This is only a preview of the May 1996 issue of Silicon Chip. You can view 24 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 Voltage Insulation Tester":
Articles in this series:
Articles in this series:
Items relevant to "KnightRider Bi-directional LED Chaser":
|
SERVICEMAN'S LOG
It was a dark & stormy night
Yes, it was; very dark and very stormy.
The storm had blacked out several Sydney
suburbs and, in the process, created a line
surge which damaged the set featured in this
month’s notes. And it was a dark and stormy
exercise correcting the damage.
The set was a National Panasonic
model TC-68A61, fitted with an M16M
chassis. It is a 68cm set, featuring remote control plus all the latest bells
and whistles, and retails for around
$1800. It was quite new, being only
about 14 months old.
I discovered later that this was not
the only TV set to be damaged in this
40 Silicon Chip
and subsequent storms a few days
later. There were many more from all
over the suburbs, one of which was
described as a complete write off.
The set came in with the simple
description of being completely dead,
which it was from the customer’s
point of view. A quick bench check
produced a violent squeal from the
switchmode power supply, suggesting
a short on one of the supply rails.
Unfortunately, this model set was
a complete stranger to me. I had
never even seen one before and had
absolutely no data of any kind. Nevertheless, I decided to at least take
the back off the cabinet and check
for any visual clues. This operation
produced its own shocks. Firstly,
everything was jam-packed in
– a real servicing nightmare.
Secondly, the cabinet was
of relatively flimsy plastic
so that, when the back was
removed, it distorted noticeably under the weight of the
large tube.
There were no obvious
signs of damage, so I decided
to pull the chassis for a closer
look. This was a difficult operation,
due in part to the distortion of the
cabinet, although I realised later that
there were some tricks which made
it easier.
Anyway, with the chassis out, my
main aim was to try to find whatever
it was that was obviously overloading
the power supply, as suggested by the
squealing. I went first to the horizontal
output transistor, Q551, and checked
for voltage on the collector. There
was none so I pulled this transistor
out, expecting it to be shorted, but it
was intact.
So it looked as though the fault
was closer to the power supply but,
without a circuit, it was impossible
to identify the various rails or even
to know how many there were. My
best effort was to find that there was a
dead short to chassis from a test point
labelled TPD1, which appeared to be
one of the rails.
At this point, I realised that it was
hopeless to proceed without a manual
or at least a circuit. Fortunately, I was
able to find a colleague who did have
a circuit and he was quite happy to
lend it to me. It amounted to a total
of six A3 pages! These cover a swag
of boards or modules, designated alphabetically. I ran out of fingers trying
to count them but I make it about 16.
The accompanying illustration is part
of the D board.
Just as importantly, my colleague
was able to pass on a lot of valuable
information based on his own experience with this model set. Of particular
value was a warning about powering
up the set after a repair. It appears that
the set is very easily damaged if other
faults are overlooked.
This was a kind of “good news/
bad news” situation; I was extremely
grateful for the warning but not very
happy about the need for it.
Circuit details
Anyway, now that I had a circuit I
could at least begin to sort things out.
The set has two switchmode supplies:
(1) a main one supplying the high
voltage rails; and (2) a subsidiary one
supplying a 5V rail for the remote
control functions, plus a 12V rail. This
12V rail is very important because,
among other things, it powers standby
and protection circuits. And it func
tions continuously.
The main supply centres around
transformer T801 and the short I had
Fig.1: portion of the D board on the National Panasonic TX-68R71. The
subsidiary supply, involving T881 and its associated parts, is at top left, while
the main switchmode supply involves T801 and transistors Q801-Q805. IC801
is at bottom centre, IC802 to the right and SCR Q821 above it.
found was in fact on the main HT rail,
normally operating at 139V. It involves
transformer pin S2, diode D808, filter
capacitor C828 and IC801.
I connected the ohmmeter between
TPD1 and chassis and progressively
removed components from this line,
including IC802, C828 and some other
minor components, until I came to
SCR Q821. I pulled it out and the short
cleared, which meant that the SCR had
broken down.
But what was the SCR’s function
and, most importantly, why had it
failed? Once again I am indebted to my
colleague for saving me from having
to try to work this out for myself. SCR
Q821 is part of an over-voltage protection circuit, particularly guarding
Q551 and the horizontal output stage
in general.
And it had done a good job, to the
point of sacrificing itself. But the implication from such a drastic reaction
could only be that it must have been a
very severe voltage overload. So how
could I fire up the set safely to make
further tests?
Normally, I would use a Variac for
this job, possibly with a series lamp in
the mains line as a current limiting device. Unfortunately, another colleague
had passed on some hearsay advice
that a Variac could not be used on
these sets, although the explanation
was hopelessly garbled.
As it turned out, this was a furphy.
It appears to have arisen from a warning in the manual, which I saw later,
against depending solely on a Variac
for protection before all the recom
mended tests had been performed.
But that was later and, right now,
with various warnings ringing in my
ears, the best I could do was settle for
a 200W series lamp in the mains lead.
I also took the precaution of disabling
the horizontal output stage by shorting
the base and emitter of Q551.
Then, with a meter monitoring the
main HT rail, I switched on. The reaction was quite dramatic – the meter
shot up to over 200V, clearly indicating
something seriously wrong with the
power supply regulation system. And,
as if to confirm this, in the few seconds
I took to absorb the reading, there was
a loud bang.
The excessive voltage had proved
too much for C760, a 0.47µF electrolytic rated at 160V, which had exploded.
And when they explode they don’t
muck about. Fortunately, this was
easily fixed and there appeared to be
no other damage.
At this stage, I encountered another
colleague who was able to loan me
a copy of the service manual. This
includes a section entitled “Service
Hints for M16M Power Supply Repair”. And almost immediately, it
begins listing “possible causes for a
power supply primary shutdown”.
Among other symptoms, it mentions
the mains fuse, F801, being obviously
O/C, and transistors Q803 and/or Q805
being physically blown apart! It also
suggests checking IC801, with a low
ohmmeter, in anticipation of it being
“absolutely S/C between all three
terminals!”
The manual goes on to list all the
components which should be checked
in the event of a “primary shutdown”.
And it includes instructions as to how
components should be tested, strict
warnings about the critical nature of
many components, and the risks of
using substitutes.
All told, it lists no less than 16 components which should be tested before
applying power. The risk appears to be
that a serviceman may follow the usual
practice of progressive testing; ie, find
and replace a faulty component, then
reapply power, check performance,
and search for further faults if necessary. The manual warns that this
approach could likely result in further
severe damage.
It’s not the most encouraging introduction to a strange set!
Voltage regulation
But at least I had been warned. And
my attention was now directed to the
voltage regulation system; to find out
May 1996 41
Serviceman’s Log – continued
how it worked and why it didn’t. Once
worked out and explained, it is not
hard to follow but it wasn’t easy coming to it cold. It all hinges around IC801
and D812, the latter an opto-coupler
IC801 is a 3-terminal device. Pin 1
connects to the 139V rail, pin 3 connects to chassis, and pin 2 connects
to pin 2 of the optocoupler, which is
the cathode of its internal LED. Pin 1
is the anode of this LED and goes to a
12V rail from IC802. The other half of
42 Silicon Chip
the optocoupler is a transistor, with the
collector connected to pin 4 and the
emitter to pin 3. The base is activated
by light from the LED.
In operation, IC801 conducts between pins 2 and 3 when the voltage
on its pin 1 terminal reaches 139V.
This completes the circuit between
the 12V rail and chassis via the LED in
the associated optocoupler. The LED
now glows and turns on the transistor
between pins 3 & 4 of this device.
Pin 3, in turn, drives a transistor
network consisting of Q802, Q803
and Q801. The latter is at the heart of
the switchmode supply and switches
the primary of transformer T801. By
controlling the oscillator activity when
the main rail reaches 139V, that voltage
is maintained.
At this point, I decided that the best
approach would be to order all the
components listed as likely needing
to be changed and put the set aside
until these arrived. This would save
time and any components not needed
could go into stock. I had an idea that
this would not be the last of these sets
I would see. The only snag was that
I was quoted up to three weeks delay
on some parts.
This was an irritating setback but I
decided to make the most of the time
by trying to pinpoint the most obvious fault – the failure of the IC801/
optocoupler combination to regulate.
Testing IC801
As already mentioned, the manual
suggests that IC801 is a prime suspect,
most likely going short circuit. Well,
I’d already cleared it of short circuits
but it could still be faulty. How to test
it? Well, not in situ, since power could
not be applied.
A preliminary resistance check
revealed no continuity between any
of the terminals but that didn’t really
mean much. Once again my colleague
came to the rescue. He had already
made up a simple test jig and gave me
the details.
It was a simple enough arrangement
to knock up and I soon had it working. And it worked very well; so well
that it clearly indicated that IC801
had carked it, which was one good
reason why the HT rail was not being
regulated.
What about the optocoupler? The
manual had made the point that if the
optocoupler proved to be faulty, then
IC801 should be replaced automatically. Would the reverse be true?
The manual suggests testing the
optocoupler using an ohmmeter and
I have no doubt that it is technically
accurate. However, the optocouplers
are very small devices and trying to
test them in this manner is fiddly, at
best. So I added to my colleague’s jig,
making it a combined tester.
It was all very nice in theory but I
needed a known good IC801 to make it
work. This was one of the components
on three weeks delay, so I cheated by
connecting the prods of an analog
multimeter (low ohms range) across
pins 1 and 2 of the optocoupler to
energise the internal LED. I could get
no response from the original device
but the new one, which arrived early,
produced an immediate response from
the external green LED.
Eventually, the remaining parts arrived and I replaced the SCR (Q821),
IC801 and the optocoupler. I had
already replaced C760 which I had
blown up earlier and had spent some
time checking and double checking
all the other components listed – as
well as some that weren’t. In theory,
I should have been able to switch on
safely.
However, the manual suggests a
proper routine for switch-on at this
stage and I wasn’t prepared to take
any chances. What this amounts to,
in essence, is to disable the horizontal
output stage, replace it with a dummy
load, then wind up the supply voltage
via a Variac.
Talk about a belt and braces approach!
In greater detail, the procedure
involves lifting a 1.2Ω resistor (R561)
on the X board, which is in the 139V
rail to pin 9 of the horizontal output
transformer (T501). At the same time,
a dummy load, consisting of a 60W
globe, is connected from this supply
rail to chassis, most conveniently
from pin 1 of the X10 plug on the D
board to pin 1 of the X11 plug, which
is chassis. These are not shown on the
accompanying circuit.
The manual also suggests lifting
D560, which I did. This is to disable a
protection circuit involving transistors
Q553, Q554 and Q555. If this circuit
had been activated by a fault, it would
shut the set down and inhibit further
testing.
Having done all this, I connected
the set to the Variac but left my 200W
globe in series. I must admit that I was
extremely nervous about the whole
situation and felt that another belt
added to the belt and braces wouldn’t
do any harm. I also con
nected the
CRO to the collector of the chopper
transistor (Q801).
I switched on and wound the Variac up slowly. And, with only about
30V in, the CRO indicated oscillation
around Q801. Beyond this level, it
abruptly stopped oscillating. I gradually increased the voltage, eventually
reaching 150V, which was as high as I
was game to go – still no oscillation.
I backed the voltage off and moved
to the subsidiary power supply. I
checked the 5V rail out of IC803 and,
at about 100V in, it came good, as did
the 12V rail at zener diode D883. Well,
that was good news; very good news
in fact, because according to the manual, this supply is vital for the remote
control and protection systems.
Remote control switching
But it didn’t help much with the
main power supply problem. In order to follow what happened next,
it is necessary to look at the remote
control ON/OFF switching function.
Working backwards from the switch
mode section, involving transistors
Q802, Q803, Q804 and Q805, we
trace the circuit up to pin 3 of D841,
the second optocoupler. And pin 4 of
D841 connects to the 12V rail which
we had just checked. So the role of the
D841 is to switch the 12V supply to the
transistors in the switchmode supply.
D841 is controlled by transistor
Q841 between pin 2 and chassis. This
transistor is controlled, in turn, by the
remote control system on board E,
involving microprocessor IC1213 and
transistors Q1231, Q1207 and Q1209.
I won’t bore the reader with all the
details of this circuit operation – just
that it finishes coming in on pin D5
on the D board and goes to the base
of Q841.
So the remote control system
switches Q841 on or off, switching
May 1996 43
So what was wrong now? All kinds
of weird and complex possibilities
raced through my mind, without
making much sense. Then I suddenly
looked up and caught sight of the
200W lamp in series with the mains; it
was glowing a dull red. I had complete
ly forgotten that the lamp was still in
circuit.
I disconnected it and tried again.
And this time everything came good
–correct HT rail voltage, no signs of
distress anywhere, and the set actually
functioning. And functioning very
well, too.
Insurance
D841 on or off, and turning the switchmode system on or off. It’s simple
when you say it quickly.
Having worked out what should
be happening, I was able to trace the
circuit through and establish that
every stage was functioning up to the
base of Q841. But Q841 wasn’t doing
anything about it. I pulled it out and
found that the base-emitter junction
was open circuit.
This presented something of a puzzle. As far as I can work it out, this transistor must have been working when I
first turned the set on, otherwise there
could have been no HT rail voltage (the
excessive voltage which blew up capacitor C768). So, was Q841 damaged
by a kickback from this misadventure.
We’ll probably never know.
Anyway, that problem was easily
fixed. I didn’t have a 2SD1010 and,
conscious of the dire warnings about
substituting alternative components, I
hesitated initially. But it didn’t appear
that this was anything more than a
general purpose transistor so I took a
punt and fitted a BC547.
That started things working. As I
advanced the Variac the CRO indicated that the system was oscillating
and it kept on oscillating. And there
was voltage on the main HT rail at
test point D1 which, according to the
manual, should reach its normal 139V
operating voltage with an input as low
as 120V.
Unfortunately, it didn’t. At 120V
on the Variac the best I could get was
about 117V. I wound the Variac up to
around 150V, which the manual warns
is the limit if a normal HT value is
not reached. There was no significant
improvement.
But there was one more job I had
to do for the customer. Damage of
this kind is not, of course, covered
by warranty. But it was covered by
the customer’s household insurance
and I filled in the necessary details
on his claim.
As for the set itself – well, I wouldn’t
nominate its designer(s) for any Oscars. I cannot escape the impression
that they started off with a lot of surplus components and that they used
as many of them as possible!
An exaggeration? Well, maybe, but
other designs have produced the same
end result with less components and
greater reliability.
More to the point, from a practical
servicing point of view, I offer this
advice to anyone presented with one of
these sets. Do not, in any circumstances, touch it – and I mean that word
“touch” almost literally – without the
benefit of a manual.
If a manual cannot be obtained,
knock it back. To do otherwise is to
do both yourself and your customer a
SC
gross disservice.
Especially For Model
Railway Enthusiasts
Available only from Silicon Chip
Price: $7.95 (plus $3 for postage). Order by phoning (02) 9979 5644 & quoting
your credit card number; or fax the details to (02) 9979 6503; or mail your order
with cheque or credit card details to Silicon Chip Publications, PO Box 139,
Collaroy, NSW 2097.
44 Silicon Chip
|