This is only a preview of the August 1993 issue of Silicon Chip. You can view 27 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 "Low-Cost Colour Video Fader":
Articles in this series:
Articles in this series:
|
SERVICEMAN'S LOG
Little things can be big time wasters
Frustration & annoyance is the theme of this
month’s notes. It’s about the many & varied
things that can contrive to slow the job & which
collectively can add up to a lot of wasted time.
As most readers will appreciate,
most faults that turn up on the service
bench are fairly routine. Generally,
they are faults that have been seen
before in the same model set or faults
which, by their very nature, can be
recognised in any set.
These are the ones that provide the
bread and butter – and a smidgin of
jam occasionally. The real stinkers –
the ones that call for a lot of patience
and electronic detective work – may
make it into these notes but seldom
earn much income.
In between these two extremes are
those which, in spite of being relatively straightforward in a technical
sense, can be quite frustrating – often
for all kinds of silly reasons. This is
one such story and it has a very silly
twist in the tail.
It is about a Panasonic model 2970V
TV set, a 73cm model with stereo
40 Silicon Chip
sound. This set is fitted with the M15D
chassis which has been very reliable.
Incidently, the “D” in this type number indicates a “dead” chassis; ie,
one which is isolated from the mains.
There is also an M15L model, the “L”
signifying a live chassis.
This set is only about three years
old but it had been used for only
part of that time. The owner had
just moved into a new home and the
set has been in storage for about 15
months while he and his family were
living in smaller temporary accommodation. Obviously, the storage period
may have had something to do with
the problem.
The owner was typically vague
about the nature of the fault, saying
only that the set was still working
but that there was something funny
about the picture. This was confirmed
when I set it up on the bench. Yes, the
set was working and, yes, there was
something funny about the picture; it
was suffering from severe pincushion
in the east-west mode; ie, it bowed
inwards on each side.
This was one of the first frustrations, because it is quite a rare fault
these days. Worse still, it was one I
had never encountered before in this
model chassis. I wasn’t quite sure
where to start.
Another frustration involved the
complexity of the service manual.
While the material in it is very well
presented, the circuit and other data
are spread over many sheets, making it
difficult if the circuit has to be traced
from one sheet to another. On the other
hand, it is much better than having the
circuit reduced so much that essential
detail is lost.
Getting back to the fault, the pincushion circuitry is on a separate
board – designated the K board – and
this same module is used in several
models. Initially, I was unable to find
the circuit and wasted a lot of time
searching for it. I eventually ran it
to earth in the section for the M15L
chassis (of course – where else would
it be?). It is reproduced herewith and
at least I don’t have to apologise for
the quality.
But this only moved me on to the
next stage of frustration. The circuit
did not carry any waveforms or even
any voltages on the transistors. I did
eventually track down the voltage
data – on yet another sheet – but not
before I had wasted more time trying
to rationalise the voltages as I found
them. I also searched through the manual for any explanation as to how the
pincushion circuit functioned, but in
vain – I was on my own.
So, all in all, I wasted a good deal
of time before I even started.
The K board
The K board is about 150mm long
by 100mm wide and slides vertically
between two rails mounted on the
right-hand side of the cabinet (as seen
from the back). It connects to the rest of
the set via two plug and socket assemblies, K1 and K2, on leads long enough
to allow the board to be removed and
worked on while still in circuit. This,
at least, was a plus.
Pin 1 of K2 connects to the 113V
main HT rail, while pin 2 carries
vertical pulses. These were traced
back to a network connected to pin
2 of IC401, the vertical output chip.
Pin 3 has no connection and pin 4
goes to chassis.
Plug K1 connects directly to the
horizontal scan coils via pin 1 (H-) and
pin 3 (H+). Pin 4 connects to chassis.
The horizontal amplitude here is quite
substantial.
The lower part of the circuit shows
three transistors: Q701, Q702 and
Q703. Q703 is the first one in the chain
and is fed with vertical pulses from pin
2 of plug K2. The output from Q703, at
its collector, goes via the pincushion
Fig.1: the K-board (pincushion) circuitry for the Panasonic TC-2970V. Vertical
pulses come in on pin 2 of plug K2 & are fed to the base of Q703 (lower right).
Q703 drives the pincushion control, the output of which then drives Q702 &
Q701. Finally, Q701’s output is coupled to the horizontal scan coil circuitry to
provide the necessary pincushion correction. Note the absence of waveforms
and transistor voltages.
control to the base of Q702 which, in
turn, drives Q701.
Q701 is a power transistor (TO66 package) and is mounted on a
heatsink. Its output is coupled into
the horizontal scan coil circuitry to
provide the necessary pincushion
correction.
Waveform checks
I tried checking various waveforms,
hoping I might find an obvious dis
crepency, but without success. The
vertical pulses appeared to be making
their way through the chain OK but,
without any waveforms for reference,
I had no way of knowing whether the
amplitude and waveform shape were
correct at every stage. The only hint
was that the gain of Q703 was not what
I would have expected from a superficial assessment of its configuration.
But then, I couldn’t be sure.
Also shown in this part of the
circuit is width control R708 (5kΩ)
and pincushion control R710 (20kΩ).
I tried adjusting the width control
and this behaved as expected; it varied the width and nothing else. But
when I tried the pincushion control,
August 1993 41
circuit. But no, it was spot on value.
Next, I lifted C707 and measured it.
It was down to around 700µF, which
made it bad enough to need replacing,
even if it wasn’t the main fault. And it
wasn’t, because a new one made little
difference.
My next stop was C708 (47µF) and
this was where I struck oil; it was
extremely leaky, which could easily
account for the weird voltages and
the failure of the pincushion circuit. And it did, because a new one
immediately cured the pincushion
problem.
Having located the fault, I checked
the voltages again, more or less as a
matter of routine. I didn’t refer to the
manual list this time, having memorised the values well enough – I
thought – to satisfy such a check. And
so it seemed; I measured 1.85V on the
emitter, 2.5V on the base, and 9.3V on
the collector, near enough to the figures
I recalled from the manual.
SERVICEMAN'S LOG – CTD
The final twist
it behaved in a less logical fashion; it
also changed the width and nothing
else! That suggested that the fault was
in this section of the circuit.
Transistor checks
My next step, was to check all
three transistors but, as far as I could
determine, all were OK. I had not at
this stage unearthed any voltage data
for these transistors but I made a few
voltage measurements anyway, hoping
that they might provide a clue.
And they did, in a way. The voltages
on Q701 and Q702 at least seemed
reasonable, by rule-of-thumb guess
timation. But Q703 was another
matter; unless it was being used in
a very unusual way, I couldn’t make
any sense of it. I measured 17V on the
collector, 17V plus on the emitter and
17V on the base. But, while this didn’t
make much sense, it did remind me
of the apparent low gain of this stage.
But what should the voltages be?
I found them listed quite by chance
when, as so often happens, I was
42 Silicon Chip
searching the manual for something
else. At a quick glance I registered that
those for Q703 were not only nothing
like the values I had measured but
seemed to be much more reasonable.
All of which simply confirmed my
idea that whatever was wrong was in
the immediate vicinity of Q703, the
transistor itself having already been
cleared.
There aren’t many components directly associated with Q703. I started
with R712, thinking it might be open
COMMON TEST POINT VOLTAGES
E
B
C
Q701
0.025
0.62
13.8
Q702
11.9
11.3
0.62
Q703
1.85
9.3
2.5
Q802
0
0.01
16.5
Fig.2: this is the relevant portion of
the transistor voltage table from the
Panasonic TC-2970V manual. The
collector & base voltages shown for
Q703 are transposed.
And so, after a routine check and
adjustment, the set was duly returned
to the customer, putting an end to my
time-wasting frustrations. Or so I imagined. My final frustration came as a
nasty twist when I later took a second
look at the voltage table in the manual.
It was then I suddenly realised that
the voltages were not listed as I had
recalled them. Oh, the values were correct but not the transistor connections.
The manual listed them as 1.85V on
the emitter, 9.3V on the base and 2.5V
on the collector.
I did a double take on that. Those
figures did not make sense and, had
I been more observant, I would have
realised this when I first saw the table.
Instead, I read them as I imagined
they would be, rather than as they
were.
The point about these figures is
that – apart from anything else – they
imply a base-emitter voltage of around
7.5V – an impossible condition according to my understanding of solid
state theory. When I went to (solid
state) school, the maximum voltage
which could normally be developed
across such a junction would not
exceed 0.7V, and would be more like
0.65V in practice.
So what had gone wrong. My immediate reaction was to suspect that the
figures in the manual were a mistake;
that they had been wrongly set out
Fig.3: this diagram shows the front-end circuitry for the High Energy Ignition
System, as published in the May 1988 issue of SILICON CHIP. The constructor’s
problems were at the very front of the circuit.
with the base and collector values
transposed. I spent a lot of time, on
and off, thinking about the problem
and the longer I thought about it, the
more convinced I became that the
manual was wrong. Note particularly
that, if we transpose the base and collector values as given in the manual,
we then have 0.65V across the base/
emitter junction, exactly according to
the rules.
Finally, at the first opportunity, I
rang my colleague in the Panasonic
service department and put the problem to him. It didn’t take him long
to fetch the manual and look up the
circuit and chart. His reply was brief,
to the point: “Ah yes, a typo” (typographical error).
Anyway, that was the end of story
as far as the various problems and
frustrations were concerned, But I
do suggest that anyone who is likely
to be dealing with the M15 chassis,
or the manual, make a note of the
mistake.
Finally, I do have some other
comments on the fault itself. While
electrolytic capacitor failures are not
unusual, I was surprised that one as
large as C708 should deteriorate to this
extent in only a few years. The fact
that the set was stored for so long may
have been a factor, although it should
not have been.
And what about C707? This, I think,
might have been a victim. Rated at only
6.3VW, it had about 17V applied across
it while ever the fault was present. The
wonder is that it didn’t break down
completely.
In addition, there is another electrolytic capacitor in this part of the circuit
– C716 (10µF 50VW). This was also
checked and was found to be down to
about 5µF. It was replaced along with
C707 and C708.
While on the subject of electrolytics,
I find that if one reads lower than its
rated capacitance, by even a small
amount, it is time to replace it. New
capacitors invariably measure higher
than their marked value. If they drop
below that figure, they are generally
on the way out.
Kit projects
To change the scene, but still on the
subject of frustrating situations, I am
reproducing a letter from a reader, Mr
R. S. of North Melbourne, Victoria. It is
not a servicing story in the usual sense,
nor was it a particularly profound exercise, but it is an excellent example
of the problems which can arise from
the supposedly simple job of building
a kit project.
Assuming a well-designed project
and a properly prepared kit, it is reasonable to expect that it will work at
first switch-on (provided, of course,
that the kit has been correctly assembled).
But it doesn’t always happen that
way. And when it doesn’t, kit builders
react in a variety of ways. Some simply
regard the design as a bomb, curse the
designer, and chuck the whole thing
in the garbage bin.
Some strip it down and rebuild it;
a time wasting and usually futile procedure. One enthusiast, in the days
of build-your-own TV sets, stripped
down and rebuilt a complete 17-inch
TV set, in an attempt to cure a relatively simple fault – the picture was
transposed left to right.
More enlightened souls, like our
reader, assume that the design is capable of working and that its failure must
be due to a construction fault –which
is usually the case. They then set about
finding it in a methodical way. Just
how simple some of these faults can
be is shown in this example. Here’s
how he tells it.
The night before Christmas
Some months ago, my son-in-law
to be raised the question of fitting an
electronic ignition system to his motor
vehicle. Because I had built a number
of CDI (Capacitor Discharge Ignition)
and TAI (Transistor Assisted Ignition)
units, I was consid
ered a suitable
consultant.
Although very pleased with the
performance of all units tested, I have
never been able to detect either an
increase in fuel economy or engine
power. Perhaps this is because I always
cleaned and adjusted the ignition system on a regular basis – about every
3000km.
What I have noticed is that the
electronic systems require virtually
no manitenance or adjustment, unless
I disassemble the distributor for some
other purpose.
Since CDI is currently out and TAI
is in, the choice was simple. I did
not experience any crossfire with
CDI but the inverter squeal could be
objectionable.
My last TAI circuit is dated at 1982
but the “High Energy Ignition System”
unit produced by SILICON CHIP in
May, 1988 was available in kit form.
And, as his training was in the field of
electronics, a kit was purchased and
he assembled the unit.
A few days before his initiation to
son-in-law, he invited me to install
the TAI in the vehicle, as he knew
that I would be more familiar with the
automotive side of things (self-trained
also). However, I put this off while he
August 1993 43
SERVICEMAN'S LOG – CTD
was being moulded into married life
during the next two weeks and I waited
for his return.
A telephone call was subsequently received on December 23rd and
arrangements were made to perform
the change over on the night before
Christmas. He had already mounted
the box in the engine bay and all that
should have been necessary was for
me to find a suitable 12V supply and
make the appropriate connections to
the ignition system.
This particular vehicle incorporates
the ballast resistance in the loom but I
was able to find a suitable power supply connection at the fuse panel and
run a wire to the unit. Laying out the
wiring to the coil in a secure fashion
came next. Suitable checking took
place and I felt that we were ready for
the all-important smoke test.
Switching the ignition to the run
position produced no problems but
switching to the start position failed
to produce any fire in the engine.
This enabled me to demonstrate how
easy it was to revert to the old faithful
44 Silicon Chip
Kettering system, if it was required.
We had begun the work in the open
and in dry conditions, but by now a
light rain had become a heavy downpour and daylight had vanished, so we
called it a night. No smart comments
about Melbourne’s weather thanks; we
aren’t overjoyed either!
The next day, my daughter rang
and invited me to lunch with the
family. No great arm twisting was
necessary, as it would give me the
opportunity for further fault finding.
After a pleasant meal, off came the
lid and the investigation began. I was
expecting to find a fault with D5 but
it still behaved as a diode and was
oriented correctly.
Transposed resistors
The resistors were checked next and
it was found that the 22kΩ resistor to
the input (pin 5) of IC1 was 2.2kΩ. A
search was made for a suitable resistor
and we were able to find two 10kΩ
units. My son-in-law has a very limited
stock of components, as electronics is
not his hobby.
I was concerned that the additional
current through the zener diode in
IC1 might be too much for it. A re-test
took place but the engine did not start.
At this point, it was not noticed that
the 22kΩ resistor which should have
wired into the input of IC1 had been
placed in the collector circuit of Q2.
In other words, the 2.2kΩ and 22kΩ
resistors had been transposed.
The unit was removed from the
vehicle and, because there was the
possibility of serious damage to IC1,
I decided to take the unit with me for
further testing and repair. The 22kΩ
resis
tor problem was corrected the
next day. I then proved that Q1 was
intact and with the aid of a spare coil
connected, produced a spark when
the output of IC1 was taken to chassis
using a jumper lead.
Since waveforms were going to be
traced, the CRO was fired up. This was
an overkill, as will be seen. The action
of the points opening and closing was
mimicked by connecting a flying lead
to the power supply chassis. This
produced a step at the 47Ω resistor,
as was expected.
Unfortunately, when the CRO probe
was transferred to the other side of diode D5, the pulse vanished. On turning
the board over to the track side, it came
to my attention that there were more
tracks and holes than were necessary
for this project (as a last resort read the
text fully). The constructor had placed
the anode lead of D5 into the next hole
up the board.
Once this correction had been made,
a pulse could be traced to pin 7 of IC1.
By now reconnecting the test coil and
spark plug, I was able to view a nice
healthy spark. When installed the next
day, the system worked perfectly.
The constructor did emphasise that
he had performed “high reliability
hand soldering” techniques as his
employer instructs. And to his credit
the soldering could not be faulted.
But my warped sense of humour
considers that suitable connectivity
is required before conductivity can
take place.
Fair enough, R. S. and thanks for
the story. It emphasises one very important point – the difference between
field servicing, where a device originally worked but has now failed, and
production line servicing, where the
device has never worked. Production
line servicing is a completely different
SC
ballgame.
|