This is only a preview of the May 1999 issue of Silicon Chip. You can view 33 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 "The Line Dancer Robot":
Items relevant to "An X-Y Table With Stepper Motor Control; Pt.1":
Items relevant to "Three Electric Fence Testers":
Items relevant to "Heart Of LEDs":
Items relevant to "Build A Carbon Monoxide Alarm":
Articles in this series:
Purchase a printed copy of this issue for $10.00. |
SERVICEMAN'S LOG
Life’s tough without TimTams
I must be getting old because a couple of jobs
really had me on the go this month.
Fortunately, persistence won the day and
I had a really good win over a recalcitrant
VCR. If only I’d stocked up on TimTams . . .
Panasonic are up to sneaky things –
they seem to think that people cannot
remember anything that’s older than
10 years and so now they are recycling
their television model numbers (of
course, in those days they were called
National so it doesn’t really matter).
This is extremely confusing for
old codgers like myself who can still
actually remember the original model.
In this particular case, it was a model
TC1401. The original cost over $550
and was an extremely heavy, 14-inch,
portable TV set with a white cabinet
and rotary tuning (VHF only). By
contrast, the beast on my bench, also
a model TC1401, was a dual-speaker
34cm set with remote control and
a black cabinet. It weighs just 12kg
and is so light that the owner has to
be careful they don’t lose their grip if
they are carrying it in a heavy wind!
The complaint with this one was
“no memory” – I’m glad I am not the
only thing that suffers from this fault
from time to time!
When the set was switched on,
the only channel available was 1
and on opening the “Preset” menu
all the channels were designated as
“skipped”. They could be retuned in
any configuration you wanted, with
the correct TV channel numbers, but
when the Preset button was pushed
again all the information was lost.
So the fault description at least
was accurate and seeing as there is
an IC (IC1104, MN12C25D) marked
“MEM
ORY”, the obvious thing to
do was to replace it. I ordered in the
chip, fitted it and switched on. Initially, this seemed to have fixed the
problem because the set worked but
then, after about five minutes or so, it
failed again. Curses – it wasn’t going
to be that easy.
Next, I checked all the voltage rails,
especially the 5V and 30V rails, but
they were all OK. I even checked them
for ripple using an oscilloscope but
there was none. I also checked all
the other voltage rails before going
on to measure the voltages on all the
pins of IC1104 and IC1102. There was
nothing was untoward – not even a
dry joint.
By now thoroughly frustrated, I
reordered another MN12C25D IC plus
an MN151142TEA (IC1102), if only
to make sure that it really wasn’t one
of these chips. When they arrived, I
couldn’t wait to fit them one at a time.
IC1102 was a bit tricky as it is a 42-pin
high-density chip but eventually they
were fitted properly and this time the
symptoms were . . . exactly the same
as before! This was definitely not the
result I was looking for!
What next?
By now, I was totally perplexed
by all this and was contem
plating
abandoning the repair. Maybe a cup
of coffee would get the ol’ brain cells
working again? Well, maybe it did
Sets Covered This Month
•
•
•
•
•
Panasonic TC1401 TV set
Teac CTM-143 TV set
NEC FS-6325 TV set
AWA CT-1447AM TV set
Mitsubishi HS-338A VCR
help because I was closely examining the main board (rather hatefully)
when I noticed that the small glass
diodes fitted to it were completely different to the diodes fitted in modern
TVs (like 1N4148). And as I quickly
discovered from the circuit, that was
because they were completely different. Surprisingly, they were germanium OA90s, a diode rarely seen these
days except in the odd discriminator.
Yet here was a fairly modern set using lots of them and – hey-ho – they
were predominantly in and around
the memory circuit I was working on.
Well, this was all academic and of
purely historical interest except, of
course, one of the main reasons for the
shift from germanium to silicon was
reliability. Silicon diodes are much
more reliable than their germanium
counterparts and have better (read
“less”) leakage. The only drawback
with silicon is that it requires 0.6V to
bias the junction, as opposed to 0.2V
for germanium diodes.
Well, I didn’t have any better ideas,
so I began testing each diode in circuit
on the x10 ohm range of my multimeter. The reverse leakage varied a lot
between them but at least there was
a difference between the forward and
reverse directions until I got to D1129.
This diode connects pin 5 of IC1104 to
pin 20 of IC1102, the microprocessor.
D1129 measured nearly open circuit
in both directions and it just had to
be the culprit.
I rummaged around an old miscellaneous diode box and found an
OA90 and fitted it. And that solved
the problem – everything was now
properly stored and remained that
way even after the set was switched
off. Without the benefit of a block
diagram, I cannot tell the precise
function of D1129 except to say it
connects C2 with P32-C2 (I’m sure
you are all the wiser for that bit of
priceless information).
I have to admit this repair was
pretty fluky and I’m now off to buy
MAY 1999 29
a lottery ticket. Maybe I’ll crack the
jackpot?
Unhappy customer
Mr Burton wasn’t too happy about
his Teac CTM-143 TV set. It still had
the same fault as when I’d fixed it
last time, or so he claimed. Well, he
may have genuinely believed this but
it was 1996 when I fixed it last and I
don’t give 3-year warranties. And as
it turned out, it wasn’t the same fault
as last time.
On this occasion, the set was intermittently not coming on and it appeared again to be a problem with the
line drive stage in the 34cm Goldstar
PC04A chassis. The previous fault
allowed the set to start but it would
then “go off” after a short period of
time. This was caused by D402 in
the 27V rail being open circuit. This
allowed the driver stage power to start
via D401 (18V) but because D402 was
open circuit, the stage would then
shut down.
This time, the voltage on the collector of Q401 (KTC2230A) was 20V.
Sometimes there was a kind of square
wave on this collector, while at other
30 Silicon Chip
times the waveform collapsed into a
reduced waveform with large negative
spikes, which in turn produced a
waveform on the secondary of T401.
This was insufficient to turn on Q402
(KDS1555), the line output transistor.
This was baffling because the square
waveform on the base of Q401 seemed
adequate to turn the stage on and, of
course, it would have to be intermittent, just to complicate matters.
To eliminate any traces of the
problem I had addressed last time, I
connected an external variable power
supply to the junction of C404 and
T401 and pumped in 18-28V. It made
no dif
ference, thus eliminating the
power supplies. The square waveform WF2 was correct at all times. I
replaced Q401 and Q402 and the fault
went away for one week but it was
back again just after I had confidently
given Mr Burton a quote for the fault.
Next, I checked all the components
in the collector circuit of Q401 and in
the base circuit of Q402, to no avail.
So what was I overlooking? Basically
Q401 is biased and switched on by
the square wave arriving via C401
(which had also tested OK) but the
waveform became distorted on its
collector. Why? It’s always the way;
the simpler the circuit, the harder it
is to find what’s wrong.
The vital clue came when I monitored the waveforms with an oscilloscope. This showed that the
amplitude of the waveform was
much greater before R402 than after
it. Certainly, the difference was much
greater than I expected, considering
that R402 is nominally only 560Ω.
When I removed R402 from the
board and measured it, I found that it’s
value was actually 750Ω, an increase
of almost 50%. Replacing this resistor
increased the waveform amplitude
at the base of Q401 and the set now
remained on. I soak tested it for a
week and crossed my fingers when
Mr Burton collected it.
No TimTams
It was a hot day and I was praying
that Mrs Norris’ NEC FS-6325 TV set
was going to be straightforward. I was
running late because the previous job
had taken far too long, due mainly to
the client’s addiction to talking – she
could talk the hind leg off a donkey!
Serviceman’s Log – continued
Anyway, if this next job was easy, I
could still make it back to the workshop in time for a leisurely cup of
coffee and a couple of “TimTams”
before knocking off for the day.
When I arrived, I quickly unscrewed the back and it was easy to
see why the set was dead – F601, a 2A
mains fuse, was as black as the ace of
spades. I unplugged the degaussing
coils and measured the resistance
across the bridge rectifier – it was
still nearly a complete short circuit.
By following the path from the bridge
rectifier, I soon established that IC601
(STR
41090) was short circuit and
hopefully the cause was due to the
obvious dry joints on C609, the main
tuning capacitor.
As luck would have it, I had a new
STR41090 in the van. I quickly ran
out, found it, shot back into the house
and replaced it before you could say
“Micky Finn”. I then switched the
set on and prayed hard but nothing
happened. It looked as though I was
snookered.
Suitably chastened, I turned the set
off and measured the main HT voltage
across C609. It was still 340V
which told me that the circuit
was stable DC wise but wasn’t
starting up. I desperately
looked around and saw a
1MΩ resistor (R607). Hoping
that this was the critical startup bias resistor, I replaced
it and switched on again.
Much to my frustration, there
was still no response – my
TimTams were melting away
as in a mirage.
That was when I spotted
that R610 had a little chip
missing from its body. There
was enough of it left to
determine that it was once
a 1Ω resistor. I measured
Q601 (sandwiched between
R610 and R607) to find that
it was short circuit as well.
I tore back out to the van,
rummaged through the mess
in the back, found the parts I
needed (miracles do happen)
and rushed back inside and
fitted them.
This time, the set fired up
and the picture and sound
32 Silicon Chip
were good. Thank you God, thank you.
I quickly scribbled out the bill, put
the back on the set, hopped in the van
and shot back to the workshop. I went
straight for the percolator the minute
I got back. Great; there was still some
coffee left but what – NO MORE TIM
TAMS. Life lost all its meaning!
Predictable trouble
Mike Tester’s (the name is changed
to protect the guilty) AWA CT-1447AM
was always going to be trouble. You
see, Mike always fancied himself as
a technician and he also lived near
the sea. This combination meant his
set was always breaking down and
he was the kind of guy who liked to
have a go. Well, this time there was no
picture but the sound was good and
there was a raster with the on-screen
display working OK.
Anyway, it looked as though he had
lost only the video between the IF
detector and chrominance/luminance
decoder IC. As he is a good friend of
mine – despite his foibles – I tried to
help him over the phone but I really
didn’t have a clue as to the exact cause
of the problem. Initially, I told him to
check all the voltage rails, especially
those feeding the signal circuits. This
he dutifully did but everything measured fine, so I told him that the only
course of action was to feed in a signal
from a colour bar generator and trace
it through with an oscilloscope. After
a bit of coercion, he finally agreed and
dropped the set off at my workshop.
When I removed the back, I could
see how rusty the whole set was from
the salt air. I started by confirming
everything he did by checking all the
voltage rails. These all proved OK,
so I hooked up the oscilloscope and
followed the video from the video
detector (pin 10 of IC101, MS51496P)
through to Q1A0 TP12 (waveform 1),
thence to LC201 DL/BPF and finally
to pin 18 of IC201 (M51412SP). After
that the scent became very cold.
I then spent an inordinate amount
of time examining the contrast control
circuits but got nowhere. By now I
was beginning to think that the fault
was somewhere in the beam limiting
circuit. I started at pin 8 of the flyback
transformer and traced the circuit
until I got to the two beam limiting
test points designated PT1 and PT2.
It was then that I noticed that R555
and R556 were badly corroded. I de
soldered them and measured
them to find that they were
both nearly open circuit.
Replacing them fixed the
problem completely but I had
to warn Mike to keep the set
dry, otherwise it wouldn’t
last very long. Unfortunately,
he didn’t listen too well and
within another three months
the set was worse than before
and he was forced to bin it.
A Heath Robinson job
Mrs Daniels, a widower
living in a housing commission flat, was a very keen soap
watcher and loved to record
her serials every day on her
beloved Mitsubishi HS-338A
video. She first brought it
in complaining of poor fast
forward and rewind, which
just turned out to be belts and
tyres, but a month later it was
back. This time, the complaint
was “snowy pictures”.
At the time, I felt sure it
was just dirty heads but after
cleaning them vigorously I
came to the conclusion that the heads
were worn out, especially as (with the
same tapes, at least) I was getting almost clear pictures with Pause/Freeze
Frame/Still. I removed the heads and
checked them on my tester to find
that they were indeed low – enough,
I thought, to be causing the problem.
When faced with the news, Mrs Daniels was very stoic, accepting that as it
was in use every day, the heads were
bound to wear out eventually. And
although she could hardly afford new
heads, she would find a way to come
up with the money as it really was her
main source of entertainment.
I ordered in the new heads, fitted
them and confidently switched the
machine on. To my horror, I found
that the problem was just as bad as
before, although the picture was still
OK when paused. It was obvious that
I had misdiagnosed the fault.
I got the CRO out and examined the
FM envelope at TP-2A to find half of
it missing. It was unlikely to be the
new heads but it could be the head
amplifier IC, the switching pulse or
worse still, the toroidal transformer
inside the drum itself.
Using the second channel of the
CRO, I quickly established that the
switching pulse (FF or flipflop on pin
2 of IC201 M51473P) was exactly in
phase with the FM envelope switching. From there, it didn’t take long to
find that the toroidal transformer primary measured 100kΩ between pins 1
and 2 of plug SB. I removed the entire
drum assembly, then removed the
drum motor and upper transformer to
reveal the lower primary coils glued to
the bottom with – yes, you’ve guessed
it – the notorious brown glue.
There was nothing that could be
done to fix it as the coils were only
accessible on the underside and the
transformer was glued too tightly to
the base. Unfortunately, the trade cost
of a complete drum assembly was a
prohibitive $416.18 (if indeed it was
available), so I tried to obtain a junked
machine from one of my colleagues
in the trade.
Two heads or three
When I enquired, one young technician asked me whether it was the
3-head version (which it is) or the
earlier 2-head HS337A. At first, I
didn’t quite realise the significance of
his question but he went on to suggest
that I substitute the pause head and
Fig.1: this diagram shows how the connections to the transformer windings
inside the drum were modified. Fig.1(a) is the original circuit, while Fig.1(b)
is the modified circuit. Fig.1(b) also shows how the leads to the heads were
modified on the top of the drum.
Fig.2: here’s how the connections to the
head terminals on the top of the drum
were modified. The pins were desoldered from the PC board at all points
marked A and B and the two pins at
A then connected to pins C using short
lengths of insulated wire.
This photograph shows the modified
drum assembly. Amazingly, it worked
and produced quite a good picture.
its winding for the open circuit winding. At first I thought that this was an
absurd idea, knowing the tolerances
these heads are made to, but having
had no success in obtaining a second
hand drum assembly, I decided to
at least give it a try.
First, I completely reassembled
the drum, refitting the old heads in
the process. I then fitted two jumpers across the toroidal transformer
primary windings, connecting
the pause head on winding (L1)
in parallel with the open circuit
winding (R). There was an added
complication in that the open circuit winding (R) shared the centre
tap with the good winding (L).
I now had to guess which head
was which on the upper drum, as
they are not marked anywhere.
The pause head is L1 and I reasoned that this would be mounted
close to play head L, which would
be diagonally opposite head R. I
then unsoldered head R, fitted two
links to the pause head winding
(L1) and tried it out. It made no
difference, which meant that I
probably had the L and R heads
mixed up.
Next, I assumed that the heads were
arranged as shown in Fig.2, with the R
head adjacent to the pause head (L1).
I then rearranged the leads as shown
so that the active heads were L and
R but I still didn’t really expect it to
work. However, I was thrilled to see
that it actually did work and what’s
more, the pictures were pretty good.
Even more surprisingly, the pause
mode wasn’t bad either.
I then made a recording and played
it back and the picture was still quite
acceptable. Frankly, I was amazed that
this had worked at all and I am full
of praise for my friend. Obviously, it
isn’t perfect but Mrs Daniels thought
it was acceptable under the circumstances, especially as I didn’t charge
her for the new heads and put them
SC
back into stock.
MAY 1999 33
|