This is only a preview of the March 1996 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:
Articles in this series:
Items relevant to "Programmable Electronic Ignition System For Cars":
Items relevant to "Automatic Level Control For PA Systems":
Articles in this series:
Items relevant to "A 20ms Delay For Surround Sound Decoders":
Articles in this series:
|
SERVICEMAN'S LOG
Sound reasons for confusion
Yes, audio problems are the order this month.
And before you sniff disdainfully, let me assure
you that audio problems in TV sets “ain’t what
they used to be”. Like everything else, they’ve
gone high-tech.
The first story is about an NEC
colour TV set, a model N-4830 using
a C500 chassis and made by Daewoo.
The fault was straightforward enough;
it was completely dead. More than
that, the reason was obvious, at least
to anyone who had seen it before.
This set has an inherent weakness.
The power supply uses a 10-pin IC
switching regulator (I801), a type
STK-73410II. This IC overheats and
fails, taking a lot of components with
it. And that is what had happened in
this case. But there is more to the fault
than that.
It appears that the failure is not
the fault of the IC itself. Rather it is
in the associated circuitry and the
makers have issued a modification
instruction to cover this. Attention to
this is most important because if I801
is simply replaced, the set will back
again in a few months with exactly
the same fault.
The modifications are quite extensive and I don’t propose to set them
out here. Quite apart from the space
needed, I under
stand that since I
did this job, further modifications
have been issued which I have yet
to receive.
My original modification sheet
came from NEC Spares, 23-25
Coombes Drive, Penrith, NSW 2750
– phone (047) 21 7655. It would be
wise to contact them for the latest
information.
Suffice it to say, for the purpose of
this story, that I carried out all the
modifications prescribed at that time,
replaced the IC, then switched on with
my fingers crossed. And, initially, all
seemed well. There was no smoke,
48 Silicon Chip
fire or confusion and the picture tube
warmed up normally and presented
a first class picture. There was only
one thing wrong; there was no sound.
In the good old days, audio faults
were relatively rare and, when they
did occur, they were usually quite
easily found and fixed. It’s not quite
the same these days, at least on some
of the more elaborate sets such as
this one.
Signal tracing
I commenced searching with an
audio probe attached to a test amplifier, starting where the audio comes
out of the IF module, at pin 9 (AUDIO
OUT). Well, there was audio there all
right, which was encouraging. All I
had to do now was trace it through
to the output stage and speaker. And
that, as they say in the classics, was
the hard part.
The audio path is a real aroundthe-world-for-sixpence arrangement.
From pin 9 it goes to pin 7 of I701 (on
a separate page), comes out of I701 on
pin 5, goes back to the IF module on
pin 8 (ATT IN), comes out again on pin
10 (ATT OUT), and is finally fed to the
audio output stage, I602.
Tracing all this out on the circuit –
with much turning of pages – was an
exercise in itself. Then I had to relate
this to the set itself, which involved a
similar order of complexity. But laborious though it was, it paid off.
Audio was being applied to pin 7
of I701 but there nothing coming out
on pin 5. At this stage, I hadn’t paid
a great deal of attention to the role of
I701 but now I needed to know.
It wasn’t hard to work out. This set
Fig.1: the terminal connections
for the IF module in the NEC
N4830. Audio comes out on pin
9, goes back in on pin 8, and
comes out on pin 10 to be fed to
I602 (the power output IC).
features audio and video input sockets
(J202, J703) on the rear of the chassis,
which allow external video and audio
signals (eg, from video recorders and
cameras) to be fed directly to the appropriate parts of the set. This avoids
the losses involved in modulating a
carrier, feeding it through the front
end, then demodulating it to recover
the original signals.
And I701 is a switching module
which feeds the audio and video sections of the set from either the front
end or from the external audio/video
inputs.
Why no audio?
So why wasn’t audio coming out
of I701? In an effort to clarify the
situation, I fed an audio signal into
the audio input socket, J702, and, lo
and behold, the signal went straight
through to the speaker. What’s more,
this situation remained, regardless
of the switching signal instruction
to pin 11. In short, the audio signal
path through I701 was jammed in the
external position.
The most likely possibility was that
the IC itself was faulty. The only other
one was that power supply or switching voltages to it were at fault but I
tended to rule this out on the grounds
that the video switching function was
normal.
Nevertheless, I went through the
motions. The IC is powered from the
12V rail at pin 6 and this was quickly
cleared. The switching voltage is applied at pin 11 and this was turning on
or off according to instructions from
the control unit.
That didn’t really surprise me. After
all, the device was switching the video
signal, so it should have been switching the audio signal as well. More to
the point, it pointed the finger fairly
and squarely at the IC.
So I pulled it out, fitted a new one,
and that was it. And as is often the
case, it all seemed so simple and obvious once I’d found it.
Why did the IC fail the way it did?
Who knows? It could have been a
simple random failure which just
happened to occur at this time, or it
could have been a byproduct of the
power supply failure. But it’s worth
keeping in mind.
The set involved was a Philips type
02CR035, model 20CT6750/75Z, using
a CTO-S chassis. It is a 48cm set, about
10 years old, with remote control and
options built in for Teletext. More
about the type number and options
later.
I didn’t deal directly with the customer – a hazard in itself, as it turned
out – but with a colleague who was
short of time and needed some help.
The complaint was intermittent
loss of picture, with a bright raster.
He had tried to find it but the intermittent nature had made it difficult.
He suspected a dry joint and nomi
nated “the board on the left” as his
prime suspect.
The board on the left was the VST
decoder board and I had my reservations about this diagnosis. It seemed
unlikely that a fault here would produce the symptoms described and
my experience with this board is that
Another sound failure
My next story is also about a sound
failure but that is the only similarity.
By comparison, the previous fault was
not unduly difficult to find, whereas
this fault was one of the most frus
trating I have encountered for a long
time. And the irony was that it didn’t
even start out as a sound problem.
Fig.2: the audio from the IF module in the NEC N4830 is fed to pin 7 of
I701 and comes out on pin 5. Note the external audio/video jacks, J702
and J703.
March 1996 49
Fig.3: the relevant sound circuitry in the Philips 02CR034/35. Audio comes
out of IC7664 (left) at pin 8 and goes to board 1070 via plug 3M12. It is
subsequently applied to pin 7 of IC7681 via plug 2M12.
it is remarkably free from dry joint
problems.
As it turned out, all this speculation
was wasted. When I switched the set
on, the fault appeared immediately – a
brilliant white raster with no picture
and no sound. And, as it turned out,
this condition was permanent.
I immediately suspected a voltage
error around the picture tube; most
likely a loss of cathode voltage, which
is effectively the bias between the
cathode and G1. And since the raster
was white, rather than coloured, it
suggested a fault common to all three
guns. The cathodes are coupled to
the RGB drive transistor collectors,
so this voltage supply was the prime
suspect.
And so it proved to be. The collectors are fed from a 180V rail but there
was no 180V. It didn’t take much
tracking to find out why. A 4.7Ω resistor, R3583, in the power supply
was open circuit. Apparently it had
been intermittent but had finally failed
completely.
I fitted a new resistor and we had our
180V and a first class picture. So far
so good but there was still no sound. I
wasn’t particularly worried about this,
as I imagined a spot of audio signal
tracing, as described earlier, would
quickly solve the problem.
Before doing this, however, I went
over the motherboard, checking it for
50 Silicon Chip
dry joints. In particular, I checked the
horizontal output transformer connections and remade several which looked
suspicious.
Having done this, I came back to
the sound problem. And, in order to
follow what happened next, it is necessary to describe the relevant part of
the circuit – see Fig.3.
The section shown is from the
motherboard. The audio comes out
on pin 8 of the demodulator module
(IC7664 – left) and goes via plug 3M12
to an auxiliary board designated 1070
(Tone Control/Headphone Panel). It
is then supposed to come back to pin
7 of IC7681, the power output IC, via
plug 2M12. I say “supposed” because
it didn’t. Using the audio signal tracer,
I detected audio at 3M12 but there was
nothing at 2M12.
Fun & games
So the fault was on board 1070
which, as its name implies, provides
remote control of the volume, treble,
and bass levels. And this was where
the fun and games started. I didn’t
have a manual for the 02CR035; the
closest I had was one for an 02CR034
which I believed used essentially
the same circuit. And it did, at least
for the main circuit, and I had used
it to track down the original power
supply fault.
But board 1070 was another matter.
There was a general similarity in that
the same IC, a TDA1524 (IC7001), was
used for the level control functions.
However, there were significant differences between my circuit and the
board in this set, which used additional circuitry and components.
This didn’t worry me unduly at first.
I confirmed that audio was coming into
the board (connection 4C6) and to pin
15 of the IC but there was nothing at the
IC’s output (pin 11). Next, I checked
the supply voltage to the IC (pin 9)
and all the associated components. I
could find nothing wrong and that put
suspicion squarely on the IC itself. I
ordered one and fitted it but no joy;
everything was just as before.
I was getting worried now. The IC
and everything around it tested OK,
yet I couldn’t get any audio through
it. I needed to delve into things at
greater depth and, clearly, I needed an
updated circuit. Fortunately, I found
a colleague with a CR035 manual but
my complacency was short lived.
Although this circuit accounted for
some of the additional circuitry, it
still wasn’t complete. Apparently
the set on the bench was a different
version again!
Service department
My next stop was the Philips service
department. I set out the circuit problem in some detail and was assured
that a manual was available –price $84.
I did a mental double take on that; it
was high, even by current standards.
Still, I needed it, and could need it
again. I placed an order.
The manual duly arrived and I
turned eagerly to the 1070 circuit.
Imagine my frustration when I realised
that it was no different from the one
my colleague had loaned me. I went
back to the Philips service department
– they were just as puzzled as I was
and could find no record of a circuit
which fitted the details I described.
Granted, after 10 years, the trail would
be cold. However, they agreed to take
the manual back and credit me and,
at the time of writing, the search is
still on.
In the meantime, the silly situation
remained; I could find nothing wrong
with the IC or its associated circuit
but it would not process the signal. I
sought the assistance of various col
leagues but without any real success.
Some had experienced a similar failure
but it had been traced to a faulty IC.
And they, too, had sought an updated
circuit without success.
In desperation I finally decided to
trace out the extra circuitry on the
1070 board in the hope that it might
provide a clue. And it did. The extra
components were connected to pins
1 and 17 of the IC. There was a 10µF
electrolytic capacitor (C2010) to chassis; a 6.8kΩ resistor (R3010) to plug
1C3; and a 100Ω resistor (R3000) to a
single additional plug, C7.
Hot trails & smelly rats
This plug connection went to a
yellow lead which ran to the VST
board, 1220, via pin 1 of a 3-pin plug
and a mating connector on that board,
marked V17. And this was where I
began to smell a rat. Pin 1 on the VST
board went straight to chassis which,
even with the limited knowledge I had
gleaned of the circuit, didn’t seem to
make sense.
Hot on the trail, I disconnected
the yellow lead, most conveniently
at the 1070 board end, whereupon I
had sound – lots of it and quite uncontrollable. That was both gratifying
and confusing but it clearly indicated
that there was something amiss with
that part of the circuit.
And there was something else I
noted at that time. In both the manuals, the board pattern showed V17 as
a 7-pin arrangement, whereas the set
actually used a 3-pin combination, as
already mentioned. But, with all the
other differences, I didn’t attach much
importance to it.
I also noted that the yellow lead
was anchored at the corner of the VST
board and that it was pulled quite tight
Fig.4: the circuitry for board 1070 in the Philips 02CR035, as shown in the manual. The
actual board used in the faulty set contained extra circuitry and components. Audio goes
into the IC on pin 15 and comes out on pin 11.
March 1996 51
Serviceman’s Log – continued
The result was music to my ears
and this time it was completely controllable. But it was not without some
mixed feelings.
Predominant, of course, was the
relief at having solved the problem.
On the other hand, I muttered some
rude words about those responsible
for the debacle in the first place.
Mental abuse
against the anchor tie to reach V17.
Again, I thought little of it.
But another observation proved
more productive. Apparently, this
board was designed to permit retrofitting of Teletext circuitry and there
were several unused connectors on the
board. And one of these, V19, was a
52 Silicon Chip
3-pin type, sited close to and at right
angles to V17.
Way ahead of me? I’ll bet you are! I
pulled the plug off V17 and fitted it to
V19, noting in the process that the yellow lead was no longer strained against
its anchor point. Then I switched on,
for the umpteenth time.
My colleague copped some of this
mental abuse as well but that wasn’t
really fair. Most of it was directed at
whoever the bright spark was who put
two identical plugs almost side by side
on the same board. It was a disaster
waiting to happen.
With the benefit of hindsight, it was
easy to work out how it had happened.
My colleague had told me that he
had suspected dry joints on the VST
board, which meant that he must have
removed the board for an inspection
and work over. And in putting it back
in place . . .
Talk about Murphy’s law. And
I don’t mean the fictitious leprechaun-like character who lurks in
the corner of the workshop, wrecking
everything we try to do. I mean the real
Murphy. More correctly Lieutenant
Murphy of the US armed forces (I’m
not sure which branch).
As I understand it, Lieutenant
Murphy was in charge of a series
of tests involving acceleration
and impact effects on military
equipment, using a rocket
propelled sled. In one
such test, in which the
equipment was tested
to destruction, the
measurements which
were supposed to be
recorded during the run
and impact were lost.
The reason turned to be – wait for
it – that two interconnecting leads,
fitted with identical plug and socket
connections, had been transposed. I’ve
no doubt Lieutenant Murphy muttered
some rude words of his own but,
for the record, he offered the classic
statement “if anything can go wrong,
it will go wrong.”
I couldn’t agree more. But, again
with the benefit of hindsight, I suppose I should have woken up sooner,
knowing that the board had been
removed and replaced. I wonder if
anyone else out there has been caught
SC
like that?
|