This is only a preview of the August 1990 issue of Silicon Chip. You can view 57 of the 112 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 "Digital Sine/Square Wave Generator; Pt.2":
Articles in this series:
Articles in this series:
Articles in this series:
|
It was the last thing I tried
"It was the last thing I tried". How often have
you heard a colleague say that? It's a pointless
remark, unless the speaker is silly enough to go
on looking for the fault after he has found it.
What he means is that it is the last thing
anyone would think of and that's the theme of
this story.
The story concerns a National
portable (34cm) colour set, model
TC1480A. This is a relatively recent model, released a couple of
years ago, and the particular set
had been in service for only about 6
months - which meant that it was
still under warranty.
In greater detail, it was one of a
~
batch sold by my dealer colleague
~·
to a local motel. He dropped it into
,v.::::;.~~ ~ the shop with the explanation that
{~ "It goes for a while and then it loses
~ the picture - the screen just goes
black. But the sound is OK".
That was a more detailed
description than I get from most
customers; in many cases, the
-OF
~qO-\ \
~U'R.~e:., ,-r w~s
11-\E: LAS, t"HING :C. TR\~D
46
SILICON CHIP
symptoms have to be prised out of
them. At the same time, I was a little surprised by it. Usually, a nopicture, no-raster situation suggests failure of the horizontal
deflection system, from which most
auxiliary voltages - including that
for the sound channel - are derived. So I sensed that this was likely
to be a "funny" one.
Anyway, I set it up in a corner of
the bench and let it run. It played
perfectly at switch-on and for the
next hour or so but then, just as my
colleague had said, it switched to
the no-picture, no-raster situation
but with normal sound. At this
stage I hadn't even fished out the
service manual and my memory of
the chassis - which is used in a
number of models - was a bit
vague. The last time I worked on
one was some years previously,
when the chassis first appeared.
At this point, it seemed that the
easiest thing to do was to take the
back off and look for anything obvious. Fortunately, the back forms
the larger part of the ea binet and
when it is removed, most the boards
are readily accessible.
I made a careful visual inspection of the boards using a glass to
check for dry joints or similar faults
but there was nothing obvious.
Then I switched it on and, again,
there was a perfect picture. I went
through the usual bash, wriggle and
shake routine to see if I could provoke the fault but it just kept on
working.
In fact, it kept on working for the
rest of the day. And all the next
day. And the day after that. So had
my prodding and bashing accidently cured the fault? It wasn't a very
satisfactory solution if it had, but
these things happen. If the worst
came to the worst, I could return
the set with instructions to bring it
back when the fault recurred.
I replaced the back and gave it
another test run. That lasted about
an hour, then it was up to its old
tricks; just the sound with a black
screen.
Well, that changed the whole
outlook. It was obviously a temperature sensitive fault which was
critical enough to sense the difference between the "back-on" and
"back-off" conditions.
I put it through this cycle a couple of times and, in the process,
learned something else. When it
failed it needed to be switched off
for only a minute or so for it to
recover, after which it would run
for another half hour or so before
failing .
This reminded me of an important feature of this chassis; a very
effective protection circuit, designed to sense a number of abnormal
conditions and shut the set down.
And when this happens, the set will
not function again, even if the fault
is cleared, until the protection circuit is reset by switching the set off.
(I mentioned a similar feature in
these notes, involving a National
TC2258, back in December 1987;
how time flies!).
Unfortunately, this can make servicing quite difficult in some circumstances. Is there a fault which
is being sensed by the protection
circuit? Or is it a typical intermittent, which commonly responds to
the off-on cycle anyway? All I could
do was press on and hope that more
clues would emerge.
Since it was necessary to leave
the back off in order to work on the
chassis, but also necessary to raise
the temperature to initiate the
fault, I fished out an old blanket I
keep for such situations and draped
it over the set. That worked fine; I
was able to bring up the fault in
about half an hour.
That was a help but it served to
reveal even more about the switch-
TV TEST EQUIPMENT
(AUSTRALIAN MADE)
off/switch-on cycle; switching the
set off for only a few seconds was
enough to restore performance which only added to the confusion.
Up until this point I had been paying more attention to the symptoms
than to the manual and the circuit.
Now, having established a behavioural pattern, it was time to
study the circuit and try to relate
the two. My first query was why the
sound system continued working
when, apparently, the horizontal
deflection system was failing.
The explanation was simple
enough. The sound section, including the sound output IC, is fed
from a 12V rail derived from the
transformer in the switchmode
power supply. This carries an additional winding which, after rectification and preliminary filtering ,
delivers a bout 17V. This is fed to a
3-terminal regulator which provides the 12V rail. So that was that
- the sound system power supply
was independent of the horizontal
deflection system.
That much clarified, it was time
to tackle the fault directly. The first
check was to determine whether
PCB POWER TRANSFORMERS
SHORTED TURNS TESTER
Built in meter to check EHT transformers including
split diode type, yokes and drive transformers.
$78.00 + $3.00 p&p
HI-VOLTAGE PROBE
Built-in meter reads positive or negative O-5OkV. Can
also be switched to check negative voltages. For
checking EHT and Focus voltages as well as TVs arid
Microwaves. Hi-tension voltages.
$89.00 + $5.00 p&p
LOW-VOLT AGE PROBE
Ideal for checking microwave ovens and TVs. The ~
ranges are from Oto 5kV and from Oto 1 OkV. Double ~
insulated for safety.
$79.00 + $5.00 p&p
D
DEGAUSSING WAND
Strong magnetic field, larger than usual coil with
multicore centre. Double insulated for safety with
momentary switch operation. 24OVA/C 2.2
amps. As important as having a soldering iron!
$75.00 + $10.00 p&p
Proudly manufactured in Australia by
TUNER REPAIRS
We repair most tuners on an exchange basis.
From only $17 .00 + p&p
Cheque, Money Order, Bankcard or MasterCard
,,
TUNERS'I
V•____
•
216 Canterbury Road,
Revesby, NSW 2212
Phone (02) 774 1154
HARBUCH ELECTRONICS Pty Ltd
90 George St., HORNSBY, N.S .W., 2077
Phone (02)476-5854, Fax (02)476-3231
AUGUST 1990
47
::a
::t:
z
n
n
0
~
CJ)
~
""'
~
,,,
1vwr1
~
- - ~cl:l! :).l)K.
~
. IS
_£°
470 ~
MOAIVEI.
'l!.Q?-,,
f~
'"°°"
~"""
~:21'
,.., ,-,o--v-·' -
:~i
:I[
- ;: ~ ~
~
~•\1::·1•
U) J
Lilli ISO
£~~;
cso,
.
! SC92S- lo
""'
c:YOR4M.n
R413
1-'«
.
II
1
'
T;;
t
--1.'5!0
._ , .,
~
- •
.
u
·· ~t
-:::z cs.:s
.:-sA~
,.
''"""'!( ) I :-,, j .
:81(
! ~~~
~
I i
. -
- - --
1
; ou,
•
:
j
i
I
C50l
~ =k=
~
·---=~. r~
...
'"'
'·"
R674
18.n: •. • r c- 140UUl/ZRIO ~
llo.)K ••• :_C-l401Al
11
'
It,~,
;
! ______,
00~~~!_ ;
JS03'
Uj
;>:zf ~ f
:
f',. ~V
"~
I
·.;, "~ool
" ii
I
,~...,.,,.
4,'T,J,.
;U; .~ I
~
·;
.
~~
~
~
2SAll2.l
0521
120
1.670
-
,."'
I
i;s
,,..
,...
""'
Jlf
!
""'!
1
I
~
I
i
~
• I
M.U, ~
r:::;'
;Y32\
mm
""
E~[§~
1 7' .
~
--
ID_5tj.
IOO
: ~
l
~
:1<at>l;~
~
:,
i
:I
ii
;j
·!j
6
r--
· '1
,ir--
f~
' 0521
ir
}
.,....,
~~
!
"\'t.'51
L:_;_
-
.•
<at>I :~
..,,,.'-J"
.
'- i«~TD
6i
_.7
.
~
i :c!,~'®! ';~
®I - ""
~
, ,. .l. . ~ :o ~
I !~
~
470
16V
•C6l'l
i·z:t
25"13094
. n-,..
::1
I
M.Al61l
06)0
~tls
Fig.1: section of the National TC1480A circuit. The horizontal output stage is at bottom left,
the protective circuit to the right of it, and the video/chroma IC at top centre.
l
pi;
1'
I
i
b i ~~
,.~KV
~
T i~&p
~e I
,g. ~
- R412
2-21(
"""'1.01"' ''·""""1
A
....
-·
l :"" ""
11~
t ~,~ ""
,;J.11
"'"
!~
!W
~
..Rl,,, 4Jf .
-j---- -- - - - -- - -- --'~"':_f
......
:.--
,
~
E-BOARD
TPN197059AB . .. FOR TC·1401AR
TPN197059AA .. . FOR TC·1401MR
TPN197059 .. . .. FOR TC·1401ZR
VIOEOICHROMA
JUNGLE
Q630
there was any EHT. Well, there
was some, but only a paltry 3kV
rather than the 25kV I expected.
This was the first real clue.
From here I moved back to the
collector of the horizontal output
transistor, Q501 [lower left of circuit). This should have been showing an output of around 900V p-p .
Instead, the CRO showed about
300V. That in itself was not surprising in view of the poor EHT figure,
but there was something else
wrong.
The frequency of this waveform
should, of course, be 15,625Hz. In
fact, it was more like twice this frequency which was a good enough
reason for the poor EHT performance. But why?
Back tracking from Q501 takes
us to Q502, the horizontal driver
transistor, and from there to pin 41
of IC601, the video/chroma jungle
chip [top centre of circuit). The
waveform here should be a 5.6V p-p
square wave at 15,625Hz. In fact, it
was well down in amplitude, was
nothing like the specified shape,
and was also at twice the correct
frequency.
So, did we have a temperature
sensitive fault in the chip? One way
to prove this was to to spray it with
freezer while in the fault condition.
But this was easier said than done.
Because of the aforementioned
protection circuit, simply spraying
the IC and waiting for a result could
well be pointless. If the fault had
tripped the protection circuit
(which seemed likely since the
horizontal output stage was involved), any change would not be apparent until the set was switched
off and on again. But, as I said, this
action might also cure the fault.
Then, a possible way out of the
impasse presented itself. Continued
testing with the blanket in place
revealed that, if the set was left on
for some time after the fault occurred, so that the chassis became
really hot, it might need up to 15
minutes in the off condition before
the set would play again.
So, after a good long heat soak, I
gave the chip a generous spray of
freezer, then switched the set off
and on again, just long enough to
reset the protection circuit had it
been activated. Result - a virtually
instant response as the tube warm-
TETIA TV TIP
Hitachi CWP139 (NP6C chassis)
Symptom: Set apparently dead .
Closer investigation shows that
11 0V rail is OK, as are the 800V
screen supply, the 180V video
output supply and the EHT.
Cure: 1 2V rail missing . In this case
there was an almost invisible dry
joint at the anode of D705. The
configuration of this supply is
unusual in that the diode is placed
between the earthy end of the
supply winding and ground, rather
than in the supply rail itself.
TETIA TV Tip is supplied by the
Tasmanian branch of The Electronic Technicians' Institute of
Australia. Contact Jim Lawler, 16
Adina St, Geilston Bay, Tasmania
7015.
ed up. Well, it hadn't done that for
the previous super heat cycles I had
inflicted on it. I reckoned that was
it.
I had a spare chip on hand and
42 solder-sucked and resoldered
pins later it was in place. Brimming
with confidence, I then tossed the
blanket back over the set and switched it on for a final test.
My euphoria lasted exactly half
an hour. Suddenly, the set was back
in fault condition and I was back to
square one.
Brain in top gear
So where to now? I suppose the
shock of the set-back put my brain
into top gear, or at least directed it
along a different path. I began to
think more and more about the protection circuit and the role it had
played in that other incident back
in December 1987.
On that occasion it was the protection circuit itself that was at
fault. So was that what was happening here? Had I been chasing a
false alarm? One way to clarify this
point would be to disable the protection circuit and see what
happened.
But first, a few points about the
protection circuit. It involves four
transistors, Q451, Q521, Q503 and
Q504 (bottom left of the circuit),
plus various diodes, zener diodes
and resistors. Just how it senses all
the parameters it does - or even
how many it senses - is quite complex but the manner in which it
shuts down the set is simply
explained.
The main voltage supply for the
video/chroma chip is derived from
the 110V HT rail via a 6.8k0
resistor (R519}. This applies 8.5V to
pin 42. And pin 42 is also connected
to the emitter of Q503 via a 560
resistor, R536.
Q503 is normally turned off but
when a fault is sensed, it turns on
and this pulls down pin 42, reducing its voltage to a small fraction of
its normal value. This shuts down
the most vulnerable parts of the set.
It was now fairly obvious that what
I had taken to be a fault in the chip
was really the result of the protection circuit doing its thing.
To disable this circuit, I had only
to lift the 560 resistor. But it wasn't
quite as simple as that. If there
really was a fault, there was a very
real risk of an expensive component
being destroyed before I could hit
the power switch. But that was a
risk I had to take.
To reduce this risk as much as
possible, I set up a meter to monitor
the HT rail and the CRO to monitor
the waveform out of the horizontal
output transistor. Then I tossed the
blanket over the set, switched it on,
and waited.
And I waited. After about two
hours, with the set still playing, I
was ready to believe that the protection circuit was the culprit. Then
suddenly there was no picture. My
gaze flashed first to the meter, then
to the CRO. But neither had changed. And a further check confirmed
that the EHT was normal.
I checked the picture tube G2
voltage, the value of which is not
given on the circuit. It was at
something over 400V which I was
happy to accept. I then checked the
voltages on the collectors of the R,
G, B driver transistors but all were
normal.
It was all becoming very disconcerting because I was now rapidly
running out of ideas. As far as I
could see, all the picture tube
operating voltages were correct,
yet the set just wouldn't work. It
was almost as though the electrons
had gone on strike.
At a more rational level I reasonAUGUST 1990
49
SERVICEMAN'S LOG -CTD
1lrd
£.1-E:C...--rR ONS ON ST'R \ Ke A, 1"H €"
PI C..-TORE: "TV~E: H~-re::R c...tRC.U
ed that there was only one thing I
hadn't checked - the picture tube
heater circuit (lower right of the
circuit). There isn't much to this;
the winding on the horizontal
transformer, a 30 resistor (R523), a
plug and socket, and the tube
heater.
I checked all these with the ohmmeter and all were OK. But there
had to be something wrong and I
wanted to make a visual check of
the heater . Unfortunately, the
physical layout makes this difficult
unless one knows exactly where to
look.
I switched the set on again, switched off the lights, pulled down the
blinds, and confirmed that the
heater could be seen. But it could
only just be seen and then at a fairly critical viewing angle. I restored
normal lighting and let the set run.
When it failed I darkened the room
again and took a careful look.
Result - no heater.
The culprit -
at last!
So I was on the track at last. But
was it the tube, or the resistor, or
50
SILICON CHIP
l'r...
the transformer? About the only
way to determine this was to
measure the heater voltage;
something that needs a special
meter because the supply is at
horizontal deflection frequency.
Fortunately, my Fluke 87 meter will
handle this with ease and that was
the answer; there was no heater
voltage.
Well, that cleared the tube which was a relief - and put suspicion squarely on the transformer.
But I wanted to be sure before I
pulled the transformer out and
replaced it. What was needed was
a continuity test but, from observations thus far, this would be valid
only when the system had been
heated to the point of failure - and
it would have to be done smartly.
I disconnected the lead from the
30 resistor and restored the connection with a clip lead. I set up a
meter for resistance measurement,
connected one lead to the other end
of the winding (pin 8), switched the
set on and let it run until it failed and then for a while longer to make
sure it was really hot.
Then I switched the set off and
quickly changed the clip lead connection to put the meter across the
winding. Result; open circuit. But I
continued to check the meter and,
within 20 seconds or so, it dropped
to zero ohms.
That clinched it of course but I
had to order a replacement
transformer. When this arrived, I
fitted it and gave the set a thorough
workout. It behaved perfectly, even
with both the blanket and the
cabinet back in place.
OK, so why did the failure in the
heater winding trigger the protection circuit? The answer is that this
heater voltage, via diode D523 &
filter capacitor C582, provides a DC
voltage to the protection circuit.
Presumably, this is . designed to
monitor the whole horizontal output
circuit but it was tricked by the
winding failure.
The protection circuit also
monitors a 25V rail which is derived from pin 6 of the horizontal
transformer. This rail also supplies
the vertical output chip (IC401, pin
7) via a 1.50 resistor (R411), one
end of which goes to the emitter of
Q451. Apparently, this arrangement is designed to sense any excessive current drawn by IC401.
The 110V HT rail is apparently
monitored by Q521.
But why did the set take so much
longer to fail after the protection
circuit was disabled? This is a real
puzzle and the only theory I can advance is that the winding failure
was not sudden. Instead, there was
probably an increase in resistance
initially. As a result, the heater
voltage may have been reduced
enough to trigger the protection circuit but not enough to seriously impair tube performance. But that is
only a theory.
And that's about as far as I have
been able to work things out so far
so it will have to suffice for the present. But about that heading;
perhaps it should have read " ... the
last thing left to try".
Who done it?
And now, for a complete change
of pace, here is a short story
without any serious technical aggravation. The main point of the
story is: "Who done it?".
It concerns a Sharp colour set,
model 9C143, a set that has been
around for about 10 years. The
owner, a new customer, simply
dumped it on the counter, said "It
doesn't go", and wanted a quote to
fix it.
I'm not into quoting in the normal
sense. It's a risky business. By the
time you've found out what is needed, most of the hard work has been
done, with the real risk that the
customer will shake his head, say
"Too dear" and walk out.
But, not wishing to come on too
strong with a new customer, and
because it involved a total failure, I
said I would have a quick look at it
and try to get some idea of what
was involved. If it was easy to
diagnose that would be fine but
beyond that I wouldn't commit
myself. After a moment's hesitation, he accepted the offer.
When I pulled the back off the
set, the first thing I checked was the
mains fuse. It was open circuit; not
blackened or even visibly faulty,
but open circuit nevertheless.
Whenever I find an open circuit
fuse, the next thing I suspect is the
degauss circuit. A common fault is
a short-circuited thermistor, although the fuse reaction is usually
more violent.
· So a typical routine is to unplug
the degauss circuit, fit another
fuse , switch on and watch what
happens. If the fault is elswhere in
the set, the fuse reaction can often
reveal a lot about it. As an added
precaution, I usually fit a lighter
fuse than the one prescribed, since
it no longer needs to survive the
degauss surge.
But in this case someone had
been there before me; the degauss
circuit was already unplugged and,
what was more, the plug had been
neatly taped to the winding, well
out of sight. OK, why had the fuse
failed?
The prescibed fuse is a 2A delay
type, so I fitted a 1.5A quick blow
type and switched on. Nothing happened; nothing untoward, that is.
The set came to life immediately
and produced a first class picture.
And strangely, there appeared to
be no purity error.
I let it run for a while, then
checked the degauss thermistor.
And, of course, it was shorted so
that was easily fixed. I then refitted
the degauss plug, changed the fuse
to a 2A delay type as prescribed,
and tried again. And again, the set
worked perfectly. I let it run for the
rest of the day but there was no
sign of trouble.
So why had the original fuse failed while the degauss circuit was
unplugged? A closer look at it pro-
vided the clue; it wasn't the
prescibed type but a 1.5A quick
blow type, similar to the one I had
used experimentally.
Well, I reckoned that was
pushing things a bit. Such a rating
is OK on a short term basis but is
almost certainly too light for continuous use, even with the degauss
circuit unplugged.
On that basis, I felt confident that
the set could be safely returned to
the customer. But what was the
history of all this and "Who did it"?
I hate to say this but I suspect that
it was a professional serviceman.
Granted, it could have been a do-ityourself type but, if so, would they
have thought to fit a lighter fuse?
And would they have bothered to
tape the plug out of sight, as if to
hide what had been done?
I doubt it. I fear it was one our
own kind who simply couldn't be
bothered doing the job properly and
took the easy way out. If so, I would
nominate him for the Striped Apron
Award for the year.
Of course, I quizzed the owner as
discreetly as I could but it was obvious that he couldn't (or didn't
want to) help. So I backed off and
that means we shall never know
what really happened.
But we can do without "servicemen" like that.
SILICON CHIP BINDERS
These beautifully made binders will protect your copies of
SILICON CHIP. In distinctive two-tone green high-quality
vinyl specially selected for SILICON CHIP, and with heavy
board covers, each binder holds a year's issues (the 14 issues of
Vol.1 or the 12 issues of Vol.2). It will look great on your
bookshelf.
* High quality
* Hold up to 14 issues
* 80mm internal width
* Gold printed with SILICON
CHIP logo on spine
Price: $All.95 plus $3 p&p each (NZ
p&p $A6). Send your order to
SILICON CHIP Publications Pty Ltd,
PO Box 139, Collaroy Beach, NSW
2097 (use the handy form on page
112). Fax 979 6503.
AUGUST
1990
51
|