This is only a preview of the August 1995 issue of Silicon Chip. You can view 31 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 "Vifa JV-60 2-Way Bass Reflex Loudspeaker System":
Items relevant to "A Gain-Controlled Microphone Preamp":
Articles in this series:
Items relevant to "Build The Mighty-Mite Powered Loudspeaker":
Items relevant to "Computer Bits":
|
SERVICEMAN'S LOG
It took a little longer than usual
At the risk of seeming to state “the bleedin’
obvious” – as our English colleagues would say –
my main story this month is an unusual one. But
then they usually are. OK; so this one is more
unusual than usual, if that makes sense.
The story really started over two
years ago but can only be told now
because that’s how long it took to
finalise the job. Well, as I said, it is an
unusual story.
The device involved was a Philips
colour TV set, fitted with a KT-3 chassis. The KT-3 chassis was fitted to a
whole range of Philips sets and this
one was a 48cm model about 14 years
old – the age being a matter of some
importance as it turned out.
And the complaint was simply
that the set would fail intermittently.
Sometimes it would fail at switch-on,
sometimes after it had been running
for some time. More importantly, from
the owner’s point of view, what had
begun as a very occasional problem
had become progressively worse. It
was now likely to be off more than
it was on.
On the bench it was just as the customer had said; it was only nominally
intermittent and faulty most of the
time. But when it was functioning, it
performed very well. But there was a
major symptom which he had not noticed; the power supply was hiccuping
away quite merrily.
This symptom normally indicates
an overload of some kind, such as a
failed horizontal output transistor, or
something in this part of the circuit. It
looked fairly straightforward, as over
loads are not usually all that hard to
track down. The trouble was, I didn’t
see Murphy lurking in the corner.
I pulled out the horizontal output
transistor and checked it. It checked
OK but I replaced it anyway. I doubted
that this was the culprit. I’ve never
known this component to fail inter
40 Silicon Chip
mittently; when they fail they don’t
muck about but there is always a first
time. I also checked the insulating
washer. It looked OK but I replaced
it also.
What about the horizontal output
transformer? Could be, except that I
would expect it would start overheating if allowed to run in fault condition
for even a short period. But no; no sign
of trouble there.
A puzzling aspect of the fault was
the effect on the HT rail. This was
down to around 35V but varying up
to about 50V at times. This seemed
to rule out a dead short and, in fact,
resist
ance measurements failed to
find any evidence of short circuits
anywhere. This seemed to suggest
that it was either an AC fault of some
kind, or something breaking down at
operating voltage. But what?
Replacement boards
In a sense, I had come to something of a dead end. All the usual
approaches to a fault of this kind had
failed and I had to think of a new one.
Fortunately, over the years, scrapped
chassis had provided me with a useful collection of boards for this and
similar chassis. So this was my next
step; replace each suspect board until
I found the culprit.
In all, there are six plug-in boards,
plus a plug-in IF pack and the ELC2060
tuner which is permanently fitted. I
imagined swapping a couple of boards
would probably be enough to give me
a clue, unless I was very unlucky.
And the truth is, I was very unlucky.
I finished up changing every board and
was no closer to solving the problem.
The only thing that had changed was
that the fault was no longer intermittent but was now permanent. This was
a minor plus in terms of convenience
but no help otherwise.
At this point, I had really run out
of ideas. I needed time to think and
there were other more urgent jobs
waiting, so I put it to one side. I find
it is often a help to take a break like
this; one can dwell on a problem for
too long and frustration clouds one’s
judgement. It is surprising how often
an idea will suddenly pop up when
least expected.
It didn’t quite happen like that this
time but I’m sure that the break did
help. When I pulled the set out of the
corner of the bench a couple of weeks
later, I could take a broader look at the
problem. What had I missed; what
hadn’t I checked?
Well, I hadn’t checked the scan
coils. That idea was a long shot –
scan coil failures are extremely rare.
I doubt whether I have encountered
half a dozen in the last 20 years. More
particularly, I had never had one in
a Philips set. So, against that background, one tends to take them for
granted. But I couldn’t take this one
for granted; the scan coil assembly
had to come out.
It’s a simple enough operation – release the neck board and ring convergence magnets, then undo the clamp
holding the scan coil assembly and
slip it off the neck of the tube.
My idea was to hook it up to my
shorted turns tester, a shorted turn
being the most likely fault. And I was
right about that. But I didn’t need
the tester to tell me; one glance was
enough. There was a large blackened
patch on one of the horizontal windings, surrounded by spots of green
corrosion which had obviously caused
it all in the first place.
So that was it; the scan coils were a
write off. Well, at least I’d diagnosed
the problem, even if it had taken more
effort than I would have liked. And the
solution seemed simple enough – a
new set of coils.
But it wasn’t that easy. A new set of
coils would be in the $100 plus category which, with labour costs, might
be difficult to justify for a 14 year old
set. And, unfortunately, scan coils for
this set was one thing I didn’t have in
the junk pile.
So I rang the customer with a typical
good-news-bad-news report. The set
could be fixed but the cost might be
hard to justify. All I could suggest was
that I might be lucky enough to score
a set of coils if another customer’s set
was written off. But, of course, we had
no way of knowing when, or even if,
this would happen.
He thought about it briefly, then
decided that a new set of coils was not
a proposition. On the other hand, he
asked if he could leave it with me for
the present, in the hope that another
set of coils might turn up. He is a good
customer, so I readily agreed although
I wasn’t very optimistic.
And that is about the end of that part
of the story, which all happened over
two years ago. I relegated the set to
spot in the junk store and more or less
forgot about it except when another
such set came through the workshop.
Unfortunately for this customer, they
were all routine jobs.
Back to the present
Which brings us to the present
time. And to another character who
became part of this story. I’ll call him
Lance for convenience but that is not
his real name. He is a young married
bloke who, to put it mildly, has had a
pretty rough trot.
Lance’s main activity is repairing
or rebuilding discarded sets which he
donates to a local charity that has stood
by him and his family over the years.
His main source of scrapped sets is a
dealer/serviceman in a nearby suburb
but he also drops in on me – and some
of my colleagues – from time to time
for a word or two of advice and to
scrounge a few parts from stuff earmarked for the tip.
Soon after I first met Lance, I raised
the matter of a set of scan coils for the
KT-3, asking him to a look out for such
a set. That was over a year ago and he
had had no luck whatsoever.
Then, a couple of months ago,
Lance walked in with what looked
like a complete KT-3 set. In fact, it
wasn’t complete, consisting only of
the cabinet (in very good condition)
picture tube and scan coils. There was
no chassis.
A visit to my store room produced
a chassis plus a set of boards and so
we had the makings of a complete
set. Unfortunately, it didn’t work out
that way. When we fired it up, the set
went into a hiccup mode exactly like
the previous one.
I wasn’t sure it was the same fault,
of course, and I made a few routine
checks initially but I could find nothing else obviously wrong. So off came
the scan coils and, yes, that was it;
August 1995 41
exactly the same pattern of corrosion
and self-destruction.
Which put us back pretty well to
square one, except that we now had
two sets needing scan coils. The only
difference was that set number two
was something of an unknown quantity – we had no way of determining
the condition of the picture tube.
Nevertheless, we agreed that it was
now worthwhile looking for two sets
of coils.
That was being somewhat optimistic, I suppose, but he did eventually
find one set of coils, which he turned
up with a few days ago. They looked
to be in good condition, so I lost no
time in fitting it to my customer’s set.
And it worked. However, before
doing any setting up, I pulled the coils
off again and fitted them to Lance’s
set. Unfortunately, it wasn’t such a
good result this time. The tube was a
write-off, one gun being completely
dead. I even tried boosting it but to
no avail.
So the coil assembly was refitted to
my customer’s set and I went through
42 Silicon Chip
the setting up procedure, after which
the set delivered a first-class picture.
Finally, I rang the customer and advised him to come and collect the set.
He was happy that the repair bill had
been kept to a reasonable level and I
was happy to finally be reimbursed for
my time and effort.
Regrettably, Lance didn’t come out
of it quite as well, although he is still
looking and hoping. But he needs both
a picture tube and a scan coil assembly
now, so he will need to be extra lucky.
That said, he has scored a chassis
and learnt something about scan coil
failure and the symptoms it produces,
so it hasn’t been a completely wasted
effort on his part.
All of which just goes to show
what can be achieved if the customer is prepared to wait. But I must
concede that this has to be regarded
as a one-off; one that just happened
to work out.
The flasher
My next story is about a lady customer who was troubled by a flasher.
No, not one of the raincoat mob –
rather, a TV set. To be more precise,
it was a Samsung 51cm colour set – a
model CB-515F fitted with the P-50F
chassis.
According to the lady, the problem
was random white flashes on the
screen. Initially, this happened only
occasionally and she put it down
to interference from some external
source. This theory was reinforced by
the fact that she lives in a gully which
is a relatively poor signal area.
More recently, the problem had
become more frequent and, at times,
much worse. Sometimes, it was so
bad as to make the picture virtually
unwatchable. Still thinking that it
might be interference, she took the
opportunity to operate the set while
staying with a friend in a much better signal area. And it did clarify the
point; the problem was just as bad in
this location.
And so the set finally landed on my
bench, along with the above explanation. I put the set on the air while
the lady was still in the shop but, of
course, Murphy was lurking in the
corner – it behaved perfectly.
All I could do was suggest that she
leave it with me, which she did. I set
it up in a corner of the bench and it
Fig.1: the front end of the Samsung CB-515 colour TV set. The tuner (TU001) is at top left, the IF IC (IC101)
at lower right, & the SAW filter & its associated components below the tuner.
ran for several days without any sign
of trouble. But then came the first hint;
a brief white flash, no more that a few
centimetres long on one line. Blink
and you’d miss it.
Nothing more happened for about
a week then, one day, it really turned
on an act with flashes all over the
screen. These became progressively
worse until it was quite unwatchable.
Well, at least I’d seen the problem and
that is always useful. And, in fact, I’d
already made a tentative diagnosis – I
was sure it was a front-end problem,
most likely the tuner (TU001).
Fortunately, this was a relatively
simple theory to prove – or disprove.
I had a spare tuner on hand and it was
easy enough to substitute it. And the
set ran perfectly after that – for about
a day and a half. Then it was back in
flashing mode.
OK, but I still felt sure that it had to
be somewhere in the front end. So the
next thing to try was the transistor in
the first IF stage (Q161, 2SC388), mainly because it was a simple operation.
But again, no joy.
Desperation measures
Things were looking somewhat
desperate now but there was one good
point; I had a good stock of spares
from previously junked sets, which
meant that I could replace almost
any component, at least in the front
end. This is a useful approach in
such circumstances but it can be time
consuming.
So, in turn, I replaced the SAW filter
(Z101), it’s associated matching transformer (T101), the video detector coil
(T171), the AFT balance coil (T172)
and, in desperation, the IF IC (IC101,
LA7520).
As readers will appreciate, all this
took a lot longer than it does to write
about it. In fact, the entire process took
several days, taking into account the
time taken to monitor each change.
And in the end it was all to no avail;
I was back to square one.
So was it in the front end? Or could
it be a fault in the horizontal scanning
system; a breakdown or flashover
which was generating interference? I
rang the Samsung service department
and spoke to a contact there who has
always been very helpful. He wasn’t
able to offer any ideas based on actual
cases but he did agree that the idea
of interference from the scan system
was worth investigating. And he went
on to offer some ideas as to how the
front end could be operated with the
scanning systems shut down.
As a result, I finished up with the
rear end of the set shut down and the
front end operating on 12V and 33V
from a bench power supply. I then
hooked up the CRO to monitor the IF
envelope – it has enough bandwidth
to do this – and fired up the front end
independently.
Glitches
Sure enough, in the fullness of
time, I could clearly see a succession
of glitches on this envelope. I had no
doubt now that I had been right the first
time; the fault was somewhere in the
front end. But where? I had checked
or changed all the likely components
in this section.
Or had I? No, there was one section
I hadn’t checked; the channel selector pushbutton assembly, shown as
PWB-SELECTOR. There are in fact two
versions of this unit, an 8-key assembly
and a 12-key assembly, the latter being
the one shown.
This assembly carries the push
buttons which are used to select
the preset channels, plus the preset
controls themselves. The pushbuttons are shown in the centre of Fig.2,
August 1995 43
SERVICEMAN’S LOG – CTD
designated as SW01, SW02, etc. They
have two sets of contacts, those on the
right performing the channel selection
function and those on the left selecting the appropriate band. These latter
contacts also activate the associated
indicator LED (DL01, DL02, etc) from
the 12V rail. These LEDs are connected
in series and all but the wanted one
are shorted out.
There are two preset controls for
each channel. On the extreme left are
the 3-position band selector switches,
marked VL (VHF low), VH (VHF high)
and U (UHF). These are preset for
the band appropriate to the channel
chosen for that position. They are
connected to the 12V rail when a
channel button is activated and apply
base voltage to one of three transistors – SQ101, SQ102 and SQ103. The
selected transistor then turns on and
connects the main 12V rail to the appropriate section of the tuner.
The chosen channel is selected by
the corresponding variable resistor on
the right (VR01, VR02, etc). These are
fed from the 33V rail and feed an appropriate voltage to the varicap diodes
in the tuner (terminal VT).
As an example, switch SW06 is
shown in the active position (with the
LED illuminated), VR06 is connected
to the 33V rail, and the 12V rail is
44 Silicon Chip
connected to the VH position of the
band selector switch.
All of which should give the reader
some idea of the complexity of these
assemblies. And to be truthful, they
have more than their fair share of troubles although, until now, these have all
involved channel selection problems.
These are easy enough to diagnose and
there is only one practical solution;
replace the entire assembly.
However, I have never experienced,
or heard of, these units causing the
kind of trouble evident in this set. But,
with all other likely culprits exonerated, this one had to be a suspect even
though it was something of a long shot.
Fortunately, I had a spare unit on hand
and this was duly fitted.
And that was the answer. The set
was run for several days with no sign
of the fault and I eventually returned it
to the customer. But I warned the lady
to contact me immediately if it should
reappear. That was many weeks ago
and all is quiet so far.
More flashers
That wasn’t the last of the flashers.
Within a few weeks, I had no less
than three more and all from the same
cause. The only difference was that it
involved different bands; one was on
low VHF, one on high VHF, and one
Fig.2: this diagram shows the
channel preset & channel selector
circuitry in the Samsung CB-515.
Note the active setting for switch
SW06.
on UHF. And they were not all CB515s; one was a CB-349 and one an
Akai CT-K115, both of which use the
P-50F chassis.
Granted, these were the bands
favoured by each customer and my
checks confirmed that the fault occurred only on the particular band.
And there was a fourth set with an
even weirder fault in this section. The
set worked normally on UHF but suffered from very low gain on low VHF.
For some reason (probably due to
leakage), when it turned on the low
VHF transistor (SQ101), it also turned
on the high VHF transistor (SQ102)
at the same time. The effect was to
completely wreck the low band gain.
So be warned; any similar funnies
and you’ll know where to look. SC
|