This is only a preview of the November 1994 issue of Silicon Chip. You can view 29 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. Items relevant to "A Novel Alphanumeric Clock":
Items relevant to "80-Metre DSB Amateur Transmitter":
Items relevant to "Twin-Cell Nicad Discharger":
Articles in this series:
Articles in this series:
|
SERVICEMAN'S LOG
Tread carefully with a new brand name
How does one cope with a set of unknown
brand? Is it an orphan brought in by an overseas
traveller, or is there a local importer? If so, what
backup service in the way of technical data &
spare parts is available?
In an extreme case, where the set
is an orphan or nothing can be found
out about it, there is often little option
but to bow out right at the start. It may
not do much for one’s reputation in
the short term but the alternative is
to risk much greater damage. If lack of
data and replacement parts means that
32 Silicon Chip
the job ultimately has to be abandoned
anyway, the customer is no better off
and the serviceman in down the drain
for his time.
Of course, all this is leading up to
the fact that some time ago I encountered a set carrying a brand name I had
never heard of and had to go through
the above mental gyrations in order
to decide how to handle the situation.
While that is now history, what followed is technically interesting and I
thought it worthwhile to point out the
need to investigate the background in
all such cases.
Unknown to me at the time, the
story really started a couple of years
earlier when a local motel changed
hands. The new owner had come from
Victoria – a point of some importance
as it turned out – and one of the first
decisions he made when taking over
was to replace all the TV sets.
It was a logical decision. From
what I knew of the original setup,
the sets were approaching the end of
their commercial life anyway (life in
a motel can be pretty rugged at times)
and were all VHF-only models. With
a significant number of guests now
wanting to watch SBS, the lack of UHF
was a serious shortcoming.
So, after carefully studying what
was available in his state of origin,
he settled for 12 Contec 51cm colour
TV sets, model MSVR-5383. And this
was what I was presented with when
he approached me sometime later to
undertake the service of these sets.
From a business point of view, of
course, it was an attractive offer. But,
initially, I hesitated to be become involved. I had never heard of the Contec brand and needed to be reassured
along the lines already discussed,
before committing myself.
As it transpired, the owner had
done his homework pretty well. The
sets had been purchased from a Melbourne firm, Freecor International
and, yes, he had investigated the
service and backup situation and was
able to give me the name and phone
number of the manager of their service
organisation.
Which wasn’t a bad effort. But he’d
gone one further; he had secured some
circuits. And as circuits go these days,
they are quite good for the most part.
The only snag is that, in the original,
-31V
5V
4
F
3
F
12V
1
F
2
F
1
E
2
E
8
1
D510 C514
47
7
6
240V
D511
3
4
2
5
R519
1k
IC
510
IC502
330
C515
470
0.1
Q506
T501
Q505
T502
114.9V C
5
6
Fig.1: the power supply circuitry
for the Contec MSVR-5383 (note:
primary side of the switchmode
supply not shown).
some parts have been shaded, leaving
a dot pattern which can make some
values hard to read.
However, with all that information
to hand, I felt reasonably confident
about tackling these sets. And, initially, most of the faults turned out
to be fairly routine. But then, about
12 months ago, I encountered the
first difficult one, which I am about
to relate, followed by a couple of real
weirdos. More about those in notes
to come.
Lost stations
So what was this one? The customer’s complaint was that if the set was
turned off at the power point, as can
easily happen when motel rooms are
being serviced, then all the channels
pro
grammed into it would be lost.
To get it working again, it had to be
reprogrammed. But this only occurred
if the power point was turned off; it
did not occur if the set was turned off
into the standby mode.
With the set on the bench, I found
the owner’s description quite accurate;
the only additional factor, which the
owner himself added, was that the
power point had to be off for something
like half an hour before the memory
D516
C
C523
C
1
2
3
Q508
Q509
was lost (the exact reason for this is
still a mystery). And so, having confirmed the situation, I was faced with
the problem of where to start looking. I
had never encountered such a fault or
anything remotely like it at that time.
Nor had I encountered any literature
explaining in detail how these systems
worked.
I turned to the circuit for inspiration but it didn’t help a great deal.
All I learned was that it was most
likely in or around one of three ICs:
IC801, IC802 and IC804. IC801 was
the 42-pin central processing unit
(CPU); IC802 was a 14-pin unit with
internal boxes captioned “memory
transistor array”, “address register”,
“address decoder”, etc; and IC804
was a 16-pin unit that contained an
oscillator, a display timing generator
and various other circuits. Both these
latter ICs were closely associated with
IC801 (the CPU).
Of those three, IC802 looked the
most likely possibility. But that was
– at best – little more than an educated guess. I needed more than that; I
needed some real help.
I decided it was time to put the
owner’s backup research to the test. I
rang the service organisation and was
put through to the service manager.
And that was a real bonus, because
he turned out to be most co-operative
and was familiar with many of the
firms and personnel that I dealt with
in Sydney.
More importantly, he was a mine
of information about the set. As soon
as I described the symptoms, he was
onto them. And I was right about one
thing; it did involve IC802. He drew
my attention to pin 2, which is marked
as -31.65V.
This voltage is derived from a small
50Hz power supply, based on transformer T501, on the main power supply board. This is the standby supply,
which means that it is activated while
ever the power point is turned on. It
provides the -31V rail from pin 8 of
T501 via a diode (D510), a 47µF filter
capacitor (C512), a 1kΩ resistor (R519)
and a zener diode (IC501). This goes
out on pin 4 of plug/socket F.
Subsequent analysis revealed that
this supply also provides a regulated
5V rail for the CPU and the remote
control receiver, from pin 7 of T501,
via IC502 and a simple filter network.
This goes out on pin 3 of plug/socket
F. And there is a 12V supply for IC904
(the audio output stage) from pins 3 &
November 1994 33
SERVICEMAN’S LOG – CTD
4 of T501, via a full-wave rectifier and
regulator transistor Q506. This goes
out on pin 1 of plug/socket E.
Having pinpointed pin 2, the service
manager came straight to the point.
“Check that 31V rail. You’ll probably
find it either zero or very low. We have
had cases where IC802 has failed internally and taken out the 1kΩ resistor
in the power supply”.
Well, that was about the most
succinct diagnosis I can remember.
I thanked the gentleman and went
back to the bench. And he was dead
right; the voltage on pin 2 was down
to a couple of volts and the cause was
R519, which had gone very high. But
there was more; diode D510 was also
faulty. Not completely open circuit;
more partially broken down. It may
still have been providing some rectification.
After that, the job was pretty much
routine. I replaced R519 and D510
and fitted a new IC for IC802, and we
were back in business. Of course, the
set had to be reprogrammed but once
that was done, it would hold the program regardless of the condition of the
power point.
So the set went back to the customer
and 12 months later it is still behaving
itself.
But it was a valuable experience,
and I learned a great deal from it. And
the experience was to prove invaluable
more recently when, as I have already
hinted, there were more problems
in this area but with quite different
symptoms and different causes. More
about those in future notes.
Another motel set
My next story concerns a Samsung
34cm colour set, model CB-349F, one
of several belonging to another local
motel. In fact, these sets have featured
in these notes before.
The complaint – or rather the
problem – was lack of bright
ness.
And I make this distinction because
the complaint was poor colour; they
claimed they couldn’t adjust the colour properly.
Taken at its face value, such a complaint would suggest lack of colour
saturation. Unfortunately, some people have difficulty in differentiating
between colour and brightness – it’s
34 Silicon Chip
all the same to them. In fact, there was
no colour problem; it was simple loss
of brightness.
Where the loss is only slight, one
might be tempted to adjust sub-brightness control VR203 – a 2kΩ pot. Another possibility is to increase the setting
of the “screen” or G2 voltage control,
or even try adjusting both.
However, these tricks smack of a
quick fix approach and are best avoided in most cases. But it is wise to check
these two controls, in case someone
has had a fiddle (it does happen) –
they should be in about mid-position.
In any case, unless the G2 voltage is
significantly below normal, it is best
left alone.
There was no temptation in this
case. The brightness loss was considerable, which clearly indicated a fault
of some kind. And I had a pretty good
idea where it would be.
The most frequent cause of this
problem in these sets is resistor R208,
which has a nasty habit of going high.
It is part of the beam limiting circuit
and is connected between the 125V HT
rail and pin 4 (pedestal clamp) of the
main IC, IC501. (It’s a swine of a thing
to find on the circuit, being tucked
away down below the horizontal output transformer).
Its value is not given directly on the
circuit but in a table on the side. For
14-inch and 16-inch tubes – which
covers this one – it is given as 127kΩ
0.5W. For 18-inch or 20-inch tubes,
the value is 110kΩ 0.5W.
And it has something of a history.
Back in May 1990, this resistor was
mentioned in a Samsung service note,
advising that it be checked for an increase in value. In fact, several such
cases were found. At the same time,
there was some initial confusion as to
what value these resistors were supposed to be, since they carried a colour
coding which didn’t seem to make
sense with the values in the table.
However, that’s all by way of background because, in this case, the resistor measured spot on. From there,
I checked the G2 voltage and found
that it was down significantly. And
I seemed to recall that there was another common fault which produced
these symptoms but, for the moment,
I couldn’t remember the details.
A caffeine fix helped and the memory suddenly clicked. Of course, a
capacitor on the neck board – in particular, capacitor C519, a 330pF 1kV
disc ceramic which bypasses the G2
line to chassis. As soon as I saw it, I was
even more convinced that I was on the
right track; it was a blue disc ceramic
and I recalled encountering these in
the power supply on several previous
occa
sions, where their tendency to
leak caused some nasty problems. And
I had mentioned their unreliability in
these notes at the time.
Well, that was it. I pulled it out,
confirmed the leakage, and fitted a
new one. And that was it; normal
brightness returned and I had another
satisfied customer. Naturally, the new
capacitor was a different brand to the
original – one about which I feel a lot
more confident.
It hadn’t been any big deal but it did
make me think about some of those
previous faults which I had temporarily forgotten. One’s memory needs to
be jolted from time to time.
The HMV portable
To finish off, here is another story
which in itself was no big deal. In fact,
it was little more than routine but I
decided to tell it because it presents
an opportunity to discuss a couple of
important points.
The set involved was an HMV portable TV set, model 8010501, a 34cm
unit actually made by JVC. It’s getting
a mite long in the tooth now, being
some 15 years old at least, but is still
a goer for all that.
The customer’s complaint was
straight to the point; no picture and
no sound. When I set it up on the
bench this proved to be literally true.
But there was a raster and some white
noise from the speaker; it wasn’t much
use to the customer but was quite valuable as far as I was concerned.
Even more valuable was a very
prominent hum pattern on the raster
and the fact that this was also shrunken
on all four sides. So it didn’t take a
genius to deduce that we had a power
supply problem. And finally, there was
the HT rail. This should have been
around 110V but was actually only
about 80V.
Likely causes? The first thought
is almost automatically to blame the
main filter capacitor but there are
other possibilities. For example, failure of the voltage regulator transistor
I had measured it properly, or whether
there was a fault in the tester. A few
quick checks soon ruled out those
ideas, so I patched it back into circuit
and gave it another try. I could have
saved my time; it simply would not
work.
Well, it’s not the first time I’ve struck
such contradictions. And it emphasises the old rule that the final test for
any component is whether it will work
where it is supposed to work.
So what was wrong with the capacitor? It could be one of several faults
that show up in electrolytics but my
guess is that it was suffering from high
internal resistance. It is a known fault
and it means that, even if the capacitance value is correct, it cannot charge
or discharge fast enough to provide the
required function.
Capacitor compatibility
can produce a variety of symptoms,
including those listed above, this
depending on the exact nature of the
fault.
Another possibility is a faulty bridge
rectifier. Failure of one diode will
result in only half-wave rectification,
with reduced voltage and lots of hum.
But these thoughts were quickly
put on hold. Time enough to worry
about them if the most likely culprit
was cleared. So I went straight to the
main filter capacitor, a 600µF 180V
electrolytic.
The easiest way to check this is
simply to clip another one across
it. But hold on – not while the set is
turned on. A large value discharged
capacitor is, in effect, a short circuit
and connecting it across a HT supply
with another capacitor already in circuit will create an almighty splat. And
the spikes such a splat can generate
on the HT rail can produce some unpleasant surprises – like defunct ICs
and transistors.
And so I switched the set off, fished
out an appropriate capacitor and
patched it into circuit via a couple
of clip leads. And that was it; when
I switched the set back on, we had
110V on the HT rail, normal picture
and normal sound.
But while I considered the point
proved, there was one surprise. When I
pulled the old capacitor out, I put it on
the capacitance tester. And according
to that it was OK; it measured just a
whisker under its rated 600µF.
For a moment, I wondered whether
So that clarified the diagnosis. But it
was not quite the end of the job; there
were some practical problems still to
be solved. The test capacitor I had used
was physically incompatible, as were
all the others of suitable value which
I had in stock.
The closest would fit in the space
OK but its leads did not match the
mounting holes in the board. And
considering the age of the set, finding
an exact replacement would, at best,
call for considerable time and effort;
time which would cost money and
inconvenience the customer.
In these circumstances, I felt that
a certain amount of improvisation
would be justified. In fact, it wasn’t
all that difficult. There was enough
space around the mounting area, on
the underside of the board, to permit
drilling a couple of new mounting
holes to suit the replacement capacitor. The copper pattern was cleaned
around these and the new lugs soldered to them. It made a perfectly
satisfactory job, with a minimum
of delay.
But there was a rather interesting
aftermath. A couple of weeks after I
had finished the job, I came across
the faulty capacitor on the bench and
I hooked it up to the capacitance meter again. And this time it read about
200µF. Later again, when I came to
write these notes, I tested it again and
it measured virtually zero.
Well, it had taken a long time to
completely die but it had finally given
up the ghost. And good riddance. SC
November 1994 35
|