This is only a preview of the September 1992 issue of Silicon Chip. You can view 51 of the 104 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 "A General-Purpose 3 1/2 Digit Panel Meter":
Articles in this series:
Articles in this series:
Articles in this series:
|
SERVICEMAN'S LOG
Death rattles: a thorny problem
We go south again this month, to commiserate
with one J. L., who regularly livens up these
notes. Maybe he lives in the Apple Isle, but
everything was not apples for several weeks
while he wrestled with an obscure fault in a
Thorn colour set.
This is a tale of woe but with a
happy ending. The woe has lasted
three weeks or more but ifI had known
when it started what I know now, it
could still have taken three weeks or
more.
The set was a 34cm English Thorn,
model 9418, now close to 15 years
old. The complaint was that it often
stopped working, for no reason that
the owner could discover. When it
was going, it worked well and showed
what he believed was an excellent
picture. However, it had been getting
worse and he was finally forced to
have it attended to.
I fired it up on the bench before he
left and, according to Murphy's Law,
it started and kept running. The owner
said it would fail within five minutes
loss
Our -IS
your gain
- ---,
'
I
I
II
'
L ~
-· . ·
Fracarro's
ESVU2. The
really low
loss VHF/
UHF mixer
with areasonable
price tag.
Peter C. Lacey Services Pty. Ltd.
P.O. Box 678 (74Fulton Ad.) Mount Eliza 3930
Tel:03 787 2077 Fax:(03) 787 3460 ACN006893438
64
SILICON CHIP
but half an hour later it was still going.
It did the same next morning but
just before lunch it gave out a loud
clattering noise and quit cold. I whipped the back off the cabinet, then
switched it on again. There was another clattering protest from the chassis and nothing more.
I located the 115V rail test point
and hooked a meter across it. Several
start-ups later I had learned nothing
more than that the power supply could
only deliver about 80V before it shut
down. There was nothing to indicate
what was causing the shutdown and
finding the cause was going to give
me some nasty headaches in the weeks
ahead.
·
I had previously serviced the set
some years ago for a minor problem
but that didn't require a circuit diagram. This time, it was going to be
different so I set about my files with a
vengeance.
Most of the Thorn models were actually an AWA chassis in disguise (or
should that be the other way round?)
but there was nothing like a 9418
among them. Neither was there anything similar among the uniquely
Thorn models.
Finally, in desperation, I turned to
a list of AWA models that had just
then arrived in the mail. The list was
helpful, in an unhelpful sort of way.
Instead of telling me that the 9418
used a so-and-so chassis, it said simply "Refer to Radio Rentals". Which
didn't bode well for me, since the
model was so old that the company
was now unlikely to have any infor-
mation on it in its files. Fortunately, I
have a couple of mates who work for
RR, so I fronted up and begged for
assistance.
All they could tell me was that they
wished me luck with the TX9, as they
called it. There hadn't been one in
their workshop for many years, for
which they were very thankful. They
remembered the chassis as · the very
devil to work on and a horrible thing
to troubleshoot.
In fact, this chassis is an English
model and is grossly over-designed,
as is their wont. It wasn't hard to see
why my friends were glad that it was
somebody else who had to fix the soand-so thing.
On the credit side, they did have a
copy of the circuit diagram somewhere
and eventually turned up a much photocopied drawing, many generations
from the original. But at least it could
show me was how the circuit was
arranged, even though many component values and part numbers were
almost indecipherable.
As a first approach to fault finding,
I gave the board a thorough going over
with a strong magnifying glass, look-
••• ,w 6:NGL\Slol T'~N-THlii::'
COM'PLJ:~1N-r WA.~~ 1-r
OFllis-N ~ WORKING,
FoR~~0~11tS:
OW~Q C.OV'-1:> l>\SCOV'~.. ,
likely to have a suitable replacement.
At this point I put
the set aside until I
could contact the
owner. It seemed
that from this point
on whatever I did
was going to take
some considerable
tim e , whether I
prov ed the transformer faulty or not.
I wanted to know if
he valued the set
enough to pay for a
long and involved
repair.
As it happ ened,
he was away for a
couple of weeks and
I had time to talk to
various colleagues
about the baulky
Thorn. My main interest was the noise
it was making as it
turned off.
.,..,... AN'9 11-\\S O'PEiNE-D
The noise could
best be described as
FIR.S"'r CP\N OF WORM~ "'0-\AT
someone play ing
'I: s-r,;r.uc.t<: VJ rr t-\ 11-\,s
the castanets . It was
a loud, wooden , rating for dry joints or other physical
tly sound that sometimes went on
problems. This revealed nothing that and on until I switched it off to save
could explain the intermittent nature
my sanity. Mostly though, it just gave
of the trouble.
a short rattle and then quit.
I had never heard anything like it
A can of worms
and I tried to find out from others if
Next I tried monitoring the collecthey could explain it. More particutor of the line output transistor. The
larly, I hoped they could suggest any
shape of the pulse at this point can
source other than the line output transoften indicate the type of fault that's
former.
eventually exposed. And this opened
Nobody had any more ideas than I
the first can of worms that I struck had and mostly they just shook their
heads and murmured "you poor-----"
with this set.
Back at the bench, I needed to devise
The line output collector waveform
a way to check the line output transis usually a single spike of anything
between 800 and 1500V, at line freformer. It was one of the very early
quency. In this set the pulse was there
diode-split types and I have not found
but at many different amplitudes and
the usual LOPT testing routines to be
at many different frequencies. In fact,
very helpful for checking these transI couldn't get the CRO to resolve anyformers.
thing but a jumbled mess in the few
It seemed as though I had two
seconds before the power supply shut
choices. One was to power the set
down.
from an external 115V source. The
I didn't like the look of the way this
other was to isolate the set from its
job was unfolding, since the loud clatown power supply and see if the lattering noise seemed to be coming from
ter would run on a dummy load.
the line output transformer. If that
The first idea was the easier of the
proved to be faulty, the set might well
two and would have be en implebe a write-off, since it was unique to
mented ifmy high voltage power supRadio Rentals and nobody else was
ply had not chosen to kark it. The
°™~
5~,...
supply is a simp le DC source that I
drive from a Variac. The output is
filtered but unregulated and does a
good job in cas es like this. As far as I
could tell, the only trouble w as that
the smoke had escaped fro m one of
the diodes but I wasn't in the mood to
try to replace it at that time.
So that left me with option two - to
get the set's own supply running. And
this, at last, pointed m e in the right
direction.
The circuit diagram (see Fig.1)
shows a link (extrem e right) th at can
be used to isolate the power sup ply.
This implied that the supp ly should
run indep endently of the line output
stage and that it was not one of those
fi endish designs needing a line pulse
to keep it running.
But w hat kind of power supply was
it? Most ofmy coll eagues referred to it
a switched mo de supply but I had
some doubts about this terminology;
p erhaps it de pends on on e's definition of a switched mode supply. I w as
more inclined to regard it as a simple
transformerless arrangement w ith a
switching functio n to facilitate phase
controlled voltage regulation.
Jose's Spanish Dancers
Anyw ay, terminology aside, I fo und
the link on the board and took to it
w ith a pair of sidecutters. I th en wired
a dummy load (a 60W light bu lb)
across the supply to simu late th e absent receiver load. An d in next to no
time I h ad the supp ly u p and rattling
like Jose's Spanish Dancers. If anything , the n oise was now even more
like the rattle of castan ets.
With the line stage disconnected, it
was apparent th at the rattle was coming from a large w ind ing n ear the
back edge of the chassis. This turned
out to be L65 (top centre), which appears to be a simple filt er choke.
As m entioned earlier, this set was
grossly over-designed, as is so much
English equipment. The power supply alone employs three SCRs, three
tran sistors, 16 diodes and innumerable res istors and capacitors . Finding
out w hich of that lot was fau lty was
not going to be easy.
By sw itching the set on re peatedly,
I learned that it h ad two m odes of
failure. On e was immediate in that it
started clattering straight away. The
other allowed the supply to get up to
sp eed but then run with the nasty
rattle already described.
SEPTEMBER1992
65
SERVICEMAN'S LOG - CTD
By monitoring the output, I found
that the first type of failure caused the
supp ly to shut down as the output
reached about 80V. With the other
failure, it ran long enough to actually
reach 115V.
However, this was varying wildly,
swinging some l0V either way. At
other times, it would swing even
higher and the supply would shut
down. So that was what I was faced
with and somehow I had to sort out
the trouble.
pacitor C147 (bottom right with C148).
So this implied that all current before
D82 had to be very lumpy DC. That
was the start.
Next, CSR2 (right of centre) was
obviously wired as a crowbar protection circuit, driven either by transistor TR66 (extreme right) in the overcurrent mode or via a 130V zener D85
(below TR66) in the over-voltage
mode. This circuit is designed to blow
the main fuse in the event of an overload. Since this wasn't happening, I
felt that I could ignore this part of the
circuit, at least for the time being.
CSR3 (left of centre) is another small
thyristor and this triggers the gate of
CSRl via transformer Tl. From here, I
theorised that since the base of TR65
(centre) was linked to the slider of a
trimpot marked "set 115V", this transistor had to be the error amplifier (its
collector drives CSR3's gate). And because the trimpot was in a circuit
across the first main filter capacitor
(C147), the transistor could only be
there to set a DC level on the gate of
CSR3.
Switching pulses for CSR3 are derived from the output of the bridge
rectifier via TR62, D72 and a filter
network consisting ofC138 and R168.
Transistor TR62 is fed from a divider
across the rectifier output and pro-
How it works
My first problem was to work out
how the system was supposed to function. With unknown circuits like this
one, I begin by redrawing the circuit
diagram, omitting everything except
the transistors and the main capacitors and inductors. This leaves me
with the bare-bones outline of the circuit and I can imagine what needs to
happen to produce the desired outcome.
In this case, since there was no
input filter capacitor after the bridge
rectifier, we had a raw 100Hz DC supply on the anode of the main switching thyristor CSRl (top left) .
The current out of CSRl passes
through L65 and diode D82 (to the
right of L65) before reaching filter ea-
C
~
..
"',. '°"
.
,.,.._.
"
"
'
:::
~
'.,
071
u
II
.._
__
11U1
111('),l(t-'
\.-
. ..0
~.
0 • .0-~
~/~\C
~~~
MJ,.
66
SILICON CHIP
~•..t\
&Clll
·~-~
:.a
0
8
UO'll't
!
"
.
u,u
DC )Of
Fig.1 (below): as can be seen, the only
available Thorn 9418 circuit was
pretty grotty. Our draughtsman has
fitted fresh labels to help identify
those components mentioned in the
text. Basically, the circuit uses CSRl
to phase control unfiltered DC from
the bridge rectifier. The output from
CSRl is then filtered by L65, C147 and
C148. TR65 is the error amplifier.
s .
~
.. "'"°,f
duces a 100Hz square wave on its
collector (point 2). This is then fed
via D72 to the filter network to produce a sawtooth waveform [point 3)
which is in phase with the 100Hz
unfiltered DC. Thus, CSR3 triggers
earlier or later in the cycle, depending on the voltage on its gate.
As can be seen, there are an awful
lot of components additional to those
mentioned and any one of them could
have been causing the trouble. What I
needed to do was to isolate the various parts of the circuit to see ifl could
find out just where the trouble was
occurring.
Although the circuit diagram shows
four oscillogram numbers (in circles),
I didn't have copies of these. All I
could do was check these waveforms
and try to make sense of what I found.
Numbers 1 and 2 made perfect
sense. No.1 was the l00Hz ripple one
would expect out of a bridge rectifier.
And No. 2 was a steady square wave
_
,w:r
11.11"
ec llt
o·
I C~I - H
~, _,.
__.
~cao.....
,.,,._
--E::
~JC\
A( .... . 111 ...&1.1
'fl{'"W{O
nlOw
~~~
"lS.S IOII
derived from the ripple. So far, so
good.
Waveform number 3 was a different story altogether. Although it still
bore a similarity to the waveform at 2,
it was jumping about all over the place,
changing both amplitude and frequency in a completely random manner. And the waveform at 4 was following 3.
My problem was to find out whether
the irregularities at 3 and 4 were the
cause or the effect of whatever the
fault turned out to be. I began my
attack by changing each of the semiconductors in turn. And while each
was out of circuit, I tested the resistors around it for value. After this
exercise produced nothing of any note,
I started on the capacitors, removing
each and testing for capacitance and
leakage. Again , I drew a blank.
The only thing I couldn't test was
the 130V zener, D85 . But, as previously mentioned, the crowbar circuit
didn't seem to be affected by the fault,
so testing D85 had a very low priority.
Doing it the hard way
Eventually, it got to the stage where
TETIA TV TIP
AWA C2341 TV set
Symptom: only even numbered
channels can be selected from
the front panel controls. There is
no problem with the remote control and all channels are readily
available.
Cure: this was caused by a fault
in the control microprocessor. The
front panel selector buttons communicate with the chip on different pins to the remote control receiver, hence the faultless performance of the handpiece.
TETIA TV Tip is supplied by the
Tasmanian branch of the Electronic Technician's Institute ofAustralia. Contact Jim Lawler, 16
Adina St, Geilston Bay 7015.
I had tested virtually everything on
the board. There was absolutely nothing that showed up faulty under static
tests. That meant that I had to start
substituting for suspect components,
AUSTRALIAN MADE TV TEST EQUIPMENT
10% discount on all test equipment.
15% discount on two or more items.
SHORTED TURNS TESTER
Built-in meter to check EHT transformers including
split diode type , yokes and drive transformers .
$78.00 + $4.00 p&p
since some faults only show up under
operating conditions (ie, with voltages
applied).
First, I replaced the three thyristors
and the three transistors but this gave
me no joy. Next, because there were
less capacitors than diodes, I elected
to replace the capacitors. And it was
here that I solved the problem, although I still don't know exactly what
the problem was.
I began with the electrolytic capacitors: C127, C143, C144, C146, C147
and C148. The smaller of these I simply replaced but the last two were
combined in one can and could only
be tested for capacity and leakage at
working voltage. To do this, I had to
resurrect my high voltage supply but
the effort was pointless. None of the
electros showed the slightest sign of
trouble.
Next, I started on the ceramic capacitors. I ignored C136 and those
around the bridge rectifier on the
grounds that there was no sign of trouble on that side of CSR1/TR62.
The next capacitor was the previously mentioned C138. It's a 150nF
ceramic rated at 100V. I don't have
ANTRIM
TOROIDAL TRANSFORMERS
HIGH-VOLTAGE PROBE
Built-in meter reads positive or negative voltages from 0-50kV. For checking EHT and other HT voltages.
$98.00 + $5.00 p&p
DEGAUSSING WAND
Strong magnetic field. Double insulated for your
safety with momentary switch operation. 240V
AC 2.2 amps. As important as a soldering iron!
$75.00 + $10.00 p&p
REMOTE CONTROL TESTER
Designed to test infrared or ultrasonic
control units. Supplied with extension
infrared detector lead. Output is via a LED
and piezo speaker. $84.00 + $4.00 p&p
From $19 exchange
plus post & pack
Z'Ji
I -~- T~NE~-~
Cheque, Money Order, Visa, Bankcard or Mastercard
Phone (02) 77 4 1154
216 Canterbury Rd,
Reve~bJ~t~;i:. 2212,
Fax (02) 77 4 1154
•
•
•
•
Now made in Australia
Standard range 15VA to 2kVA
Large ex-stock quantities available
Competitive pricing available for OE Ms
& resellers
• Simply the best transformers all round
HARBUCH ELECTRONICS PTY LTD
90 George St, Hornsby, NSW 2077.
Phone (02) 476 5854 Fax (02) 476 3231
SEPTEMBER1992
67
c.EN-rRe:)
CSR~ ("R\G-H"r OF
W~'S OSV\005\..V W\R&t:> AS
~ CROW6AR ?'R0"t'"e.c,...,-1ON •••
.. o
any 150nF capacitors in my stock. In
fact, I don't know anybody who does.
But I do have a good supply of 0.15µF
capacitors and I wasted no time fitting one.
And that was all it took. At next
switch-on, the supply came up to 115V
and sat there without a flicker. At last
the clattering noise that had so bedevilled the workshop had been eliminated.
I reconnected the link to the line
output stage and the set came good
with an excellent though somewhat
blue picture. A brief tweak on the
screen pots soon put that right and
the set is now producing a really firstclass result. Considering its age , it's
quite remarkable that it works so well.
Naturally, the owner was delighted
. that it proved not to be the expensive
transformer that I had originally suspected and was happy to pay my larger
than usual bill. And I don 't know
what was wrong with the capacitor.
But it certainly made a mess of an
otherwise nice little TV set.
Thanks J. L. I sympathise concerning what must have been a really nasty
68
SILICON CHIP
one. A colleague to whom I showed
the circuit commented that it looked
like the Thorn 8000, which was the
basis for the early AWA colour sets.
But he agreed that the similarity was
largely superficial.
The things I'm asked
And from my own notes, a brief one
to finish off. As I imagine most of my
colleagues will testify, servicemen get
asked some very funny questions at
times - funny peculiar, that is. It happened to me again quite recently. I
answered the phone and the caller
introduced himself as a friend of one
of my customers, who had assured
him that I could help.
And the problem? How to receive
interstate country radio stations, particularly in Queensland and Victoria .
And it was quickly evident that the
caller hadn't a clue. He fondly believed that it was only a matter of
buying a sufficiently sensitive receiver
and the problem would be solved.
His query set me back a bit. I suddenly realised that I had completely
lost touch with such activities. It was
all the rage when I was a teenager
(yes, I was one once!) and advertisements for top of the line receivers
always boasted that interstate reception was guaranteed. And many a tall
story was told about the number of
distant stations logged in one night.
But all that had died out decades
ago. So why the sudden interest? And
what was the technical situation now?
Sure, transmitter powers have increased markedly but there are also many
more stations on the air, many of them
sharing frequencies.
I summarised the situation, as best
I could, on the basis that there was
little future in it. Sure, some stations
might be heard at night, but not very
reliably. Fading would be a problem,
accompanied by distortion and adjacent channel interference. If he simply wanted to log a station, for the fun
ofit, OK. But ifhe wanted to seriously
follow a particular program for any
length of time, well, forget it.
Which prompted the logical question: why did he want to do it anyway? This caused a pause at the other
end of the line .
"Well, I'll tell you the truth. The
TAB has been taking a lot of money
from me and not giving anything back.
And I've worked out a system which I
reckon will beat them." (I'm afraid I
laughed outright at the word "system", but he was obviously deadly
serious).
"The trouble is, the local radio stations are not allowed to broadcast the
result of interstate races for half an
hour after the · race. But if I could
listen to the interstate stations I reckon
my system would work".
I said, "Well, I'm sorry mate but I'm
afraid you're flogging a dead horse:
(no pun intended). "There's just no
way you can receive signals on this
basis."
·
So he thanked me for the advice
and added, "At least you've saved me
spending a couple of grand for nothing". A couple of grand? Just to beat
the TAB!
So .what was this "system"? I'm no
follower of the Sport of Kings. I recall
putting a couple of bob into an office
Melbourne Cup sweep many years
ago but it must have been rigged because I didn't get anything back either.
So I'm not competent to comment.
But a couple of grand to beat the TAB?
Oh well, it takes all sorts I s'pose. SC
|