This is only a preview of the August 1998 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 "Build A Beat Triggered Strobe":
Articles in this series:
Items relevant to "15W/Channel Class-A Stereo Amplifier":
Purchase a printed copy of this issue for $10.00. |
SERVICEMAN'S LOG
Neither cat proof nor kid proof
Many faults in TV sets are due to human
factors, or sometimes “cat factors”. Flower
vases, cats, boisterous children and TV sets
often don’t mix well together.
This month, we were blessed with
not one but two recent model Sony
53cm TV sets fitted with the current
BG-2S chassis and barely out of the
egg. It is unusual to be servicing such
new sets (particularly two of them),
as they would normally be covered
under warranty with a Sony Service
Centre. But these two came in by accident – literally.
The first one was a KV-G21F2
belonging to Mrs White, who very
foolishly left a vase of freshly watered
flowers on top of the set, while she
went shopping. And she left Pookie,
the pedigree Siamese cat, in charge
of the house.
Pookie decided, in his wisdom,
to leap onto the TV set, presumably
on his way to climbing the brocade
curtains. Unfortunately, there wasn’t
enough room for both the flowers and
Pookie on top of the TV. The result
was inevitable, with most of the water
finishing up inside the TV set.
The situation was compounded
when the lady switched the set on
that night, to be confronted by a pyrotechnic performance on the wrong
side of the screen, followed by complete silence and the sickly smell of
something burning. The result was
not only a technical problem but a
diplomatic one as well. The lady’s
husband wasn’t all that enamoured of
Pookie at the best of times and now
Pookie was really in the dog house –
which is a dreadful thing to happen to
any cat and doubly so for a sensitive
pedigree Siamese cat.
There was even talk of finding a
new home for Pookie!
Anyway, the upshot of it all was
that, by the time I arrived on the scene,
it was pretty horrible mess, with a
surpris
ing amount of corrosion. I
suspect that the vase contained more
than plain water, probably having been
enriched by something designed to
keep the flowers fresh plus some natural chemicals from the flower stems.
I cleaned it and dried it as much
as possible, then made a visual inspection. I was surprised to find
that the main fuse (F601) was
intact but R611, a 0.1Ω 0.25W
resistor supplying power to pin
2 of switch
mode transformer
T601, was open circuit. From
there, I traced the circuit from
pin 5 of T601 to pin 1 of IC601,
an STRS6708. And this IC had
broken down completely – it had
obviously gone short circuit and
taken R611 with it.
Before replacing these parts, I
made a few checks on the various
rails and found a short on the
main 114V rail, off pin 13 of
T601. This short could be anywhere but I decided that, for
August 1998 27
Serviceman’s Log – continued
the moment, I would start close to the
power supply and check from there.
I opened the rail by disconnecting
resistor R131 and inductor L802, then
bridged this gap with a 100W globe,
with a voltmeter across it.
With the short isolated, I replaced
R611 and IC601 and was rewarded
with normal power supply operation
and a 114V rail, up to the 100W globe.
But I still had to track down the short.
This wasn’t hard; the horizontal output transistor, Q802 (2SD1878), was
the logical suspect and sure enough,
that was it.
I replaced Q802, removed the 100W
globe, reconnected R131 and L802,
switched on and the set came to life.
Well, sort of. The first problem I observed was intermittent vertical scan.
I let it run for half an hour or so to see
what would happen and whether I
could pinpoint any likely cause.
I couldn’t and the vertical timebase
problem only became worse until it
failed completely. Again, I picked on
what appeared to be the most likely
suspect – the vertical output IC (IC551,
LA7830). I was right once again but
I must admit that it took no great
mental effort – the whole thing was
28 Silicon Chip
messy and badly corroded. Replacing
it cured that problem.
But I wasn’t out of the woods by
any means. I had fixed the most obvious and immediate problems but
a prolonged soak test revealed that
a few more subtle ones were still
lurking. In particular, there was some
horizontal tearing and either wrong
or no colour at switch-on, although
these problems usually cleared after
few minutes.
By now I was beginning to worry
that the soaking might have created
long-term damage which would continue to surface long after I had fixed
all the obvious faults. The problem as
far as I was concerned was how could
I possibly guarantee the repair under
these circumstances. All I could do
was press on and hope for the best.
In more practical terms, the chief
suspect was the jungle IC (IC300,
TDA8375A), a 56-pin monster which
had scored a direct hit from the water.
Quite frankly, I didn’t fancy having to
replace it. It would not only be expensive but the job would be time-consuming as well.
Nevertheless, that seemed to be the
next logical step so out it came. This
revealed several damp patches that
had been under the IC, along with
some wire links that were already
showing signs of corrosion. What ever
it was that had been added to the water, either deliberately or by accident,
it was a pretty potent brew, attacking
everything in its path.
I cleaned and dried everything and
fitted new links to the board. At this
point, I was all set to procure a new
IC but suddenly wondered whether
this might be overkill; it could be that
there was no fault in the IC itself and
that the faults had been due to the
damp patches.
Was it worth taking a punt on that?
If I refitted the original IC and the
fault(s) reappeared, I would have to go
through the whole replacement routine again. But if it fixed the problem,
I would have saved the cost of the IC
plus the delay in getting it.
I took the punt and I won; the old
IC behaved perfectly and even after
a soak test lasting several days, there
were no signs of trouble. The only
complaint came from Pookie’s master,
who was frustrated at the time it took
to do the job. But as I explained to him,
it was essential that I (soak) test the
set over several days to be sure that
nothing had been overlooked.
And what happened to Pookie?
Nothing drastic as far as I know, apart
from a few black looks and some
nasty comments from his master. I
did, however, suggest that the flowers
should be moved elsewhere to prevent
a repeat incident.
The fallen Sony
The demise of the second Sony
KV-G21F2 was more violent than the
first. It had actually fallen over onto
a carpeted concrete floor. As a result,
the cabinet was cracked and the set
was dead.
I didn’t enquire as to how this set
had met this untimely fate. No explanation was offered but the Morris
family has four boisterous young boys
so it wasn’t too difficult to imagine
what had happened.
A close examination on the bench
revealed that the main PC board had
sustained several large cracks, the
most obvious ones being around the
horizontal output transformer and
the front panel controls. These cracks
were all repaired but it wasn’t going
to be as easy as that.
The large cracks were easy; it was
the hairline cracks that were the real
problem. There are several approaches to finding these. Visual inspection
using a good light and a jeweller’s
loupe is one method and it really
pays to examine the board carefully
adjacent to where large cracks have
occurred.
The other approaches are basically electrical. If a voltage or a signal
cannot be found at one end of a track
but is present at the other, the break
is somewhere along that track. But
pinpointing it can still be difficult.
The light and jeweller’s loupe may
be sufficient but it often requires a
very sharp pointed probe which will
penetrate the board lacquer. And in
the end, it’s back to visual inspection.
It took a long time to find and fix
some of these hairline cracks and
it was a most frustrating exercise.
Finally, I was confident that I had
found the last one. I switched the set
on, there was a rush of sound and the
picture came on. My elation didn’t last
long – there was a sudden click and
then silence.
After some detective work, I found
yet another hairline crack which had
been concealed by paint and was lying between closely packed adjacent
tracks. But it was of no immediate
help; the set was still dead – only the
standby light could be switched on
and off with the remote control.
So where to now? Were there more
hairline cracks which I still had to
find or was there a component fault
as well? Either way, all I could do was
go over the whole thing again to see
if I could find a clue.
I went back to taws and checked
the power supply and rail voltages.
The supply was working perfectly
and the main HT rail was at 114V and
was applying voltage to the collector
of Q801, the horizontal drive transistor (2SC3209). And this gave me my
first clue – the voltage was too high at
that point because the transistor was
completely cut off. The reason for this
wasn’t hard to find; there was no base
drive to the transistor.
The set is switched on via pin 30
of microprocessor IC001 (system
control), which switches on 9V regulator IC521. Among other things, this
regulator supplies 8V to pins 12 and
37 of IC300, the jungle IC which had
reared its head in the previous story.
The regulator also switches a +15V
rail to the audio output IC (IC203), via
transistors Q207 and Q208.
These circuits were all working but
there was no horizontal drive signal
(15,625Hz) from pin 40 of IC300,
which accounted for the high voltage
on the collector of Q801. But was this
due to a fault in the IC or was it due to
something connected to IC300?
I went over everything I could think
of that might be the cause. I soon
established that the 4.43MHz crystal
oscillator reference wasn’t working
but couldn’t be sure whether this
was a cause or an effect so I put that
observation on hold. There is a “hold
down” function (HD.SW) associated
with pin 27 of IC001, which controls
transistor Q801 and whose function it
is to shut down the horizontal output
stage in the event of a fault. I disconnected the circuit from pin 27 in case
it was shutting things down but no joy.
Finally, I took the plunge and replaced IC300. And that was it; the
set burst into life and everything
functioned perfectly. And it kept on
working after several days soak testing.
And that was about as far as I could
push it. There was no way of knowing
whether there were any more hairline
cracks lurk
ing in the background.
There could be but until a fault develops, there’s nothing more that can
be done.
I mentioned earlier that the cabinet
had been cracked. It was a relatively
minor fault and a new cabinet would
cost a couple of hundred dollars. In
view of the cost, the owner was happy
to live with it.
So why did IC300 fail? My theory is
that it was damaged by spikes, probably from the horizontal output stage,
before all the cracks had been fixed.
So considering everything, the owner
was lucky the set was not written off.
And perhaps his four boisterous boys
should be consigned to the dog house
along with Pookie!
The Akai stereo TV
Mr Gavin is an amicable bloke who
was somewhat disappointed that
his 18-month old Akai CT2167A TV
August 1998 29
Serviceman’s Log – continued
set had broken down so soon. This
Chinese-made model is a 53cm unit
with remote control and Teletext.
Unfortunately, the owner had found
it increasingly difficult to turn the set
on with the remote, until eventually
it failed completely.
As it came to me the set was dead
except for the standby indicator LED.
This at least told me that the switch
mode power supply was working, as
the 5V rail for the LED is derived from
a 7V rail via Q901. In fact there are
four voltage rails, the highest being at
105V. My first step was to check this
voltage at the collector of Q402, the
horizontal output transistor. This was
present and was switchable with the
remote via Q905.
As for the Sony TV in the previous
story, the collector voltage on the horizontal driver transistor (Q401) was
high, once again suggesting a lack of
drive to that stage. This should come
from pin 41 of jungle IC IC301. And
again, there was no 8.8MHz signal
from the crystal clock oscillator at
pin 2 of IC301.
It fact, there was no voltage at all on
that pin, or anywhere else on the IC
for that matter. In particular, there was
little or no voltage on pin 42 which
should have been 8.6V. This voltage is
normally derived from the 105V rail
30 Silicon Chip
via 5W resistor R306 (6.8kΩ).
At least, the circuit showed it as
a 5W resistor. But if the undersized
unit in the chassis was rated at 5W
I was a monkey’s uncle. I pulled the
resistor out and measured it, only to
find that it had gone way up into the
megohm range. I fitted a fair dinkum
5W unit in its place and the set burst
into life and everything functioned
as it should.
My theory is that the increasing
reluctance of the set to turn on was
almost certainly due to the progressive increase in value of that 6.8kΩ
resistor. Just how the wrong wattage
resistor found its way into the set is
anyone’s guess.
Unusual video recorder
My final story for this month concerns an unusual video recorder that
Mrs Small brought in for service. It
was one of the NEC 9000 models; one
of a long-running series with many
different features and cabinet styles.
In this case, it was a model N9034M
which is a multi-system variant of the
N9034A, both of which were available
in Australia, although it was originally
designed to sell in the Middle East.
The “M” model covers PAL, NTSC
and MESECAM (a version of SECAM).
That is by way of background. The
actual problem involved the power
supply, which is designed to automatically operate from any voltage from
90-260V AC (either 50Hz or 60Hz).
It’s obviously aimed at the widest
possible market and is a set that you
could take almost anywhere.
An interesting point here is that the
power supply construction is completely different from the rest of the
video. In fact, it’s quite possible that
the power supply is not made by NEC
at all but by some other manufacturer.
That said, the power supply is obviously very well made. It’s built inside
its own metal cage and is relatively
easy to remove from the body of the
recorder. But that’s the end of the easy
part; working on it within the cage is
a job in itself.
The cage is made in two parts,
each roughly “L” shaped. One half
forms the top and one end, while the
other half forms the other end and
the bottom. The cage itself functions
as a heatsink, the heavy duty power
transistors and other larger compon
ents being mounted on one end. The
remainder is the circuit is mounted
on a PC board.
But there’s more to it than that.
From a servicing angle, this power
supply has more tricks than a cartload of monkeys. Any mistakes in the
diagnosis of this switchmode unit and
you can blow the lot in a nanosecond.
When that happens, you are right back
where you started, except that you are
at least $50 lighter.
Unfortunately, the replacement
parts are very expensive. At trade level
the transistors cost $24 each and the
resistors $1.62. The exception is 5W
resistor R2 which was listed at $4.15
before it became “no longer available”.
A quick check showed that resistor
R2 (20Ω) had failed. This resistor is in
the bridge rectifier secondary circuit
and it blows instead of fuse F1 which
is rated at 2A.
More to the point, of course, was
why did that resistor fail? And that
is the tricky part of servicing this
power supply. The circuit is such
that a failure in one component can
create a chain reaction that can take
out several other major components,
particularly the expensive power
transistors.
The safest procedure appears to be
to make as many individual component tests as possible before applying
power. In this case, the basic cause
SILICON
CHIP
This advertisment
is out of date and
has been removed
to prevent
confusion.
Fig.1: the universal power supply in the NEC 9034-M video recorder can
be tricky to service. A failure in capacitor C11 and/or C12 can also take
out power transistors TR1 and TR2, along with resistors R2, R10 and R11.
P.C.B. Makers !
If you need:
• P.C.B. High Speed Drill
• P.C.B. Guillotine
• P.C.B. Material – Negative or
Positive acting
of the fault was the failure of either
electroly
tic capacitor C11 (220µF
10V) or C12 (220µF 15V), or both.
This causes transistors TR1 and TR2
(2SC6378) to go short circuit, which
takes out 5.1Ω resistors R10 (5W) and
R11 (1W).
From this, it is easy to understand
how a newcomer could be trapped
by such a circuit. Simply replacing
one suspect component at a time is
not good enough. One needs to be
sure – or as sure as possible – that all
faulty components have been found
and replaced before applying power.
Fortunately, none of these dire
predictions eventuated. Having been
forewarned by colleagues, I was
extremely careful and the job was
completed without incident. So I was
lucky the first time.
Finally, here is a little snippet from
the reception counter. A teenage lad
came in with a video recorder and was
concerned about the loss of the small
flap, fitted on many machines, which
cover the various controls. Could he
buy a replacement and how much it
would it cost?
My assessment was that it would
cost far more than it was worth; that
it would be easier and cheaper to
simply forget about it. But the lad was
obviously not convinced.
“But what about the cockroaches?”
“The cockroaches? What cockroaches?”
“Well, isn’t the flap put there to
keep out the cockroaches?”
Well, maybe the flap was performing that function in its present environment but a I doubt whether that
was what was in the designer’s mind
SC
when he specified it.
• Light Box – Single or Double
Sided – Large or Small
• Etch Tank – Bubble or Circulating
– Large or Small
• U.V. Sensitive film for Negatives
• Electronic Components and
•
•
Equipment for
TAFEs, Colleges and Schools
FREE ADVICE ON ANY OF
OUR PRODUCTS FROM
DEDICATED PEOPLE WITH
HANDS-ON EXPERIENCE
Prompt and Economical Delivery
KALEX
40 Wallis Ave E. Ivanhoe 3079
Ph (03) 9497 3422
FAX (03) 9499 2381
• ALL MAJOR CREDIT
CARDS ACCEPTED
August 1998 31
|