This is only a preview of the March 1997 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. Items relevant to "Plastic Power PA Amplifier":
Items relevant to "Build A Jumbo LED Clock":
Items relevant to "RGB-To-PAL Encoder For The TV Pattern Generator":
Items relevant to "Audible Continuity Tester":
Articles in this series:
Purchase a printed copy of this issue for $10.00. |
SERVICEMAN'S LOG
The rich tapestry of servicing
What makes a non-technical person fiddle with
his VCR when something goes wrong? And why
do young children like “posting” money into the
cassette wells of VCRs? It’s all part of the rich
tapestry of servicing.
The tall, distinguished looking
gentleman who wandered into the
shop clutching his prized video to his
bosom didn’t look the type to have a
go – he looked more like a lawyer than
a serviceman. Anyway, I wasn’t presumptuous enough to enquire about
his profession; instead, I politely asked
him for his particulars and asked what
was wrong.
He explained that the tape would
go in and down and wrap around the
drum motor but it wouldn’t play, fast
forward or rewind. He freely admitted
that he had had the covers off and so I
decided to carry out a few preliminary
checks while he was there. I connected
the machine – a Teac MV505 –to the
power and pushed in a tape to confirm
what he had said.
His description of the problem was
spot on but that wasn’t all. I also found
that the tape wouldn’t eject because
the cassette flap wouldn’t open, which
meant that he had also removed the
front escutcheon and not replaced it
properly. This he sheepishly admitted
was the case.
After he left, I removed the covers
and the front panel and re-engaged the
door flap lifter so that the tape would
now eject properly. Anyway, that was
only a minor detail; I now had to track
down the main fault.
Preliminary checks
The deck, surprisingly, was a Mitsubishi Fo swift mechanism and I
could see that the tape was not lacing
up fully. The drum motor was spinning
but there was no sign of life from the
capstan motor.
My preliminary diagnosis was that
something was wrong with the loading
mechanism. But what? Was it jamming? Were the gears out of alignment?
Was it the timing? Or was it a faulty
mode select switch?
I began by inserting a tape and when
the loading motor stopped with the
tape 3/4 laced up I continued to rotate
it by hand, consciously feeling for any
resistance. I couldn’t feel any and I so
I continued to turn the motor until the
arms were almost completely laced,
at which point it would unload itself.
Because the loading motor turned a
squirrel gear, I concluded that this test
may be misleading. Because of the gear
ratio, I would not necessarily feel any
resistance at my fingertips.
My next step was to check if there
was anything preventing the arms
from completing their travel to the
end stops. They seemed quite free and
loose and so I concluded that either a
52 Silicon Chip
gear had jumped a tooth in the loading
gear chain or the mode select switch
was at fault.
Unfortunately, as I discovered when
I removed the bottom cover, it’s not
easy to check the gear alignment as
there is a printed circuit board covering the master cam, along with several
sliding plates. However, the mode
select switch is easy to access and
so I decided to check that first. This
switch is soldered to the PC board via
five connections and there is an alignment point which must marry up in
the eject position.
I replaced the switch but there
was no improvement in the loading
sequence. Regretfully, it looked like
major surgery was required and I was
extremely grateful that I had a full
set of instructions for this deck, even
though these were for a Mitsubishi
VCR.
It is hard to summarise the next
hour of invective and bad language.
The instructions make it all sound so
easy and I suppose it is if you work
on this deck all day every day. If you
don’t, then it’s not quite so straightforward.
Anyway, I removed the reel belt,
capstan brake spring, cam plate B,
three gears, the loading gear arm and
five screws, before desoldering the
leads to the full erase head. At this
point, the deck PC board is ready to be
prised off – at least in theory. However,
on this model, the lower moulding
that supports the deck is somewhat
generous and the PC board won’t come
out unless the whole deck, including
the ejector, is removed.
A closer inspection revealed that it
would be necessary to remove about
6mm from either side of the moulding
to free the PC board. As a result, I
decided in the interests of time that
an Australian modification was required and so I used a soldering iron
to melt away the plastic so that the
board could be removed (this didn’t
alter the strength of the structure in
any way).
Finally, I had access to the main
cams (1 and 2) and, after wiping away
the excessive grey grease, I could check
the alignment hole. Would you believe
that all was correct? The shafts and
levers were all in the right places. I
removed the cams and carefully examined them on both sides for broken
or bent galleries but all were perfect.
Even their teeth were straight.
Worse, naturally, was to come. Any
damn fool can take things to pieces –
it’s getting them back together properly
that sorts us out. Inevitably, I fell for
all the traps, in particular the record
safety lever which should be held
back whilst inserting the board, not to
mention pin “e” getting in the wrong
track of the cam slide plate B.
On the third attempt, it all finally
came together and we were back to
square one with the original fault.
It was now that I had a little bit of
well-deserved luck.
Whilst cogitating menacingly over
this vile mechanism, I noticed that it
had been fitted with a new green pinch
roller. Now the original pinch roller
arm mechanism was made of white
plastic and it is common for one of the
arms that guides it down the squirrel
gear cam to break.
This is replaced by the green type
which you can either purchase as a
single part or as part of what is called
“Abrasion Part Kit for Fo DECK (Rubber)”, whatever that means (the part
number is 789C007020). This kit comprises the arm, the reel belt, the circlip,
a sachet of grease and a comprehensive
instruction booklet.
However, if a serviceman doesn’t
March 1997 53
Serviceman’s Log – continued
know about this kit (it isn’t mentioned
in any service manual) and only fits
the new pinch roller, he usually also
neglects to clean and lubricate the
shaft it slides up and down on. And
that’s precisely what had happened in
this case. In operation, the pinch roller
started to slide down the shaft but it
was too slow because of the friction
and it was jamming the roller against
the top of the capstan shaft housing in
a way that wasn’t obvious to the eye.
Cleaning and lubricating it with grease
fixed the problem completely.
Anyway, when our lawyer (?) friend
arrived to pick it up, I asked him about
it and he confirmed that the pinch
roller had indeed been replaced just
over a year previously. What a palaver
over what, in hindsight, should have
been a straightforward simple repair.
Christmas treat
After a long cup of coffee, I tackled
the next job, praying it would be easy.
Oh the joys of Christmas – the presents, the new VCR for Dad, the odd
bit of cash for Johnny the 5-year old.
And oh what a disaster this combination can make!
54 Silicon Chip
Mr Grey brought in his Akai VSG220EA, with a tale of woe that his
youngest son had “posted” some toy
or other into it. Of course, it no longer
worked and when I shook it I could
clearly hear something rattling inside.
The coffee had definitely improved
my mood and I chose Mr Grey’s still
shiny VCR – just barely out of the egg
(I think) – to look at next.
Removing the cover gave good
access to the mid-decked VCR and I
quickly found two coins – a 10-cent
piece and a 5-cent piece – sitting on the
PC circuit board just under the deck.
I retrieved the two coins by carefully
jiggling the deck upside down in the
air, then carefully checked for more
and for any signs of damage before
powering it up.
When I switched it on, the drum
motor came to life briefly but no other
signs of life were present – not even
from the display. I pushed various
buttons and nothing happened but
when I pushed a prerecorded tape
in and pressed play, the tape loaded
normally and a picture appeared on
the TV with full sound.
I pushed all the buttons in turn and
it paraded its full box of tricks. In fact
everything was working except the
display. Unfortunately, I don’t have
the service manual for this particular
model, which meant that I would have
to tackle it blind.
Fortunately, the deck isn’t too difficult to remove. It’s simply a matter of
removing five screws and three plugs,
removing the front escutcheon/control
panel, and then lifting the deck out
vertically. This gives access to the PC
board which is held in via two screws
and half a dozen clips.
Unfortunately, the clips make it
rather awkward to remove the PC
board assembly but eventually I was
able to free it and lift it out from the
rear. This done, I gave it a careful
visual inspection but nothing obvious
was shouting back at me so I applied
power to the board and began checking
voltages around the circuit.
Because there was no display, I
reasoned that the supply rail to it had
probably gone missing. Either that or
the display itself, or possibly the microprocessor that drives it, had been
damaged.
My initial checks revealed that a
voltage was present between the two
ends of the display where the filaments
are connected. This is typically either
5V DC or 5V AC. Having found this, I
checked various other points around
the display, looking for a -28V (approx.) rail, but there was none. I didn’t
have a circuit diagram which was a bit
of a hindrance but it all screamed of a
failed -30V rail from the switchmode
power supply. All I had to do was
identify it.
There are about 16 diodes in the
secondary of the power supply, most
of which are protected by low-value
resistors. Unfortunately, no voltages
were marked on this part circuit but
it didn’t take a mental giant to figure
out which diodes were in the negative rail, as their anodes connect to
the negative side of an electrolytic
capacitor.
Anyway, I checked each of these in
turn and eventually found that D209
was open circuit. I replaced it, plugged
the machine into the wall socket
and was immediately rewarded by a
flashing “AKAI” sign in the middle of
the display. Switching on the power
at the machine then brought up the
word “ERROR”, which is normal at
this stage.
Getting it all back together again
was surprisingly simple, with the PC
board literally falling into its supports.
The deck accurately followed suit and
I powered it up with a tape in place.
This time, the display worked correctly and I gave it a thorough soak test
before calling the customer.
The house call
After lunch, I was asked by a little
old lady to do a house call on her
aging Sony KV2764EC which had
sound but no picture. This set is now
about 10 years old and not getting any
younger. In fact, assuming average
use, this is about the “use by” date of
a TV receiver.
I don’t like doing service calls on
these sets as access is to the main
circuit board is quite poor. However,
she couldn’t possibly bring the set to
me, so I had to go to the set!
When I got there, I found that the set
was on a low table near a window and
so the lighting was good. I switched
the set on and the symptoms were as
described – sound but no picture.
This set used a PE3 chassis rather
than the Rx chassis. The tube filament was alight and you could hear
the familiar rustle from the EHT at
switch on.
As a first step, I measured the screen
voltages on pins 3 and 4 (G1 & G2) of
the tube. These were both around 500V
which is what I would expect them to
be. However, the cathodes were too
high at nearly 200V. I then switched
the set off and it momentarily flashed
a white line. Ah, ah, I thought – a
vertical deflection failure.
When I finally managed to remove
the motherboard, I could see what
looked like a number of dry joints
and even though I worked them over,
I knew that the problem just had to
be IC552 (TDA3652). However, this
IC is no longer available and is now
replaced with a TDA3654. At the same
time, you also have to replace R518
(6.8kΩ) with a 1.5kΩ resistor.
As this is a well-known problem, I
had the parts on boards that I carried
with me and installing these quickly
restored the picture. However, on departure, I advised her to start saving
for a new TV.
Portable players
Back at the ranch, a couple of Pye
ND-20 portable CD cas
sette stereo
radios had come looking very much
the worse for wear. My instructions
from the owner’s financier, namely the
father of two teenagers, was to make
one good unit out of the two.
Much as I hate working on these
cheap units, I reluctantly agreed to
have a go. Gaining access to the circuitry of such units if often a problem
but in this case, the two halves of the
cabinet shell can be split after removing only eight screws.
One of the sets was completely dead
and I decided to work on this first,
as I hoped that it would be a simple
power supply problem. In the event,
the power was OK and I could trace
the problem to the cassette deck and
function switch SW2. Unfortunately, I
didn’t have a circuit and so I couldn’t
quite work out the sequence of events
from there on in.
The function switch was fairly
complex and as power was going in
but not coming out, this had to be the
logical suspect.
To confirm this, I wiggled and twisted the switch and sprayed contact
cleaner inside it until finally I managed
to get some sound out of the speakers.
That was enough to confirm my theory
and I placed an order for a new switch
right away.
What about the other machine?
Well, it had a smashed control panel
and for a while I contemplated removing its function switch for use on the
other machine. However, it wasn’t
worth the time that would be spent
removing and refitting it, particularly
without knowing its condition. Of
course, the other option was to repair
the smashed unit with parts from the
other machine but the hole was too big
and the damage too severe.
Tarzan’s TV
My next customer was a young man
who arrived in a small 3-door hatchback. To my astonishment, he removed
a 63cm stereo TV from the back seat
and effortlessly carried it into the shop
as though it was an empty cabinet. He
plonked it down on our small counter
and cheerfully informed me that there
was no picture.
The set was a Philips 2B-S chassis
KR5987R 25CT8883/75, circa 1988.
This was a fairly popular model and
is one that I am quite familiar with.
March 1997 55
Fig.1: the video control chip in the Philips 2B-S chassis. As the set ages, it
is sometimes necessary to add a 33kΩ resistor between the +13a (12V) line
(pin 6 of IC7300) and pin 26 (RGB output stage current sensor).
Not bothering to even catch his breath,
“Tarzan” continued to elaborate on
the set’s problems. Apparent
ly, the
picture had been a little “unclear” and
intermittently had taken longer and
longer to come on.
Initially, I was rather reluctant to
take the set on, as I didn’t fancy the
prospect of it taking up so much bench
space for a week or two while I chased
an intermittent fault. I mumbled that
it could take quite some time and
suggested that, in view of the set’s age,
he might prefer to spend his money
on a new set.
His response was that he wanted
the set fixed and that I could take
as long as I liked. It was the wrong
response but still, you couldn’t help
liking him for his cheerful manner. In
then end, I relented and promised to
start on it straight away. All he had
to do was move it to my workbench
and promise to pick it up as soon as
it was ready.
No problem – Tarzan made the set’s
removal look as though a genie had
instantly answered my wish. If only
he could have fixed it too!
When I switched it on, there was
no picture but both sound and EHT
were apparent and the CRT filaments
lit up. Now, this chassis will not give
a picture until the beam current has
56 Silicon Chip
reached a certain level. This feature
is achieved using a video control chip
(TDA4580/V2 – IC 7300), which also
deals with brightness, contrast, saturation, beam cutoff stabilisation and
beam limiting.
However, a problem arises as the
tube ages, in that it takes longer and
longer for the picture to come on. Often, there is only a white line at the
top of the screen but this can usually
be fixed by adding a 33kΩ resistor
between the +13a (12V) line (pin 6 of
IC7300) and pin 26 (RGB output stage
current sensor).
Unfortunately, this same symptom
(ie, the white line at the top of the
screen) can also be produced by a
variety of other faults. And, in fact,
it showed up within a minute or two
from switch on.
Access to the PC board in this set is
not good and I find the best approach is
to turn the cabinet on its side, or even
upside down, to get to it. The first thing
to tackle when you do get access is to
remake any suspicious-looking joints,
particularly around the transformers
and ICs. In this case, it wasn’t too bad
and my efforts made no difference to
the problem.
Next, I replaced C2571, a 100µF
electrolytic capacitor in the vertical
output stage. This is something that
I always do as a matter of course
with these sets, as past experience
has shown that this capacitor can
give problems. Again, it made no dif
ference.
Finally, I stopped working as an
automaton, put in my re
m aining
braincell, and started measuring
voltages and checking waveforms.
First, I checked the 1.2V nicad battery
which was OK (this battery backs up
the memory for the microprocessor).
This done, I checked gating pulse
waveforms 43 and 44 to the chroma
decoder (pin 8M10), to pin 9 of IC7550
(TDA3870/V2), to pin 10 of IC7300
(TDA4580/V2) and to pin 7 of IC7570
(TDA3654Q).
I was drawing blanks everywhere,
so I decided to go back to first principles and examine the CRT voltages.
And as luck would have it, I found
the cause almost immediately. As
soon as my 100kΩ/V analog multimeter touched pin 7 of the CRT socket,
the picture came on and stayed on.
There was just one problem – it was
out of focus.
The voltage on pin 7 of the CRT is
marked as 650V and is derived from
the flyback transformer 7kV connection via the focus and screen control
pots. And, fairly obviously, the very
small current through the meter was
necessary to make the focus control
function.
Replacing the focus control pot
(33MΩ) fixed the problem. Unfortunately, the picture still wasn’t the best,
even after adjusting the focus, and I
suspect that the emission was down.
However, it wasn’t down far enough
to justify the modification mentioned
earlier.
When I stripped down the faulty
pot, I found that the printed circuit
on the ceramic base had corroded.
Apparently, the extra leakage of the
meter was enough for the voltage
to arc across the corroded section. I
might add that, on some sets, I have
also had problems with the screen
control (4.7MΩ), especial
ly on the
KT3 chassis, and with C2471, a 68nF
capacitor which connects from the
wiper of the screen control to the
+200a voltage rail.
Tarzan, true to his word, turned up
not long after I made the call, tucked
the set under one arm (well, not quite),
and plonked it in the back of his tiny
car. Fortunately, he was as happy as
SC
Larry.
|