This is only a preview of the March 1988 issue of Silicon Chip. You can view 34 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 "Technilab 301 Function Generator":
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
|
•
A baffling exercise
It was the Count of Monte Cristo who, prompted by
his creator Alexander Dumas, made the profound
observation that, "Only he who has known the
greatest sorrow can know the greatest joy." After
battling with a particularly stubborn video
recorder recently, I felt the same way.
Fortunately, my first story did not
cause all that much frustration. It
concerns a Toshiba colour TV set; a
model C-1416, 33cm set of some five
years vintage. The job started out
as a fairly routine assignment but
caused some head scratching at
one stage.
It started off with a phone call
from a new customer who, after
describing the make and model of
the set, complained simply that it
had stopped. I tried a few discreet
questions to get some idea of what I
might be up against, but quickly
realised that this wasn't getting
anywhere; the set had stopped and
that, as far as the customer was
concerned, was all there was to it.
All I could do was suggest that he
bring it into the shop.
So it was that he turned up at the
shop a couple of days later and sat
the Toshiba on the counter. I plugged it in while he was there and confirmed that, at least as far as he
was concerned, the set was totally
dead; no picture, no raster, no
sound. But it wasn't totally dead to
my ears, because I could hear the
power supply hiccuping away merrily, suggesting an overload which
it didn't like.
I am not very familiar with this
particular set but there were a couple of points in my favour; I had a
service manual and, when I came to
work on it, I found it a lot easier to
get at than many designs I could
mention. The circuit was well laid
out and easy to follow, and appeared to be a fairly conventional
arrangement. As I said, it all looked
fairly straightforward.
Likely suspects
With symptoms like this the two
areas I first suspect are the power
supply and the horizontal output
stage. The power supply was fairly
typical; a switchmode arrangement
running from a bridge rectifier connected directly to the mains. Much
the same applied to the horizontal
output stage; a horizontal driver
transistor, Q402 (2SC2482), driving
__
PIN 1,
LOT T461
.JLf\_
......,.
/
890VP-P
C464
560pf
C463
.0022
2kV
Fig.1: horizontal output stage of the Toshiba model C-1416 TV set. Note
the protective resistor and diode built into the 2SD896 transistor.
50
SILICON CHIP
the horizontal output stage (Q404)
via a transformer (T401). The
horizontal output stage was a
2SD869, a transistor with built-in
protection (ie, a resistor from the
emitter to base and a diode from
collector to base).
The main HT rail was shown as
114V, which was applied to pin 3 of
the horizontal output transformer,
then from pin 1 to the collector of
the aforementioned 2SO869. That
much digested I decided to measure
the HT rail, which can often provide a clue as to the likely culprit. It
turned out to be well down, around
50V, and I mentally filed this figure
for future reference.
At this stage it was a toss-up
whether to move to the right of the
circuit, towards the horizontal
stage, or to the left towards the
power supply. Fully aware that
Murphy would be lurking around
the corner, to make sure that which
ever way I went would be the
wrong way, I took a punt on the output stage.
Well it seemed that Murphy must
have slipped out for a cup of Irish
coffee, because that was the right
decision.
In greater detail I simply disconnected the output transistor. These
transistors are not the easiest to
test, since they tend to show a low
resistance from collector to base,
regardless of polarity, due to the
protective diode. Similarly they
show a low value resistance, ranging from about 400 to 900 according to type, between base and
emitter. Granted, one can allow for
these characteristics but it is usually easier to simply fit a new
transistor.
But, before fitting a new one, I
switched the set on with the
original one removed. I wasn't quite
sure what kind of a HT reading I
would get, but I hoped it would
come up to something near normal,
if the transistor had been faulty. On
the other hand, there was a risk
that it might go high and trigger any
over-voltage protective circuitry.
Once again, it seemed that I had
done the right thing because the HT
rail came up almost spot on the circuit value. It seemed that it was my
lucky day; I should have the job
knocked over in short order. All I
had to do was fit a new transistor
and we should be up and running.
It must have been around this
time that Murphy finished his coffee and came back on duty. I switched the set on, full of confidence,
only to find that it was in exactly
the same condition as before;
power supply hiccuping and low HT
rail. Somewhat taken aback, I
reached for the CRO leads and
made a quick check of the signal
from the driver stage, through the
driver transformer, and to the base
of the output stage. As nearly as I
could tell, allowing for the reduced
rail voltage, this section appeared
to be functioning.
At that point the only logical
thing I could think to do was to take
the new transistor out, check the
HT rail and power supply
behaviour again, and try to decide
what to do next. And that operation
produced surprise number two the rail voltage had risen, but not
all the way; it was now sitting at
around 75V. What the heck was going on?
Smoke signals
I double checked the connections
to the transistor, confirmed that
they were correct, and began looking for any other silly mistake I
might have made. I found nothing,
but had left the set running during
those few minutes. Suddenly my
nose told me that something was
getting hot somewhere and a few
seconds later I pinpointed the
source.
A thin curl of smoke was rising
from a capacitor which forms part
of the output stage assembly; C464,
a 560pF, 2kV unit connected between the collector and the emitter
of the output transistor. It was a
blue plastic encapsulated unit
which had suddenly developed a
brown spot that grew larger as I
watched it.
Well, that was the breakthrough.
I pulled the capacitor out and then,
before refitting the output transistor, turned the set on and checked the HT rail again. And this time
it came up just above the nominated
value, exactly as it had done the
first time. Which I reckoned proved
the point.
At a more practical level I needed a replacement capacitor. My
stocks didn't run to an exact
replacement, but I did find a 560pF
unit with a 3kV rating. This was fitted, the original output transistor
wired back in, and the set given
another try. And this time
everything worked; the set gave
forth sound, the HT rail came up
spot on, and a first class picture appeared as the set warmed up.
So that was that and the set was
duly returned to a happy customer.
But it is worth speculating on what
caused the sequence of events just
related. Fairly obviously, the
capacitor was faulty all along, being unable to withstand the peak
level of nearly 900V generated by
the output stage, and pulled the HT
rail down accordingly.
But with the output stage removed no such voltage was generated,
and all it had to withstand was the
114V from the HT rail. Initially, at
least, it was able to do this but it
must have been on the point of
breaking down completely. My test
run with the replacement transistor
must have been the last straw; by
the time I tested the set without the
output stage for the second time, it
was breaking down at 114V and
went up in smoke.
Which is all very satisfying from
a technical point of view, but most
of my diagnosis was wasted; a few
more minutes running on the bench
would have produced the curl of
MARCH 1988
51
SERVICEMAN'S LOG
smoke and revealed the fault with
no effort on my part.
That's the luck of the game and it
had a happy ending anyway.
A failure to erase
My next story involves a much
more frustrating experience. It concerns a Sanyo beta video recorder,
type VTC 5005, that belonged to a
regular customer. The first intimation of the trouble came via a phone
call from the lady of the house. The
fault was rather unusual in that the
machine would record the video
signal satisfactorily but, on odd occasions, would fail to record the
sound or erase any previous sound
track.
She went on to enquire whether
this was a common fault, whether I
knew what would cause it, and, of
course, was it going to be expensive
to fix. I had to reply that it was a
fault I had not encountered before,
that I could only assess the likely
cause in broad terms, and that it
was almost impossible to estimate
the cost. However, I did promise not
to let costs get out of hand without
consulting her. In the meantime I suggested that she
bring the set in.
So the lady turned up at the shop
a couple of days later with the
machine and, commendably, she
had thought to bring a faulty tape
with it. In the event it didn't help a
great deal, except to confirm the
symptoms she had described, but
even that was useful and I wish all
my customers could be so
thoughtful. I tried the machine
while she was there, but I need
hardly add that it behaved perfectly. This surprised neither of us.
I have not had much experience
with this machine, but I do have a
manual and I fished this out immediately. In particular I concentrated on the section designated
"VD-1 Audio Circuit"; the section
which handles the audio signal taking it from the record/replay
head, or feeding it to it, as required
- and which also incorporates the
bias and erase oscillator.
It was this latter which interested me most because it was
this function that was failing
somewhere along the line. But exactly where was the real question.
It could be in the oscillator itself, it
could be the erase head, or it could
be the flexible lead and plug and
socket which connects the head to
the board.
The record/replay head seemed
an unlikely suspect, in that there
had been no replay problems as
such. And if the bias line to this
head had failed somewhere on
the board I would have expected that the system would
still erase, but fail to record
anything but a weak and
distorted signal. So the
odds were strongly in
favour of a failure within
the erase circuit, or
failure of the oscillator
itself.
:~-. ,.
~
...-......,.~
1\-\~'s ~E LUC.Ko~ "Me: GAME.
& \T
~~b F\
l'\A1)P'f e.rJl>\NG
~N'(W~'l ...
52
SILICON CHIP
~~
But I needed to be sure. And
when you think about it, this isn't
an easy function to monitor. Simply
letting the machine run in the
record mode is of little value, since
any failure will not be apparent until the tape is replayed, by which
time the fault will most probably
have vanished. But, in any case, the
time involved in such an approach
would be quite unacceptable.
As depicted on the accompanying
circuit the erase/bias oscillator
(near the bottom) consists of transistor Q2007, transformer T2001,
and a few minor components. The
secondary of T2001 feeds the erase
head directly from pin 5, and supplies the bias for the record head
from pin 6, via C2027 and tab pot
VR2003. The erase head is connected via plug and socket S2002
and the record/replay head via
S2001.
Fortunately, the circuitry involved is fairly easy to reach. In fact the
audio circuit, as shown separately
in the manual, is really a part of the
complete video/audio board, and
the print side of this is directly accessible when the main covers are
removed. After that, removal of a
few screws permits the board to be
swung up, giving access to the component side.
For a start I set the machine up in
the record mode and connected the
CRO to the active terminal of the
erase head, this being about the
most convenient access point. This
confirmed that the oscillator was
functioning, at least for the present,
so I decided to leave it running in
this way, while I went on with other
jobs, simply glancing at the CRO
from time to time to check what was
happening.
Initially, this approach paid off.
After it had been running for some
time I checked the CRO and realised that the erase signal had vanished. Unfortunately, before I could
make any further tests the system
came good. I wasn't too worried at
this stage; I blissfully imagined
that, since the fault was obviously
in a mood to happen spontaneously,
a little prodding, freezing, or
heating would encourage its reappearance.
I should have known better. I
tried every trick in the book; I froze
every component likely to be involv-
.-----------------------------------■
..---------il---->
R2001
C2009
L?K
I AUDIO IN
I
GNO
0 -L?µ/16
I
R
"'8al
I >-
a:
"'"'
52001
-- H
...
2
:
R-iN~EAD
R-P HEAD
02001 2so1011
SWITCHING
Q2002 2501011
SWITCHING
u
"'"'
2~
3--
h
I
a-
20
22K
R203',
~
Q.1 (0)
a/ ?:
I
(8 -L)
Q2003 2SC945
SWITCHING
-T ,R
r2001
~~µ"
i2
03?
02006 2SC2603
SWITCHING
c~~
'"''N'
REC/ PLAY OUTPUT LEVEL
(0-2)
'
-v
TP2001
,02 1
I --- ,..,____-+____,·,~~<. >"'
R2036" 8.2K
ijg
o,- ~ iL
u
a:
r P12002
R2039
R2038
'OK
10K
-
02004 2SA1016
SWITCHING
g~io1 2SD43B
------------------------------------
Fig.2: relevant section of the audio circuit of the Sanyo VTC 5005 VCR. Note the erase head plug at bottom left.
ed and blasted the whole area with
a hair dryer. I tugged the leads,
wriggled the plugs and sockets, and
prodded every component with
everything short of a sledgehammer. Nothing produced the
slightest hint of a fault. Had I not
known better, I would have been
prepared to swear that there was
absolutely nothing wrong.
All I could do now was to continue the tests as before. And this I
did, for several hours each day for
the next couple of weeks. But not
once did the CRO pattern so much
as flicker, even though I repeated
some of the previous brute force
tactics from time to time. It was a
stalemate.
I would have been happy to leave
things set up for as long as was
necessary, but this would hardly
suit the customer. In fact they had
already made a couple of polite enquiries and I sensed that they were
becoming impatient. Among other
things, I realised that they used the
machine more for playing prerecorded tapes from the video shop
than for recording off-air programs.
The upshot was that I explained
to them what I had discovered so
far [which wasn't really very much)
and that, until the machine elected
to fail again - which it appeared to
be stubbornly refusing to do at this
stage - there was very little
chance that I could make any progress. So it was agreed that they
take the machine away and make
such use of it as they could, until
such time as the fault worsened.
Weeks went by, then several
months, before I saw them again.
Then the lady contacted me with a
completely different problem and I
took the opportunity to enquire
about the recorder. "Oh, it's going
fine", she replied, "whatever you
did seems to have fixed it."
I know the feeling: after that
length of time I could almost kid
myself that I had done something to
fix it; almost, but not quite. Deep
down I knew it was only a matter of
time before the gremlin would
strike again.
But more months went by and a
couple of discreet checks on my
part produced the same answer,
"It's going fine. " Then came the
day when the lady was on the
phone with a tale of woe. "The
recorder is really playing up now. It
is almost impossible to record
anything." This was the best news I
had heard about the machine so far
and my enthusiasm probably showed when I suggested she bring it in
immediately. She lost no time in
responding.
Before disturbing anything I loaded a tape into the machine and
made a brief test recording. Sure
enough, the fault was there. So,
very gently, I removed the covers
and connected one of the CRO probes to the active lead at the audio
erase head. This confirmed that the
fault was still present and I reached for the second CRO probe with
the idea of checking progressively
along the board.
At which point the system suddenly came good and there was no
more I could do until it decided to
play up again. Fortunately I didn't
have to wait very long. When it failed this time I very gently removed
the necessary screws and lifted the
MARCH 1988
53
SERVICEMAN'S LOG
board so as to provide access to the
component side. This didn't disturb
anything and the fault remained.
I reached for the second CRO
probe and approached the print
side of the board in the vicinity of
the two pins which mate with
socket S2002, and which carries
the leads to the erase head. By just
touching the board, and before I
could make an electrical connection, I cleared the fault.
Suspect plug and socket
I immediately suspected the plug
and socket assembly, and this seemed to be confirmed when I wriggled
the lead and the plug and found
that, by stressing the assembly in a
certain way, I could make the fault
come and go. At last it looked as
though I was getting somewhere.
In this setup the plug on the lead
is actually the female connection,
the male contacts being two pins
soldered into the board. I was
specially suspicious of the female
contacts, particularly where they
made contact with the cable. This is
a crimped connection and it is not
54
SILICON CHIP
unusual to find the crimping does
not penetrate the insulation properly, resulting in an intermittent
connection.
Removing contacts from plugs of
this type is a little tricky. They are
held in by a small tongue or barb
punched into the contact, and
which is depressed when the contact is inserted into the plug. It springs up when the contact is fully inserted and effectively locks it in
place.
To remove these I use a long,
thin, pointed probe with which to
TETIA CORNER
Thorn 9904 (Q Chassis)
Symptom: No luminance . Sound
and colour OK. A normal picture
appears briefly if the set is switched on again quickly after switching
off. 12V rail reads high at 15.5V.
Cure: D231 (EQA01-12S) 12V
zener diode open circu it. This
diode sets the 1 2V rail and the
higher voltage when it fails blanks
the video output from IC201 .
depress the tongue and thus allow
the contact to be withdrawn from
the back of the plug. With both contacts out I examined them carefully.
As far as I could see they were
quite OK, but I also checked them
with an ohmmeter. Again they
seemed faultless, even when the
wires were vigorously tugged and
wriggled.
Nevertheless, I put them through
my own "extra crimping" process.
This involves applying pressure to
the stem of the contact, in the valley
where it is supposed to punch
through the cable insulation. To do
this, I use a pair of cutters, with one
blade lying the length of the valley. I
know it sounds drastic, and it certainly looks risky, but it is merely a
matter of applying a judicious
amount of pressure. And it does
work.
But it didn't help much in this
case. I re-assembled the plug, fitted
it back on the board and gave the
whole setup another wriggle test. It
behaved exactly as before;
pressure on the plug or tension on
the cable could make the fault come
or go. And, since I felt that I had
cleared the plug, the next likely
suspect was the pair of pins on the
board.
Conned
I examined the joints where these
pins were soldered to the board,
even though I had already been
over them once before, but could
see nothing suspicious. Nevertheless I resoldered them, just to
make sure. But this achieved
nothing either and I was forced to
the conclusion that there was probably nothing wrong with the plug,
socket, and cable assembly; that I
had been conned by their sensitivity
to pressure.
And I had been conned in more
ways than one. It had seemed so obvious that this was where the fault
lay that I hadn't even bothered to
make the other obvious check
which I had set out to make at the
beginning; whether the oscillator
itself continued to function when
the waveform vanished from the
erase head.
I quickly made amends, connecting the second CRO probe directly
to the oscillator circuit. Then I
wriggled the plug and socket
assembly again, created the fault,
and established that it was the
oscillator that was failing, not the
circuit to the head, in spite of the
symptoms.
That much established, the next
likely possibility seemed to be a
faulty component in the oscillator
circuit and, by now, I would have
been quite happy to replace every
component in this part of the circuit
if it produced a quick cure. After
all, there was only one transistor
and a handful of resistors and
capacitors. The transformer,
T2001, was about the only special
item.
But before taking that drastic
step I decided on another freeze,
heat, and bash routine. After all,
the thing was much more touchy
now than it had been when I tried
this before, and it might just work.
And so I set to, with the machine in
the record mode, both CRO probes
connected, and the fault condition
evident, this having been achieved
by much wriggling of the aforementioned erase head plug.
I drew a blank with the freezer,
and similarly with the heating, so I
reached for the sledge-hammer actually the butt end of an insulated
alignment tool - and began prodding. Nothing happened until I
came to C2029, a 100µ,F 16V electrolytic, when the lightest touch
caused the oscillator to come good.
Which was very encouraging, except that no amount of additional
prodding, freezing, or heating of
this component could reverse the
procedure. So what did it mean?
Was the capacitor faulty, or was
this another furphy like the erase
head plug where, apparently,
vibration and pressure was being
transmitted to the fault somewhere
nearby on the board?
In any event, it seemed logical to
remove the capacitor, check it as
thoroughly as possible and, if any
suspicion remained, replace it. But
it didn't come to that because, as I
unsoldered one of the lugs, the
solder came away much too readily
and I was convinced that I had uncovered a dry joint, one that had
defied my visual inspection.
Closer inspection of the lug confirmed my suspicion. It had been
tinned during manufacture but I
was convinced that the solder on
...
the board had never really "wet" it.
Most likely it was a "cold" joint;
one where the temperature of the
lug had never reached the melting
point of the solder.
I measured the capacitor as a
matter of course, and it was well
within tolerance. I cleaned the lug,
tinned it again, put it back on the
board, and made sure it was well
and truly soldered to the pattern.
Then I checked the whole system
again.
I wasn't really surprised when it
came good immediately, because it
had done that many times. But I
was gratified to find that wriggling
the erase plug no longer had any effect. Considering how touchy it had
been before, I felt it was very
significant.
But there was more to it than
that. The capacitor's position on the
board was slap alongside the two
pins which mated with the erase
plug. So wriggling the plug could
easily have aggravated a dry joint
at the capacitor lug.
I ran the machine for several
hours a day for a week or so after
that, and it never missed a beat. Of
course it had done that before too,
so I couldn't be one hundred percent sure that I had fixed it. The only thing I felt sure about was that I
had found a dry joint, even if the only evidence I had was my own
observation when I unsoldered it.
And so the machine was returned to the customer, with a strong
emphasis on the need to contact me
immediately should the fault show
again. Thafwas several months ago
and a couple of check calls have
confirmed that there has been absolutely no sign of the fault.
Only time will tell though. I'm
keeping my fingers crossed.
~
RCS Radio Pty Ltd is the only company which
manufactures and sells every PCB & front panel
published in SILICON CHIP, ETI and EA.
651 Forest Road, Bexley, NSW 2207
Phone (02) 587 3491 for instant prices
4-HOUR TURNAROUND SERVICE
MARCH 1988
55
|