This is only a preview of the April 1994 issue of Silicon Chip. You can view 28 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 "Remote Control Extender For VCRs":
Items relevant to "Sound & Lights For Level Crossings":
Items relevant to "Discrete Dual Supply Voltage Regulator":
Items relevant to "Low-Noise Universal Stereo Preamplifier":
|
SERVICEMAN'S LOG
Nothing unusual happened this month
I don’t have any stories from my own bench
this month, since nothing sufficiently unusual
has happened. So I have had to call on a
couple of colleagues, who have come to the
party with some really tricky ones.
The first story comes from my colleague on the NSW south coast and
it concerns a problem peculiar to his
area.
In fact, for readers not familiar
with this area, it is necessary to set
the scene in terms of the local TV
channels. In the early days of TV,
residents of Wollongong, about 80km
south of Sydney, managed as best they
could with signals from the Sydney
channels.
It was a chancey business. As well
as the distance, they had to contend
with less-than-favourable topography.
Tall masts, high gain antennas and
masthead amplifiers were the order
of the day. Some managed reasonably
well; others took what was there on
a day-to-day basis and were grateful
for it. Further south, around Nowra,
Bateman’s Bay and their surrounds,
it was virtually hopeless.
First relief for the area came with
the establishment of a couple of VHF
transmitters at Knight’s Hill, south of
Wollongong. Today, the area is served
entirely by five program sources in the
UHF band, using five main transmitters and five translators.
In order to appreciate my colleague’s
story, it is necessary to set out these
channels. The five main transmitters,
at Knight’s Hill, use channels between 53 (701-708MHz) and 65 (785792MHz), while the translators use
channels between 30 (540-547MHz)
and 48 (666-673MHz).
So, against that background, here’s
my colleague’s story, more or less as
he related it to me.
The fussy Rank
The set was a Rank-NEC model
C-1413. It was brought in many months
ago by one of my lady customers with
the complaint that “the picture goes
funny on some of the channels only”.
Well, I’ve had worse descriptions
although, as it turned out, it was
AUSTRALIAN MADE TV TEST EQUIPMENT
12 Months Warranty on Parts & Labour
DEGAUSSING WAND
Great for computer monit
ors. Strong magnetic field.
Double insulated, momentary switch operation.
Demagnetises colour picture tubes, colour
computer monitors, poker machines, video
and audio tapes. 240V AC 2.2 amps, 7700AT.
$85.00 + $10.00 p&p
Cheque, Money Order, Visa, Bankcard or
Mastercard. Phone for free product list
SHORTED TURNS TESTER
Built-in meter to check EHT
transformers, including split
diode type, yokes & drive trans
formers. $95.00 +$3.00 p&p
TV, VCR
TUNER REPAIRS
From $22. Repair or Exchange
216 Canterbury Rd, Revesby, NSW 2212, Australia.
Phone (02) 774 1154 Fax (02) 774 1154
40 Silicon Chip
accurate enough; it just wasn’t very
helpful. In cases like this, one of the
points I have to watch in this area is
the need to determine, right from the
start, which transmitters a customer
uses; the main transmitters, the translators, or even combinations of both
in odd cases.
This fine distinction is usually lost
on most people; they think in terms of
names – ABC, SBS, Prime, etc – with
little appreciation of how the program
comes to them. But some careful questioning in this case finally pinpointed
the main transmitters on Knight’s Hill
as the signal source.
So that was where we started. The
model C-1413 is one of a series of
sets which use essentially the same
circuit and appear under several brand
names. As a Rank, it also appears as
the C-1414 and C-2020 (among others)
but it also appears under the GE label
as GE482 and under the General label
as GC205.
I have most of these circuits, including the C-1413 and the C-2020.
As luck would have it, the 2020 came
out of the file along with the 1413 and
I left it out.
As far as the symptoms were
concerned, the lady was right and
the pictures from Knight’s Hill were
“funny”. And that wasn’t such a silly
term either. The effect isn’t easy to
describe; my best attempt would be
random pulling and rolling, with the
suggestion that this might have been
hum related. The lady was also right
in that there was no sign of the trouble
on the translator channels. In fact, I
explored this aspect very thoroughly
to make quite sure.
OK, so we had a frequency related
problem. That meant trouble somewhere in the front end; probably in
the UHF tuner itself. This set uses
two mechanical tuners; the UHF tuner
which down-converts to VHF, and a
VHF tuner which then down-converts
these signals to the IF. Of course, it
can process off-air VHF signals as
one carrying the supply rail, the AGC,
AFT and chassis connections, one for
the IF lead, and one for an auxiliary
network for the UHF tuner RF bias
control. With the two sets close together, there was enough lead length
to allow the suspect tuner assembly
to be replaced with the known good
one, without any need for mechanical
demounting.
I was fully confident that this would
confirm that there was a tuner fault. If
so, it would make things easy because
I had a couple of spare tuners from
junked sets and a replacement would
be a cheap and easy solution.
But no – the set behaved exactly
the same with the replacement tuner
assembly. This was a really revolting
development; any complacency I
had allowed myself up to this point
was completely dispelled. I had a
real stinker on my hands, defying all
the rules. If the fault was in the main
circuit, which was carrying nothing
higher in frequency than the IF, then
how did it know when the set was
tuned to something above 700MHz?
And if use of the word “know”
sounds a bit way out, it was no more
so than the fault itself.
Caffeine fix
well. It’s all quite conventional really
and I imagined that the job would be
fairly routine.
The first thing I tried was feeding in
a signal from the colour bar generator.
This happened to be set to channel
36 (575-582MHz) in the translator
group and, as expected, it produced a
perfectly steady picture.
I then reset the generator to channel
67 (799-806MHz) in the transmitter
group, fully expecting that I would
be able to observe the fault under controlled conditions. But not a bit of it.
The colour bar signal was just as steady
on this channel as it had been on the
lower one. It was a nasty setback.
OK, so it was back to the real world.
I switched to one of the Knight’s Hill
transmitters and confirmed that the
fault was still very much alive. So
what next?
Well, luck was with me; I had another identical set in the workshop at
this time. There wasn’t much wrong
with it and I could use it for a spot
of swapping. In particular, I had in
mind to swap the tuner assemblies,
thus either confirming or rejecting this
section as being at fault.
It was a simple exercise. The tuner
assembly is connected via three plugs,
After I’d had a caffeine fix and
calmed down a little, I had another
thought. Could it be a power supply
fault? A long shot surely – how could
the power supply be involved? It was
just about as far removed from the frequency selection process as anything
could be.
Yet I’ve had some very funny faults
traced to power sup
plies. We tend
to forget that there are signal paths
through or around all power supplies,
usually involving capacitors other
than the filter capacitors, and that
failure of these can create faults a long
way from the source.
More to the point from a practical
point of view, it took only a few minutes to patch the power supply from
the other set into this one and settle the
point once and for all. And it did – it
made no difference.
The next most likely possibility was
distortion of the sync pulses. I could
think of no way that such a fault could
be frequency conscious but the idea
could not be ignored.
The IF signal from the tuner goes
through an amplifier stage (TR201), a
SAW filter (FL201), and thence to pins
April 1994 41
Fig.1: the IF circuitry in the Rank C-1413 colour TV set. The IF input is at extreme left & feeds TR201, while C208 is
below IC201, between pins 11 & 14. Note capacitors C221 and C222 from pin 11 to chassis.
1 & 16 of IC201 – see Fig.1. It emerges
on pin 12 and a clear staircase waveform pattern is given for this point. So
the CRO was hitched to pin 12 and the
colour bar generator used to provide a
staircase signal.
But, again, there was no cry of “Eureka” – or a triumphant dash down
the main street. As far as I could see
the waveform was perfect; exactly
according to the circuit, with no hint
of dis
tortion or compression, from
either a channel 36 or channel 67
signal. Nor did the signal level appear
to matter. The generator can deliver
a solid signal – stronger than most
off-air signals in practice – and so I
took this right down until the pattern
dropped out of colour. The sync pulse
remained perfect.
Nor did varying the AGC adjustment have any effect. But what about
off-air signals? I checked some of the
translator signals and the sync pulses
appeared much the same as from the
generator. The pulses from the Knight’s
Hill transmitters didn’t look too bad
either, although that “too bad” implies
a qualification.
Yes, the shape was still OK but
one difference did catch my eye, although I still don’t know whether it
was relevant. As I said, the shape was
correct but there appeared to be some
rubbish, or noise, inside the pulse
rectangle. It was nothing that could
be resolved and is still a mystery.
42 Silicon Chip
All I know is that it was only on the
troublesome signals.
Hard slog
So now it was down to hard slogging
and a lot of hope. I changed transistor
TR201, the SAW filter, and even IC201.
I changed the capacitors on the AGC
line (pin 4), including C210 (4.7µF),
C212 (0.01µF) and C209 (4.7µF), plus
sundry resistors. None of these had
any effect.
By this stage, I had reached the
point where I had to stand back and
take a long hard look at the whole
situation, not only technically but
financially. I had spent a lot of time
on it; somewhat more that could
reasonably be justified for something
which was now looking as though it
might be a write-off.
I contacted the customer and
brought her up to date on the situation. She was quite co-operative,
in that she was in no hurry to get
the set back. To be truthful, I gained
the impres
sion that she was quite
prepared to write it off. But she was
prepared to spend up to $100 to get
it working. So, at least the pressure
was off. I had more pressing jobs to
attend to and so the set was put aside
in one corner of the bench.
I had fully intended to get back to
it reasonably soon but, as often happens, other jobs kept piling up and
I kept putting it off. And so several
months went by. But its mere presence provided a nagging factor; every
time I looked at it, I felt guilty – and
apprehensive. I had no idea how I was
going to tackle it.
Eventually, when things slackened
off over the Christmas/New Year break,
I knew that the moment of truth had
come. I fished it out again, determined
to settle the situation one way or the
other.
Fortunately, I had scribbled a few
notes and kept the components which
had been changed, so I was soon back
in the picture. But I didn’t really have
any fresh ideas. The best I could do
was to continue replacing likely – or
even unlikely – components and hope
for a breakthrough.
And that’s what happened. After a
couple of false tries, which included
capacitors C222 (47µF) and C221
(0.01µF) in parallel with it, I came
to capacitor C208 – a 2.2µF tantalum
electrolytic connected between pins
11 & 14 of IC201. I hadn’t tried it earlier
because it was hidden on the copper
side of the board. Only its presence
on the circuit diagram as part of the
AGC circuit sent me looking for it. Its
location may be significant, considering what followed.
I pulled the capacitor out and
checked it. It measured spot on but
I replaced it anyway. Well, sort of – I
didn’t have a 2.2µF capacitor handy, so
I settled for 1µF. The result was quite
dramatic; not a total cure but such an
improvement that I could have almost
let it go. There was just an occasional
tendency to pull.
My natural reaction, initially, was
that if I fitted the correct value, it
would complete the cure. It might have
too, but a number of things happened
to change my approach.
To explain this, I have to make it
clear that I had abso
lutely no idea
as to the function of this capacitor.
I didn’t even know the function of
pin 11; whoever drew the circuit had
omitted to identify it. In hindsight,
I could have easily worked it out by
tracing the circuit but didn’t I realise
this. But I was curious about C208.
It was then that the C-2020 circuit
stuck its nose in – see Fig.2. It had
been updated, with pin 11 marked
as VCC (the supply rail pin). On the
C1414, it connects to the 12V rail via
a 22Ω decoupling resistor (R223), bypassed by the previously noted C221
and C222.
Suddenly, C208’s role became clear.
It is a bypass capacitor for pin 14. Pin
14 is part of the internal AGC circuitry
and, while I have no idea of its exact
role, it is clear that it is held at around
7V by a 560kΩ resistor (R209) to chassis. And C208’s job is to peg this point,
at RF, to chassis.
Only it doesn’t go direct to chassis –
it goes to pin 11. And pin 11 is pegged
to chassis at all but DC by C221 and
C222. It’s a rather roundabout route
but a perfectly valid one – at least in
theory.
In the process of working all that
out, I became aware that the pin 14
bypass circuit was quite different in
the C-2020 circuit. In this case, the
bypass capacitor, now designated
C223 and reduced to 1µF, goes directly to chassis. In practical terms,
this seemed to me to be a much more
elegant approach and since someone
a lot smarter than I had preferred it,
why not try it?
The rest is history, as our political
commentators like to say; I made the
changes and it was a perfect cure. I
ran it for a couple of weeks before
calling the lady and it remained rock
steady.
So we had a happy ending. But I
realise that the story poses as many
questions as it answers. While it is
clear that the pin 14 bypass arrangement is critical – and that someone at
engineering level must have discov-
Fig.2: this is the IF circuit from the Rank C-2020. Note the changes on pin
11 & 14 compared with the C-1413.
ered this – I still can’t explain why the
fault was frequency conscious.
But then I’m not an engineer; I’m
only the poor bloke who has to try to
make sense of these weird situations
in the field. All I can do is learn from
the experience and, by passing it on,
perhaps help some other poor blighter from going round the bend – as I
nearly did.
Well, that’s my colleague’s story,
and a good one it is too – as a story.
But I can only sympathise with him
over the anguish and frustration it
must have caused. Nor can I answer
any of the questions it poses. Any
suggestions?
A thorny problem
And so to colleague number two;
old faithful, J. L. from the island way
down the bottom. Here’s his contribution.
By all that’s reasonable, the old
Thorn model 3504 should have been
junked 10 years ago. It was released by
AWA in early 1975 as the company’s
first ever colour TV set.
By 1984-85, the model was beginning to show its age and by 1990 most
examples had succumbed to the years
and were only to be found on the
municipal tip. However, a few have
survived and one of these came to my
attention last week.
I’ve been caring for this set for close
to 15 years. I don’t know to what extent
its longevity is related to my atten
tions but I do know this – when I last
worked on it about two years ago, it
still produced a good picture.
So when he called me last week and
said that the picture had gone “...all
purple” I wasn’t particularly worried. I
was confident that it wasn’t the picture
tube and fairly sure that it was going
to be a simple electronic problem. I
was even more convinced it was the
latter when he said that thumping the
cabinet sometimes restored the picture
to normal.
A purple picture results from the
loss of green content, so this problem
had something to do with the green
video output or the green gun. I firstguessed the former because, in the
3504, the video output load resistors
are etched onto a ceramic substrate
and I have found a number of these
breaking down recently.
In greater detail, the connecting
pins break away from the ceramic
where they are attached to the etched
pattern. I’ve had no luck resoldering
this connection; it is more practical
to replace the printed resistors with
discrete 10W units.
Nowadays, I don’t have to resort to
such subterfuge since I have a large
collection of good boards taken from
sets that have paid the supreme sacrifice. And so I decided it would be
easier to do the job in the customer’s
home, by simply replacing the suspect
board with a known good one from
this collection.
And this did appear to be the answer; at least for a few moments after
I had made the swap. But soon the
green part of the picture disappeared
April 1994 43
again and I had to admit that it was
really a different fault. So where to
start looking?
I seem to have a habit of making
the same mistake over and over again.
The mistake this time was to forget to
use my multimeter. Circuit voltages
are one of the best indicators of circuit performance, yet I always seem
to make this my last test instead of
the first!
If I had made a quick check of the
tube base board, I would have found
that the fault lay with the screen (first
anode) voltage, not with the cathode
or grid voltages. So having done at last
what I should have done at first, I set
about trying to adjust the voltage on
the green screen (pin 5).
The result was uncertain – the green
could be restored but it was erratic.
It was hard to say exactly how it was
varying, although the instability was
seemingly related to the position of the
screen potentiometer, R793.
All of this made me think that we
had a dirty screen pot. I’ve had these
before and they usually respond to a
44 Silicon Chip
squirt of contact cleaner. So, two or
three squirts later, the picture came
good and no amount of mechanical
abuse would alter it.
I let the set run for half an hour or
so while I had a very welcome cuppa’
with the owner. The picture never
varied and so, by mutual consent, we
declared the job done.
Pride commeth ...
What is it that they say about pride
coming before a fall? That night the
owner rang to say that everything was
back as it had been – no green, purple
picture, and all!
When I left the workshop next time,
I made sure I had packed a complete
convergence board. This panel carries
not only the many convergence controls but also the three screen pots and
their associated beam switches.
My plan was to change over the
whole board to make a quick and simple repair. What I had forgotten was
that the “new” board had been set up
for a different picture tube and would
have to be completely readjusted.
When I fitted it to the set, the picture
came up with plenty of green but the
convergence was grotesquely out of
ad
justment. I could see that it was
going to be a long operation to do a
complete convergence setup and I’d
already spent as much time on the job
as I could afford.
So I decided to refit the original
board and just change the doubtful pot
for one of the good ones from the new
board. And it was then that I found the
true cause of the trouble. As I prepared
to remove one of the screen pots from
the new board, I noticed that the three
beam switches were of two different
types. Then I remembered!
On a few occasions in the past, I
have found that these beam switches develop internal shorts. It seems
not to be a mechanical problem but
an electrical one within the switch
material itself. No amount of cleaning
compound will cure the trouble – the
only answer is to change the switch,
something that I had obviously done
on the new board at some time in the
past.
So instead of changing the screen
pot, I changed the beam switch, S752.
After a gray scale adjustment, we had
as good a picture as any I’ve seen on a
set this age. In fact, the colour, contrast
and brightness were all excellent, as
was the convergence.
When I commented about the excellent picture on a set of such advanced
age, the customer’s wife commented
that they only used the set for a couple of hours at night. News and the
early-evening soaps were all they ever
watched and it was never on during
the day.
I made a quick calculation. Two
hours a day, 365 days a year, for 18
years, makes over 13,000 hours! I seem
to recall something about 10,000 hours
being a reasonable life for a picture
tube. So this one is not only well past
its presumed lifetime but looks capable of going on for many hours yet!
The only problem is that the set has
only a VHF rotary tuner. However, the
owner professes to have no interest
in channels other than the ABC, so
perhaps he really doesn’t miss the
UHF facility.
Thanks, J. L. – I hope you can keep
the old clunker going for a few more
years. As for the UHF channels, why
not add a junked video recorder to the
set to tune these stations; one in which
the front end is still functioning? SC
|