasn't much I
could do about it, except give it a
good workout. (In fact there was a
good reason for the failure, which I
will discuss later).
So for the moment, I simply set it
up in a corner of the bench, connected to a monitor, and ran it at
every opportunity. The first day I
ran a 3-hour pre-recorded tape
through it. No problems. The second day I recorded a 3-hour tape,
then played it back. No problems.
The third day I played another
pre-recorded tape. It ran without
any problems for about two and a
half hours. Then - a noise bar. Oh
no! - fault number four.
Sweaty brow
As I wiped my brow in exasperation, I recognised the first clue; it
was a stinking hot day and my
workshop doesn't have air conditioning. I switched the machine off,
let it stand for about half an hour,
then ran it again. The fault had
cleared, confirming my suspicion
that it was a thermal problem.
I rang the owner and asked him if
he had ever experienced anything
like this fault. He was quite emphatic that he hadn't and it Wil,S
then that I realised the significance
of his previous location; well south
and somewhat elevated. The
chances of him encountering the
same thermal conditions as I had
would be slim.
In addition to the thermal clue,
the nature of the display provided
another clue. The noise bar was
drifting slowly down the screen
rather than remaining stationary as
is more usually the case. This suggested that it was due to a loss of
lock in either the drum motor or the
capstan motor. The question was,
which one? It was a 50-50 chance
of course but I had a feeling in my
bones that it was the drum motor.
,oo
11-1~ IVlACHIJ\le;:. W~'S N'OR6.
OR t-6S1S ~SHE.'D A-S\t:>6-•••
My first move was to connect the
CRO to the appropriate test points
- TP5 for the drum and TP3 for the
capstan - and check the waveforms against those given in the
manual for adjustment of the tab
pots (R724 and R777) for correct
operation.
These waveforms came up virtually spot on - or at least as far
as I could tell by comparing them
with the idealised drawings in the
manual, a point of some importance
as it turned out. Of course, the
machine had now cooled and so the
test was not really conclusive.
I had to leave things there for the
day and the following day was a lot
cooler, meaning that I had some
trouble re-creating the fault. I considered applying artificial heat but
rejected the idea. Instead, I wanted
the fault to occur as a result of the
machine's own internal heat
distribution, at least for my initial
investigations.
Incidently, bulk heating can
sometimes produce deceptive
results due to innocent components
being heated to a level far above
that which they would experience
under the worst real-life conditions.
It's a point to watch.
Anyway, the fault eventually appeared without any help and while
waiting for this to happen, I had
delved a little deeper into the circuit. The control signals for the
drum motor come from IC708, on
board PWB-A (Servo Still). The
phase control signal comes out on
pin 1, passes through transistor
Q705, and is then applied to the
drum motor. The frequency control
signal comes out on pin 8, is processed by trani;;istors Q703 and
Q702, and also goes to the drum
motor.
Since I had gleaned little from the
CRO patterns at the te~t points, it
occurred to me that it might be worthwhile checking these two circuits.
I chose the collector of Q702, which
goes directly to the drum motor.
The only snag was that there are no
waveforms in the manual covering
these circuits so that I had no way
of knowing whether what I was
observing, under no-fault conditions, was within tolerance or not.
The only obvious clue was that it
changed when the fault occurred.
Then it occurred to me that while
I had the fault in my sights, it might
be worthwhile trying to find the
heat sensitive component with the
NOVEMBER 1989
51
071 ◄
0X01 ◄ 2a
R71 t
C701
'"'"'
+
i.z~ ..-Ei-c-+---N,-,
R710
1,0v1
0702
'""
·•ro
2SC~~
croo
.o,
0703
2SA733
Fig.1: the drum control circuitry for the Sharp VC-483X VCR. Transistor Q702 appeared to be excessively heat
sensitive. But was it?
aid of some freezer spray. I started
with Q705. No response. Q703
reacted similarly. But when I hit
Q702 (2SC945) the reaction was immediate and dramatic; the fault
vanished instantly. Further confirmation was provided by applying
some heat to Q702 with the soldering iron, which quickly re-established the fault.
I quickly fitted a replacement for
Q702, then gave the machine
another series of long runs. All
seemed well at first but the second
day was another hot one and after
many hours operation, there was
the fault again. And again, a
smidgin of freezer was all that was
needed to cure it.
So what could I do? Run a plastic
pipeline from inside the machine
and give the owner a can of freezer,
with instructions on how to use it?
This and similar wild Heath Robinson ideas went through my mind,
even while I was trying to figure out
a genuine cure.
I went back to the CRO check
point [TP5) and had another look at
the pattern. I still felt that it was as
near as one could wish to the drawing in the manual but decided to try
adjusting pot R724. With the
machine still in fault condition, I
turned the pot a few degrees one
way with no result, then a few
degrees the other way. And bingo,
everything came good although it's
52
SILICON CHIP
worth noting that the effect on the
CRO pattern was virtually negligible.
Well, that really did cure it. I ran
the machine over many days, for
hours at a time, and did everything
I could to promote the fault. In all, it
must have chalked up about 30
hours of use and it never flickered
once.
So what was the true nature of
the fault? Was the first 2SC945
really faulty? Or was the original
setting of R724 marginally off;
perhaps just enough to cause trouble when the drift of some other
component values became significant? Or had the previous serviceman had a fiddle with that as
well as butchering the board?
I'm inclined to suspect the latter,
although there is little doubt in my
mind that the first transistor was
more heat sensitive than the
second.
Pertinent comments
Finally, I feel that I should add a
couple of pertinent comments.
First, several checks with the
owner have confirmed the cure; it
hasn't missed a beat since and he is
delighted. I have a permanent
customer there.
Second, why did the safety
resistor fail? Shortly after I wrote
the main part of this story I had
another VC-483X in for service. It
turned out to be a faulty idler
wheel, similar to the one described
above. That was simple enough and
a replacement was quickly fitted.
The only snag was that the machine
was now completely dead; not even
a clock readout.
Of course the penny dropped and
I went straight to the safety
resistor. Sure enough, it had failed.
Why? It was my own fault really.
In both cases I had neglected to
turn the machine off before working on it and, in removing the
cassette carrier to get at the main
deck, it is very easy to short one of
the supply rails on the small board
on the left hand side of the cassette
carrier. And oops! - one safety
resistor destroyed.
That's one I learned the hard
way.
And what of the intermittent loss
of colour which started all this?
That was almost forgotten in the
hassle of tracking down the other
faults. And for a very good reason;
it never appeared again and that includes my monitoring right up until
writing this last part of these notes.
From this I can only conclude
that it was a faulty joint, either present from the start or created by
the previous serviceman and which
was corrected when I cleaned up
the board. And that in itself is a
rare enough event to be worth
noting.
~
|