This is only a preview of the August 1988 issue of Silicon Chip. You can view 40 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 "High Performance AC Millivoltmeter":
Items relevant to "Universal Power Supply Board":
Items relevant to "Building the Discolight, Pt.2":
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
|
Sound reasons for confusion
Last month I commented on the rarity of sound
faults in TV equipment, and compared them with
the old fashioned radio sound faults. Well, hardly
was the ink dry on the papyrus - and before I'd
had time to cut myself a new quill pen - than
another one landed on the bench. And it was as
tricky as they come.
The device involved was a
General colour TV set, model
GC187 and the complaint was simple enough; no sound of any kind. I
thought the solution would be
equally simple; which just goes to
show how naive one can be.
The set belonged to a local motel
for which I have the contract for
the whole TV installation. This particular set was one of several
devices which had failed after a
particularly violent thunderstorm.
During the storm there was a lightning strike which probably hit the
antenna directly. Anyway, it knocked out the distribution amplifier,
put several other TV sets out of action, and cooked a microwave oven.
All in all, quite a mess.
Most of the repairs were quite
straightforward, though somewhat
,s rr .... SAt>?
WEL-l-,
"flie:.
FLOAAL-
AR~ANGEME.tJ1""'S
HAD 1-r. ..
expensive and not particularly
newsworthy. As things worked out,
the GC187 was last on the list.
Knowing the capriciousness of
lightning strikes I was quite
prepared to find the fault in almost
any part involved with the sound
channel; ie, it could be anywhere
between the antenna and the
speaker.
At a practical level I started at
the input to the audio amplifier.
This uses three transistors; QB0l,
Q802 and Q803. Q802 and 803 form
a complementary symmetry output
pair, driven by QB0l; a perfectly
conventional set-up. Feeding a
suitable level of audio signal into
the base of Q801 produced a
healthy output from the speaker;
nothing wrong there.
Back-tracking from this point brings us to pin 2 of IC2 (uPC1382C),
the sound IF and audio detector
system. This amplifies and
demodulates the 5.5MHz sound carrier applied to pin 14. This signal
comes from pin 12 of ICl
(T A7607 AP); the video IF amplifier
and detector. And signal into this is
via pin 1 from the first IF amplifier,
Ql and the SAW filter, FLl.
Considering that the set was producing a first class picture, I felt
that the fault was most likely to be
closer to IC2 than the front end.
Colour bar signal
''"--t nus PA~1"1CUl-AR S6,,..
w~s ONE. ~•c.~ HAO FAll..EO APfER A
PA~i,cu1..~Rl.'t V\Ot..EWr' 'O'\UN~S1'"01'M •••
28
SILICON CHIP
I fed a colour bar signal into the
set and checked the output at pin 12
of ICl. As far as I could determine
all was well at this point, so I moved
on to IC2 and gave it a full voltage
check. Again I found nothing
suspicious, all values being acceptably within tolerances.
Next I used the CRO in an effort
to check the 5.5MHz level at pin 14
of IC2. Results were inconclusive
except that the level seemed to be
well down on what I expected. So
what could cause this? ICl was obviously delivering adequate video
I UNtK
.,.
VHF/UHF
CA-335S
TP15
IF A
[IBJ
JJK
TA 760
VR1 SK(B l
VB
UI
vsw
E
TU
+ 15Y
... ,,..
:HAN NEL
;ELECTOR
!OARD
R672 100K
ED
C673
o.o,,
l
CUI
2200P
I
IC670
uPC574J
ca10
1200P
..
05
y
IF SOUND
.,-
-
BOARD
~
•uo10
OUT
-
+
~AIN
BOARD
- ISV
CEC:ill
0670
EQAOI
-15Rl
~
'f
'"
C661 _
0.01 -
C672
-1
J,. ,oisev1F
rl,-
R667
1a1<
( 1/2)
Relevant portion of the GC187 circuit. The audio stage is at bottom left, fed from IC2 to its right, which in turn is
fed from IC1 at top right. The video line goes off to the right from pin 12 of IC1.
and chrominance (colour) signal to
the rest of the set. Only the 5.5MHz
component seeming to be missing.
So was something in ICl, or even
earlier, upsetting the response
curve to the point where the sound
plateau had been lost?
I didn't think this was very likely
since the chrominance carrier was
obviously intact. Still it had to be
considered. I had a spare IC2 on
hand, and it was an easy job to fit it,
so I went ahead. Well, as Thomas
Edison would have said, I now knew
one more thing that wasn't wrong,
because it made no difference.
So where to now? Still thinking in
terms of front-end distortion of the
response curve, my thoughts turned
to the SAW (surface acoustic wave)
filter. It would not be the first time I
had encountered a SAW filter
which upset the response curve,
though the effect is usually less subtle than this. Still, there is always a
first time.
Again it wasn't a difficult point to
prove. I had a spare filter on hand
and it took only a few moments to fit
it. And that was another victory for
the Edison philosophy; I now knew
of two things that weren't wrong.
More importantly, it now seemed
obvious that the fault was not at the
front end or in ICl, but somewhere
from pin 12 of this IC onwards.
I made some attempt to observe
the 5.5MHz level at pin 12, in order
to compare it with that at pin 14 of
IC2, but results were inconclusive.
There was some loss between the
two but this could have been normal, considering that there were
several components in the signal
path. On the other hand, one of
these could be faulty.
Fortunately, most were easy
enough to test and a lk0 resistor, a
39pF capacitor and an 18µH choke
all proved to be OK. That left only
the 5.5MHz filter, CF2. This
presented a problem, in that I had
no spare on hand to substitute. I
could have ordered one but that
would mean a few days' delay, and
I was anxious to get on with the job.
This prompted a rather wild
idea. There was another filter, CF3,
in this part of the circuit (pins 9 and
10) and by all that I could determine
the two were identical. What would
happen if I swapped them over?
Granted, I could hardly expect it to
solve the problem. If CF2 was faulty
I would be simply shifting the fault
from one place to another. But I
reasoned that it would at least
change the symptoms and help confirm whether or not CF2 was faulty.
It was a simple operation and in
a few minutes I was ready for the
big test. But I was hardly ready for
the result. The set came on with a
perfect picture as before but also
with full sound; not a thing wrong
with it. But what did it mean?
Frankly, I wasn't sure. The most
likely possibility seemed to be that
(a) CF2 was faulty and (b) the filter
in the CF3 position was more of a
technical nicety than a necessity,
such that a faulty filter in this position wasn't all that serious. But
even as I formed the theory I had
AUGUST 1988
29
SERVICEMAN'S LOG
doubts. I couldn't shake off the feeling that it was just too glib an
explanation.
Dead silence
I left the set running while I
pondered over all this, and tried to
decide whether I should order up a
new filter anyway. This went on for
about half an hour, while I pottered
around tidying up the bench and attending to routine matters, then
suddenly there was dead silence;
the fault was back just as it was
before.
So what did this mean? In one
sense I was almost relieved. It
meant that the rather nebulous and
unsatisfactory situation I had been
trying to resolve was a furphy.
There was now a chance that the
fault could be nailed more positively, and with less chance of a
rebound.
But I still had to nail it; and that
meant starting all over again. Considering what had already been
done, I felt that about all that was
left was a more or less brute force
approach; to check every component around IC2, for I was convinc-
ed that it was in this area that the
trouble lay. In fact, there weren't
all that many components involved,
so it wasn't such a big deal.
The few resistors were easily
checked, in situ, and all were
within specs. Then there were the
capacitors. Prime suspects were
three low-value electrolytics; C23,
C24 and C27. These were lifted at
one end and measured for capacitance and leakage but again, all
appeared well.
That left only a few ceramic
types and statistically, these seemed unlikely suspects; ceramic
capacitors in Japanese sets are extremely reliable. I started by removing each capacitor and checking it
on the capacitance meter. But then,
since the fault was obviously inter-
r
SPURIOUS
IY\A&NE-,l'SIV\
~ SHAD0WMA59
30
SILICON CHIP
·
mittent, I decided to replace each
one anyway, regardless of how it
tested, just to make sure.
This achieved nothing initially,
the set remaining dead as each one
was replaced. Finally, there was
only one left - C30, a .OlµF from
pin 13 to chassis. It tested OK along
with all the rest but when I fitted a
new one in this position, the sound
came forth at full voloume. Had I
really found it? I re-fitted the
original capacitor and again I had
full sound, so I simply let the set
run.
This time I was lucky. After
about 15 minutes the set went dead
again. I had hoped for such situation and was ready for it. I bridged
C30 with the other .OlµF whereupon the sound came good. More
importantly, when I took it away the
sound vanished. So that was it.
I don't think there is anything
more satisfying than pinning down
a fault in circumstances like this. If
any more proof was needed it
emerged when I pulled the
capacitor out and put it on the
checker again; this time it showed
zero capacitance.
Which was the end of the story
from a practical point of view; the
set went back to the customer and a
recent check confirmed that it had
not missed a beat since. But a couple of questions remained unanswered. For one thing, I'm not
clear as to the role of C30, or why it
is so critical, except that it is
associated with the limiter circuit.
Which only emphasises how difficult it is to keep pace with all the
circuit design theory these days.
There is also the question as to
why the set failed in the first place.
Was it really the result of the
lightning strike, or was the timing
purely coincidental? We will never
know but I'm inclined to support the
lightning strike theory, with the
rider that the capacitor was probably a bit dodgy in the first place.
Lightning strikes
I suppose this raises the old
query as to just why lightning
strikes behave the way they do.
Why do they destroy one component in the middle of a circuit, often
far removed from the antenna and
front-end circuitry, while leaving
the rest of the equipment intact?
While there is no simple answer
to this question, we do know a lot
more about the effects of lightning
strikes than we used to. There are
really two major mechanisms by
which lightning causes damage.
The first is the obvious one where
the lightning actually strikes some
object, such as an antenna, and
causes quite catastrophic damage,
often melting the antenna elements
and associated cables, coils and
other components.
The second effect is more subtle
and is due to the intense magnetic
field created by the massive current flow which accompanies such
a strike. This magnetic field can induce damaging currents in electronic components, solid state
devices being particularly susceptible.
f was actually reminded of this
phenomenon, and just how intense
the magnetic fields can be, by
another case which came up shortly after the aforementioned lightning strike. The customer lives about
TETIA TV TIPS
Philips CA826 (Toshiba chassis)
Symptom: Raster narrowed
unevenly but more so at top of
screen. Picture hiccups and set
occasionally cuts out. 11 2V rail
down to only 90 volts.
Cure: 0801 (2SC1195) open circuit between base and emitter.
This transistor is the power series
regulator, in parallel with 25W
resistors R81 0/R811 . When it
fails, the resistors will pass sufficient current to keep the set going, though obviously not very
well.
50 metres from the motel and he
called me a couple of days after the
storm with a complaint about his
AWA Mitsubishi TV set. As he
described it, the colour had "gone
funny".
It was a large set (63cm) so I
made it a house call. And when the
owner ushered me into the lounge
room where the set was running, I
needed only one glance to know
what was wrong. It was one of the
worst cases of purity error I had
seen for a long time.
Turning the colour off left the
screen displaying all the colours of
the rainbow in a psychedelic pattern. But before I had time to ask
the obvious question the owner
volunteered the information that
this fault had appeared immediately after the lightning strike.
Since several days had passed
since the strike, and the set had
been in constant use, I was a little
surprised that the normal degaussing system had not corrected the
condition, and even considered the
possibility that this had been knocked out by the strike. But a quick
check confirmed that it was still intact, meaning that the spurious
magnetism in the shadowmask was
far too strong for it to correct.
Fortunately, a couple of passes
with the old faithful degaussing
wand wiped out this latest example
of modern art, and made the set
suitable for the more mundane art
form for which it had been
designed.
But the real point of the story is
to emphasise the strength of the
magnetic field which must have ac-
companied the strike, remembering
that there was a good 50 metres
between the point of strike and this
TV set.
So don't be surprised at anything
that happens when lightning
strikes.
One day at a time
And here's another contribution
from my colleague, J.L., in
Tasmania. He calls it "One Day at a
Time" and, as this implies, it
describes a typical day in the life of
a serviceman - or as near to
typical as one can get. This is how
he tells it.
Began the day by collecting an
old Akai video recorder from a
customer's home, then made some
money disappear into my bank account. I hate those state and federal
bank taxes. Sales and income tax I
have learned to live with - well,
more less - but not FID, SDT and
BAD!
Then it was across the river and
up a steep hill to a house with the
best view in Australia. I was hoping
for a lengthy repair so I could spend
some time admiring the scene. Unfortunately, I had only to put a new
fuse in an old Rank colour TV set.
Ten minutes and I was on my way,
back across the river to the
workshop.
On the bench was a G.E. colour
TV set awaiting attention. It was a
fairly common problem - R731
burnt up. This is a fusible resistor
in the 120V line to the horizontal
driver stage. A new resistor lasted
10 minutes, so I assumed it was probably a protection circuit fault.
This protection circuit, TR56 and
its drivers TR57 and TR58, cause
more trouble than enough by going
over-sensitive. They tend to trip for
no reason.
The easy solution is to remove
TR56, which is the crowbar switch.
This invariably cures the problem
and R731 never fails again. Nor
does anything else that I have ever
encountered. But first it is
necessary to check that the 120V
rail and the EHT are not running
high. They never are but they must
be checked before it is safe to
remove the protection.
(Some Rank Arena sets also have
one of these "protector" circuits
that often fail for no reason at all.
AUGUST 1988
31
SERVICEMAN'S LOG
Whether they are ever called on to
actually protect something is a
moot point)!
The next job was an old National
colour TV set and a new National
VCR, both from the same customer.
The TV set was supposed to spit
and splutter whenever a commercial came on the air. It was
reported to be OK on ordinary program material, and ABC and SBS
programs.
All of which only goes to
demonstrate its good taste. I know a
lot of viewers who behave in much
the same way - including spitting
and spluttering at some of the
tasteless commercials they try to
bash our ears with.
The VCR was brand new and had
never been tuned to the local channels. Nor could the owner find any
combination of buttons that would
allow the machine to play a prerecorded tape. I have often complained about incomprehensible
service manuals and now I will add
user manuals. This National
manual was written in tolerable
English but omitted 50% of the info
needed to enable the user to set up
the tuner and timer.
I was able to get things working
because I have done it before but a
non-technical owner might have fiddled for years and got nowhere.
The number of buttons and the permutations of them is staggering. (In
recent months I have had dozens of
calls from people prepared to pay
me to come to their homes to teach
them how to use their new "'what's
it". Often I have nearly as much
trouble as they do, sorting out the
user manual).
In this case the TV set was badly
off tune on all channels and this
was easily fixed. The VCR was not
so easy to tune but eventually
responded to some gentle prodding
and twiddling. After this·, the picture was perfect on all channels
and also off tape. I let the set run all
day but it showed no signs of any
other fault, so the owner can collect
it tomorrow.
Thorn 3500
The next set in line was an old
Thorn 3500 series. It had no blue in
the picture, due to total lack of
voltage on G2 in the blue gun. This
can be caused by an open circuit or
high resistance feed resistor, a
faulty adjustment pot, or a faulty
beam switch. The last is the least
common, but also the hardest to
find as the switch tests perfect in or
out of circuit. In this set it was
switch trouble.
It seems that the switch develops
a leakage path at the high voltages
at which it normally operates but
tests OK at the modest voltages normally used for measuring. Replacement switches are now unobtainable and, as the switch is only
needed for occasional purity adjustments, I prefer to remove it entirely and fit a jumper in its place.
Philips K11
The next problem concerned a
Philips Kl 1 with intermittent pincushion distortion. The problem
came and went at odd times; from
seconds to minutes to nearly an
hour. When the fault was present
no amount of thumping or bumping
would alter it, so there seemed to be
little chance that it was a dry joint.
I tried freezing the E-W correction transistor, TS421, when the
problem showed and again when it
vanished, but results were inconclusive. The spray did seem to
have some effect on what little
.°c,~
"o/.
"f' ,,.
~ o"•
C) --f"
I'
tt
q
p
~~":
:
r..:
. ..
-y~~~~-·:·-...__,
:,.,.. , ___,,
- -- -- . . ----- '
·
~
-------~-~-~ffw:
-· So l>Orv'-r a£ SUR?R\SEJ:> A"T' AN"<-r\-\\tvG
~
~
.
~-,,
'THAT HA'PP&NS WHE.N l-\GHTtJIN6 ST'R\~~4S.
32
SILICON CHIP
------------
'
I wasted some time on the Akai
VCR because it was unfamiliar
gear. The time I lost on this one will
be made up the next time I see one
of these machines - it will go
straight to my colleague and no
messing about.
Bread and dripping
--o~
L
t(f._ ·--..
1\-l&: TV S£.T WAS SOPf>OS~D
iO SPI, IN 'DISGUST
WHENE.VE.R A COMMERCIAL.CAME. -ro A\R .....
behaviour pattern there was, so I
changed the transistor just to see
what would happen.
In a sense nothing happened,
because there was no distortion
when I switched back on, and there
has been no sign of it since. After
several hours the set was still working perfectly, so I felt it was safe to
consider the fault cured. So,
another one down and one to go.
Akai video recorder
The Akai video recorder mentioned at the beginning was an old toploading machine. All I had been told
was that it didn't work. It was left
to me to find out in what way it
didn't work. I fired it up and was
rewarded with all the expected
lights on the front panel. I pressed
the eject lever and received an
unenthusiastic response from the
cassette carrier. It came up about
one centimetre and then stopped. It
took some gentle pressure to
release it from whatever was
holding it.
I put in my test cassette and closed the machine, whereupon the
drum motor started and nothing
would stop it. Normally the drum
does not start until the play lever is
pressed, so this looked like some
sort of clue.
Unfortunately, this was of little
value because the cassette carrier
was locked down (or jammed) and I
could find no way to get the cover
off to release it. Without a manual
and little detailed knowledge of this
particular model, I was out of my
depth. It will go into a specialist colleague in the morning.
By this time it was too late to
start another job so I spent what
was left of the day doing the books,
trying to make the income bigger
than the outgo. This has been an
easy day; Apart from the VCR, each
job was done fairly quickly because
past experience has taught me
where to look for these faults.
Well, that's J.L.'s story of a
typical day, and it emphasises a
point that I have made in these
notes many times; elusive intermittent faults and similar nasties may
make interesting reading, and even
boost one's ego - at least in
retrospect - but they don't put the
bread and dripping on the table (Oh
dear, how sad, never mind. Ed.).
That comes from the routine, day to
day faults , which one takes in one's
stride on the basis of previous
experience.
And there is another point in
J.L.'s story worthy of comment; his
complaint about the inadequacy of
user manuals. I couldn't agree
more. At one time customers were
advised: "If in doubt, read the instructions." Now, the more appropriate advice would be: "Find
someone to show you how to
operate it, then read the instructions to see if they make sense".
That may sound cynical, but it's
true. It's not that the manuals are
inaccurate but simply that they are
inadequate. They are apparently
usually written by people who are
very familiar with the device and
its workings - people who simply
cannot understand the confusion it
is likely to cause the consumer the
first time he encounters it.
I have no doubt that their
defence - if ever one should be fortunate enough to confront them concerning any inadequately explained
point - would be, "Oh everyone
knows that". Carried to its logical
conclusion, that means that we
don't need a user manual at all;
which is close to what is happening,
in effect.
So it is not until the user has
found someone who has already
been through the exercise the hard
way, and is able to instruct him,
that the manual begins to make
sense.
Which makes it all a bit pointless,
really. See you next month.
1§::l
AUGUST 1988
33
|