This is only a preview of the December 1988 issue of Silicon Chip. You can view 44 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:
Articles in this series:
Articles in this series:
|
The customer has no idea
I often wonder whether the customer ever has the
slightest inkling of what a serviceman goes through
in tracking down some of Murphy's most vicious
creations. But even as I ponder, I know the answer:
the customer's thinking is light years away, across
a chasm that can never be bridged.
And what brought on that little
burst of philosophy? Answer - one
of the most frustrating and least
satisfying exercises I have experienced for a long time. It included just about every negative aspect
one could imagine.
Quite apart from the technical
problems, there were difficult
customer relations. These involved,
among other things, a reluctance to
reveal the set's previous history
and a threat to bring consumer affairs into the situation. And I need
hardly add that the whole exercise
was a complete dead loss financially.
The job started out routinely
enough; a phone call from a new
customer enquiring as to whether I
could service her TV set. When I
said I would be happy to do so, she
said it was a Sanyo colour set
which had stopped dead in the middle of a program. Since there didn't
seem to be much point in probing
deeper right then, I simply suggested she bring the set in to the
shop.
When she turned up a couple of
days later the set turned out to be a
Sanyo model CPT 6603, which first
appeared about eight years ago. A
quick check confirmed that it was
quite dead. I made some attempt to
delve into the set's history but the
customer seemed strangely ignorant in this regard. But I also
sensed that she didn't like being
questioned.
When I opened the set the first
thing I found was a blown mains
fuse, a 2A slow-blow type. Based on
experience, I mentally nominated
three possibilities: a faulty degauss
thermistor, a faulty horizontal output transistor, or a faulty switching
transistor in the power supply. Of
these, the last named was the most
common.
The power supply board mounts
vertically on the left hand side of
the chassis (from the back) supported by a metal frame. Releasing
a screw on the main chassis frees
the frame and allows it to be swung
to one side. The power supply
board can then be freed from the
frame, giving access to the copper
side.
The switching transistor, Q361
(2SC1046L), is mounted on this
board on a fairly substantial
aluminium heatsink, about 50mm
square. This, in turn, is mounted
clear of the board by small spacers
but is still close enough to allow the
base and emitter pins of the transistor to pass through the board to
the copper tracks.
Knife & blow lamp
~t1<~
,... .:SU,- 1: AL.~ SE.NSEO SHE.
DION~ L-lKE. ge.1N6 qUE~TID~Et>...
62
SILICON CHIP
It was as I swung the board clear
that I received my first surprise;
something which, with hindsight,
should have warned me that trouble lay ahead. One glance told me
that someone had been there before
me. And unless I was very much
mistaken, he had been wearing a
striped apron and had attacked the
board with a butcher's knife and
blow lamp.
Well, you know what I mean. It
was a mess. Excessive soldering
heat had destroyed the original cop-
. ~Well, to cut it short, it was the
same situation as before; a blown
fuse, a shot switching transistor,
and a couple of faulty diodes. But
why? I would have liked to have
blamed coincidence but deep down
I knew this was clutching at straws.
There just had to be a more fundamental reason.
I went through the motions of
replacing all the components and
soon had the set running again. But
there was no way that I could
return it with the threat of another
failure hanging over it. At the very
least I owed the customer an explanation for whatever time it was
going to take to find the real fault.
So I rang the number and this
time found myself talking to the
lady's husband. And, after giving
him a brief resume of the situation,
I found him to .be as outgoing as his
wife was taciturn. I learned that
the set had already failed in a
similar fashion on three occasions,
and had been repaired by another
serviceman, whom he named.
Well, that explained a lot of
things. I knew the serviceman he
mentioned and realised that I
erence is the 2S0380 which has the should have recognised his hanhighest ratings of the lot.
diwork. He is reasonably clued up
I took the precaution of making technically but is sadly lacking the
the first test at a reduced voltage skills necessary to handle today's
via the Variac, increasing the components.
But it explained more that that.
voltage gradually in case there was
anything else amiss. But everything The fault was obviously a "stinker"
went smoothly and I soon had the that had already defied three efset running at normal voltage.
forts by the other serviceman until,
I ran it for several hours that apparently, the customer became
day, spent some time the next day fed up with the situation and sought
cleaning up the butchery as best I out someone else - Yours Truly.
could, then ran it for a couple more None of this did anything to cheer
hours. Then I rang the customer meup.
(I learned later of a further comand told her it was ready. She called in later, paid the bill - it was plication. The husband and the seraround $70 - and collected the set. viceman were golfing buddies. One
can imagine the conflict that situaAnd that was another one down.
tion created in the husband's mind.
It's not that simple
And what of the domestic situation?
Oh yeah! I wouldn't be writing While I sensed a lack of agreement,
this had it been that simple. A cou- I couldn't decide whether he was a
ple of weeks passed and then the henpecked husband or she a
roosterpecked wife).
lady was on the phone with the
complaint that the set had failed
But right then I had to content
again and, according to her, it was myself with pointing out, as
the same fault as before. I was well diplomatically as possible, that he
aware of the significance of that might have saved a lot of time and
remark but I wasn't going to argue
trouble - and money - if I had
at that stage. I simply invited her to
known the set's history in the first
bring the set in again.
place. And I added that it might
_ _._,
per tracks to the transistor pins and
these had been replaced with short
lengths of heavy scrap wire, clumsily soldered into place. And I could
see that several other components
had been removed for testing or
replacement. There was no damage
here but one could see where he'd
been.
I made a mental note to clean
things up later. I didn't want
anyone seeing that mess after I hq.d
handled the set.
But first I had to fix the fault. It
took only a moment to confirm that I
had picked it in one; Q361 was a
dead short between collector, base
and emitter. This type of failure is
not unknown in this model and I had
seen it often enough before to know
that any one of three diodes D362, D363 and D364 - could also
have been destroyed (sometimes all
three). In this case it was D362 and
D364.
Repairs were simple - fit two
new diodes and a new switching
transistor. In place of the
2SC1046L I use a 2S0350 or a
2S0380. Both have higher ratings
than the 2SC1046L and my pref-
DECEMBER1988
63
SERVICEMAN'S LOG
take some time to find the real fault.
Fortunately, they had a spare set.
Circuit checks
So it was back to the bench. I connected the CRO to the switching
transistor collector, hoping I might
detect some abnormality to account
for what it now transpired were
five transistor failures in a few
weeks. Unfortunately, this only
revealed that the waveform was a
copybook version of that shown on
the circuit.
At this stage, I suspected that
something might be generating the
odd spike or spikes but, unless I
was very lucky, I wouldn't be watching when they occurred.
In the light of this there was little
I could do except let the set run and
hope that when it failed again - as
I felt sure it would - it would provide some clue. In fact, the set ran
for several days with no signs of
trouble. Then, one afternoon, I left
· the workshop for a few minutes and
returned to find it dead.
I pulled the faulty transistor out,
found a couple of the diodes which
had gone with it, and replaced
these as a matter of course. But
there seemed little point in simply
64
SILICON CHIP
fitting another transistor and
waiting for it to fail in a few days. I
had to make some effort to find the
cause.
More in desperation than anything else I decided to examine the
heatsink mounting and the associated connections to the board, in
case an intermittent connection
might be generating spikes. So I unbolted it from the board and
carefully inspected the mechanical
assembly. I drew a complete blank;
there was nothing to arouse the
slightest suspicion.
But the exercise did prompt
another thought; what about the
mica insulating washer between
the transistor and the heatsink?
The best view of it in situ revealed
nothing so I pulled it out for a closer
look through a jeweller's loupe. And
there it was; what I had mistaken
for a tiny spot of discolouration which often occurs in mica - was
in fact a pinhole. It was so small
that I doubt that more than the
point of a pin would have passed
through it.
Well, we seemed to be getting
somewhere at last. I fished out a
new washer, examined it under the
glass just to make sure, then fitted
everything back together. The set
came good immediately and I settled down for another soak test.
Another splat
It was a brief test. The set hadn't
run for more than a couple of hours
when there was a sudden "splat"
and it went dead. My heart sank:
not only had I not fixed the fault but
I had probably done in another
transistor. And these are not the
10-cent variety- they run closer to
$10, even at trade prices.
Closer examination brought some
consolation; the mains fuse had
blown, but the transistor and
diodes were still intact. What was
more, there was a clear mark
where the flashover had occurred;
from the transistor case (collector)
to the heatsink, across the edge of
the mica washer. And it was then
that I realised that the new washer
was smaller than the one it had
replaced.
Both were nominal TO3 washers
but the original had been a good
5mm larger all round than the transistor case, while the replacement
was a neat fit. I must confess that I
was somewhat puzzled by this turn
of events. Was the design so critical
that only an oversize washer would
withstand the normal voltages in
this part of the circuit?
There seemed to be only one way
to find out; fit an oversize washer
and try again. After some searching
I found one; larger in fact than the
original, being close to 8mm larger
all round. So this was duly fitted
and, muttering something along the
lines of "Beat that, you b ... ", I switched on again. And again the set
came up and played normally.
For a couple of hours. Then there
was another splat and this time it
took out the transistor, plus one of
the diodes. And again the weak spot
was evident; the bolt holes in the
washer were marginally larger
than the previous ones and this was
where it had flashed over.
I was feeling pretty desperate
·1:
·ER..
UNIT
I
I
'
C:307
RJO<
560 K
c:i ,o
CJII
'"""' '"""'
TC~'l9
'305
560K
,.,
i
t C387
100., F
.
,
,,,'~
m
-!-
POWER-SUPPLY UNIT
':' ') OEFLl:"CTI ON
Fig.1: the power supply circuit for the Sanyo Model CPT 6603 receiver. An obscure and frustrating fault caused
repeated failure of the main switching transistor, Q361.
and disheartened by now, but there
seemed to be nothing for it but to
heed the obvious lesson and try
again. After a lot of searching I
found a large washer with small
holes, put everything back together
and tried again.
This time things looked a lot more
promising. The set played for
several days, during which time the
lady customer was on the phone
showing obvious signs of impatience. I explained that, hopefully,
the fault had been fixed but that I
wanted to run the set for a couple
more days, just to be sure.
Somewhat grudgingly, she accepted
my explanation.
So I ran the set for a few more
days and then, with no hint of trouble, rang the lady and told her she
could collect the set. And I did explain that I was making an additional charge - about $30 - purely to cover material costs - ie, the
transistors. I would not have dared
to try to charge for my time. She
seemed to accept this and duly collected the set and paid the bill. I
heaved a sigh of relief as the set
went out the door.
An irate customer
It was a misplaced gesture. The
set had been gone only three days
before the lady was on the phone
with the news that it had failed
again. And, understandably, she
was pretty irate. For that I could
hardly blame her. From where she
stood she had spent a lot of money
(including whatever the other serviceman had charged), experienced
a lot of inconvenience, and seemed
to be getting nowhere.
But what upset me was her
threat to take the matter to consumer affairs; in fact, she intimated
that she had already contacted
them. Assuming that this was not
just bluff, and knowing consumer
affairs, I imagine that she would
have been advised to take the matter back to me, then contact them
again if I refused to co-operate.
Since I had never had any such
intention I wasn't particularly worried about the threat but I was an-
noyed. I felt that she might at least
have consulted me first. And I spelt
this out, adding that I was perfectly
willing to stick with the problem, at
no additional cost, until it was solved. She calmed down a little after
that and agreed to bring the set in
again.
Naturally it was exactly the same
fault, except that there was no
evidence of a flashover from transistor to frame. I'd beaten that but
it was now clear that this was only
a symptom. And it was also clear
that I needed help. So I rang the
Sanyo service department and
eventually contacted a technician
who seemed to be familiar with the
fault.
We had a long discussion and out
of it came a list of components
which experience had shown were
likely suspects. These were: C361
(2200pF) in series with R361 between emitter and collector of Q361;
diode D361; capacitors C332 (22µF)
and C351 (220µF), connected to terminal 13 of the converter transformer (T361); and the main filter
DECEMBER1988
65
Problems?
.
SERVICEMAN'S LOG
'
..
' . . •:
... and you
don't have our
I20page
catalogue . ..
.... ..
I-\£~,
~INK
At last . ..
a TRADE
catalogue for
the consumer
ARISTA ... your one-stop problem
solver. Video plugs and sockets .. .
Video extension speakers .. . Video
flyleads . .. Video RF interference
filters .. . Video splitters .. . Indoor
antennas . . . Video switching units
. . . Down converters . . . Video
speaker controllers . . . Video
camera lights . .. Video tape
rewinders ... Video cine adaptors
. . . Video head cleaners . . . Video
splicing kits ... Video storage
cases . . . Video dust covers . . .
Video leads . . . Scart plug leads
. . . Video dubbing kits . . . Video
headphones . . . Video shotgun and
wireless microphone systems .. .
Pre-amplifiers with video inputs .. .
Video camera stands . ..
Just about anything
you want.
. . . Try us ... NOW!
Get your catalogue FREE
from your local ARISTA
dealer or send $2.50 P&H
and your return address to:
ARISTI\.
ELECTRONICS PTY LTD
PO BOX 191, LIDCOMBE, NSW 2141
66
SILICON CHIP
W~'S\½ER
M\CA
~£'RE. WAS A G~~~ MARK ~~E 1ME:
FLASHOVE."R \.\AD CXCOR~E.'t); FROM -rME
-r'R.AN~\S"f"OR CASE. 1'0
"ffie. HEP«'SINK,
AGRO~S 1'\E. E-'PGE 0~ 'THE MICA WAS\-\~,, ...
capacitors, C310 and C311, both
100µF.
But he'd kept the best bit until
last. A prime suspect was the trigger transistor, Q321. This should be
replaced as a matter of course,
along with any othe~_r of the
aforementioned components which
tested less that 100% or were
doubtful for any reason. I thanked
him very much and then, as he
wished me luck, he joked; "If that
doesn't fix it, - you're on your
own."
A comforting thought!
But at least I now had a few leads
to follow. I started with C361 and
D361. Both checked OK but I
replaced them anyway; after all, if
a minor component has to be pulled
out to be tested, it is just as easy to
fit a new one in its place .
When I came to capacitors C332
and 351 it was obvious that these
had been replaced by the previous
serviceman. Not only were they
brand new but, for once, I was
grateful for his style of soldering; at
least I knew what he had tried.
Similarly, the main smoothing
capacitors, C310 and 311, had obviously been removed, though I
couldn't be sure they'd been replac-
ed. So I pulled them out and checked them. They both tested OK, well
on the plus side in fact, so I refitted
them.
That left only the trigger transistor, Q321. Since my predecessor
seemed to have been following the
same path, I wondered whether he
had received the same advice and
had replaced this also. If he had,
we were back to square one.
Fortunately one glance at the
board reassured me. There was no
way that Butcher Bill could have
removed that transistor and put it
back the way it was. So out it came
and in went a replacement. Then I
fitted yet another switching transistor, put everything back together, and switched on again.
It was no surprise when the set
came up with perfect picture and
sound; that had happened many
times before. The real test was
whether it would last.
I ran the set every day and all
day for the next fortnight and it
never missed a beat. But as vital as
this test was, I couldn't convince
the customer of that. In spite of the
fact that they had a spare set, she
was on the phone every couple of
days wanting to know when it
TETIA TV TIPS
Kriesler 59-4/5/6
Telecommander Remote Control
Symptom: At switch on will select
channel 6 sound only at low
volume. No picture . When any
channel button is pressed, sound
increases to normal but still no sign
of picture .
Cure: IC1320 {TDB1030) faulty.
Although IC1380 does the channel selection , it is IC1320 that
does the actual switching and a
fault in this chip cuts off the tuning
voltage.
would be ready. And no amount of
explanation on my part seemed to
penetrate.
But at the end of a fortnight I
reckoned that the point had been
proved, at least to the extent that it
could ever be proved, and I returned the set to her. I made no charge
and urged her to contact me if there
was further trouble. This was accepted with moderate grace but
without any enthusiasm.
Well, that really is the end of the
story, except that several months
have now passed and I have checked the situation several times during that period. And each time I
have been assured that all is well.
So it would seem that Q321 was the
real culprit.
Of course it's easy to be wise
after the event but it is now obvious
that the flashovers around the transistor were merely symptoms of the
fault that was generating destructive spikes. The fact that I fitted
washers marginally different from
the original simply served to produce a nice red herring; one which
cost me dearly.
So why did Q321 generate those
spikes? I haven't a clue - and
neither, it would seem, does Sanyo.
Suffice it to say that it is a well
documented fault and one worth
noting. Perhaps this story may save
someone else the run around.
Another mystery
And while we are talking about
mysteries, here is another one to
mull over. It is about a brand new
National colour TV set, model
AW A Thorn 3500 Series
Symptom: Very low, wedge
shaped picture with very bad convergence. All voltages normal and
vertical output waveforms also normal. Looks very much like shorted
turns in the yoke.
Cure: Break in track on convergence board, between Plug 25
Pin 6, and L7 54. This track
passes close to one end of R751,
a known source of dry joints and
the break was at the closest point
to the resistor.
TC2959, only recently released.
This is a most impressive set,
featuring a 74cm (29in) screen, a
frighteningly elaborate remote control system, and a price tag of about
$2000 to match. It also boasts the
best picture resolution that I have
ever seen on a colour set.
Shortly after its release the
makers issued a modification sheet
concerning a corner purity problem. This was due to insufficient
degaussing current due, in turn, to
the type of degaussing thermistor
originally fitted. It recommended
fitting an alternative type, Part No.
ERPF5BOM12OG.
I have modified a few of these
sets and the modification works
well, even without any external
degaussing. So I wasn't worried
when one of my regular customers
phoned to say that they were having trouble with this model set, the
customer's description of the fault
being that " ... the colour's gone all
funny".
She explained that the set had
been working fine when they had
gone away for the weekend but had
"gone funny" when switched on
after they returned. I imagined that
it was just another modification job.
In fact it turned out to be one of the
worst purity problems I have ever
seen; the whole screen was a bright
random pattern involving every colour of the rainbow. There was no
way that this was the problem I had
anticipated.
Nevertheless, I went through the
motions of checking the thermistor,
confirmed that it was the old type,
Circuit Notebook ...
continued from page 59
plugpacks up to 15VAC are
suitable. Either way, the source
should be able to supply about
50mA or more.
The circuit operation relies on
the fact that the LM317 maintains 1.25V (nominally) between
its adjust terminal and the output
terminal. By connecting a
resistor between these terminals
it can thereby be made to function as a precise constant current source which is independent
of input voltage or load voltage.
Two resistors in series, 22n
and 2.20, set the constant current at close to 50 milliamps (actually 51.6 for a regulator
voltage of 1.25V}. Up to six cells
in series can be charged. Charging time is 14 hours (unless otherwise specified).
If you have 400mAh AA cells
they should be charged for 14
hours at 40mA. To set the ch.arging charging current at 40mA,
change the 2.20 to 8.20.
116
TV TECHNICIANS &
SERVICE COMPANIES
Have your
PHILIPS V ARICAP
TUNERS (ELC2060)
Repaired or Exchanged
Only $17 .00 ea.
Ten or more $15.00 ea.
Quick Service
3 Months Warrc1nty
T.V. Test Equipment
Shorted Turns Tester
(Silicon Chip, Sept. 88)
$65.00 ·+ Tax
Hi-Voltage Probe
0 · 50kV.
$70.00 + Tax
Send Faulty Tuners
Cheques and money orders
+ P&P $2.00 or C.O.D.
-:r.v.Tu
IIERs)
216 Canterbury Road,
Revesby 2212, Sydney Australia
Phone: {02) 77 4 1154
DECEMBER 1988
67
presto! - the problem vanished; I
had a first class picture with as
high a degree of purity as I have
ever seen.
So what was going on? My first
reaction was that there had to be
some factor in the customer's home
which I had overlooked, such as
audio speakers which might have
been moved close to the TV set. But
even that assumption posed as
many questions as it answered.
In order to change the thermistor, I had pulled the set into the
centre of the room, where it was
quite clear of any other objects. So
why, in that situation, had my
degaussing wand lost its magic?
Spurious magnetism
and fitted a new one. All of which
had no effect whatever on the problem. This didn't surprise me and I
simply reached for the degaussing
wand and gave the front of the set a
once over lightly.
But that didn't help either. When
I switched the set on again the ef~
feet was as bad as ever, although I
did get the impression that the actual pattern had changed. So I tried
again, being more careful this time
and since the back was still off the
cabinet, I went right over the inside, poking the wand in wherever
it would go.
Still no joy. I went through the
ritual several times but seemed
unable to make any impression. The
only noticable effect was that, if I
left the set running, I could observe
a changing pattern but with no hint
of reduction.
Faulty picture tube
That really rocked me. There
was only one possible explanation
in my experience and that was a
faulty picture tube. The fault is
rare but I have experienced it. It is
caused by a faulty spot weld securing one corner of the shadow mask,
allowing the mask to move back and
forth and upset the critical adjustment of the whole system. And a
68
SILICON CHIP
symptom is that any attempt to
degauss the set simply changes the
pattern.
It was also a situation which called for some diplomacy. The people
involved were very understanding
types but I hesitated to suggest that
their brand new $2000 set had a
faulty picture tube. While it would
be fixed under warranty, they'd
probably worry about a picture
tube failure for years to come!
So I simply suggested that the
fault needed workshop attention for
a few days. They still had their old
set, a Philips K9, which would serve
in the meantime.
And so the set finished up in the
shop. I left it for a couple of days,
partly to think about the problem
and partly to organise a discussion
with the National service department. It was just possible that they
may have already encountered a
problem with this picture tube.
But when I put it to them I drew a
blank. While they were well aware
of such a possibility, they had not
had any such problems. So I was on
my own.
At the first opportunity I gave the
set another try. It displayed exactly
the same pattern as before and, on
an impulse, I gave it another once
over lightly with the wand. And
Nevertheless, I rang the customer and asked her whether there
was such a speaker system and
whether it might have been moved
close to the TV set. Yes, they did
have such a system and one
speaker normally sat alongside the
TV set. In fact, it had been sitting
there when I called.
I realised later that if it had had
a mouth it would have bitten me. I
was just too intent on the TV set to
notice it and, as I said, I had pulled
the set clear anyway.
More to the point the lady was
quite adamant that nothing which
might be involved had been changed. The set was working perfectly
when they left the house on the Friday night, the house was empty until they returned on the Sunday
night and, at first switch-on, the set
was faulty.
There was nothing for it but to
return the set to the customer and
test it in its original environment.
The answer - nothing. Nothing untoward, that is. The set behaved
perfectly, even though I had taken
the trouble to ensure that it went
back exactly where it had been,
relative to the speaker.
Finally, I moved the speaker all
around the set and did everything I
could, short of shoving it inside the
cabinet, to create the fault. Nothing
worked; nothing I could do would
re-create the problem. And the set
has been running for many weeks
now with no hint of trouble.
So the mystery remains. Any
ideas, anybody?
~
|