This is only a preview of the August 1997 issue of Silicon Chip. You can view 34 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:
Articles in this series:
Items relevant to "A 500 Watt Audio Power Amplifier Module":
Items relevant to "Build A TENS Unit For Pain Relief":
Items relevant to "PC Card For Stepper Motor Control":
Articles in this series:
|
SERVICEMAN'S LOG
Just give it a flamin’ good thump
Most jobs are fairly routine but these three
jobs hardly fall into that category. Of course,
the customer doesn’t always help and the one
job that should have been straightforward
was complicated by the customer’s rudeness
and uncooperative attitude.
A month ago, I was asked by Mrs
Johnston to attend to her Sanyo TV.
The problem didn’t sound like much
in that most of the time the set worked
perfectly. It was just that, every three
days or so, it wouldn’t start although
Mr Johnston could “fix” that by
thumping the back. But it was an-
60 Silicon Chip
noying and now it did it all the time.
The set was a Sanyo CPP2601SV-00
employing an 84P-B26 chassis and
is a large 63cm stereo remote control
model. It all sounded so easy – just
whip the back off, fix the obvious dry
joint or loose plug, replace the back
and Bob’s your uncle. A piece of cake
really and what’s more, it should take
no more than half an hour. I should
have known better.
I arrived at the appointed time full
of optimism and had the back off, the
soldering iron switched on and the
meter ready before the tea arrived. The
TV switched on perfectly and did so
for the next 20 subsequent efforts. Mrs
Johnston assured me that it always
played up almost straight away but
I know all about the contrariness of
inanimate objects.
I left it on while I finished my tea
and then tried again. It still worked
perfectly. I felt like Homer Simpson –
TV goes on, TV goes off, TV goes on,
etc. The time was slipping by, with
no sign of a problem at the end of 30
minutes. It was time for a different
approach.
Using a strong flashlight, I examined every board very carefully but, as
usual, access to the chassis was rather
difficult. The leads to the front panel
are far too short and the chassis could
only be moved about 80mm out of the
cabinet. Not only that, but there is a
plastic support frame which obscures
the main PC board.
I decided to gently tap each board
in turn but that made no difference.
That was it – I couldn’t afford to waste
any more time. I refitted the back,
pushed the set back into its corner
and switched it on for the last time.
The set came on, to our mutual frustration. Mrs Johnston then had a go
herself with the remote. She switched
the set off and it wouldn’t come back
on. How did she do that? I tried the
remote and the main on/off switch –
the set was completely dead.
I was now too far committed to
back away, so off came the back and
we both tried switching the set on
and off again. And would you believe
it? – it now worked perfectly. I then
replaced the back, carefully noting
where it made contact with the chassis in case it was dislodging a plug or
socket. This time, with the back on,
Fig.1: the power supply circuit in the Sanyo CPP2601SV-00 63cm colour TV set.
the set worked perfectly, no matter
what either of us did.
Three quarters of a hour had gone
by now and we both had other things
to do. Obviously, this was a job for the
workshop and I advised the Johnstons
accordingly. As it turned out, they
were going on a fortnight’s holiday a
few days later and so we organised for
the set to be delivered to the workshop
before they left.
The fault appears
Finally, the set was on the bench
and the back came off yet again. Actually, the back wasn’t in very good
condition. It is made of a very brittle
hard plastic and had cracked in a few
places where Mr Johnston had been
hitting it. He must have been giving it
a frightful whack on occasions!
It took three days of continuous
running before the fault finally appeared. And when it did, it didn’t
take too much of a tap to make it work
again for another three days.
After about two weeks, I finally
deduced that the problem was somewhere in the power supply although
I still didn’t know the exact nature
of the fault. The standby light could
always be made to come on with the
remote control, which meant that the
+5V and +12V rails were fine from the
Power Sub unit. And during one of the
short periods while the set was not
working, I measured approximately
+325V on the collector of Q311 but
found that there was no output at all
on the five secondary power rails.
Because it took so little vibration to
make the set work again, all measurements had to be made with extreme
delicacy. Eventually, I had precious
multimeters and a CRO permanently
hooked up all over the set, waiting
for it to play up. This was a real nuis
ance as it left me with only a limited
amount of test equipment for fixing
other things.
My next step was to establish that
the optocoupler (D311/TLP632) was
working correctly. I found that there
was +1V applied to pin 1 (on the LED
side) when the set was off and 0V
when the set was on. And there was
0V across the transistor on the secondary (hot side) of the optocoupler
when the set was off and -20V across
pins 4 & 5 when it was on.
This implied that the optocoupler
was working correctly. What’s more,
leaving the meter connected between
pins 4 & 5 of this device seemed to
“correct” the fault because the set
always started reliably while ever it
remained connected. Obviously, the
extra drain of the meter was having
an effect on some part of the circuit.
My two weeks were nearly up and
something had to be done. I placed
the set on its side for better access
and shone a bright lamp onto the PC
board. Freezing the various plug and
socket contacts didn’t seem to make
any difference and they all looked
perfect. The components that did
have a reaction to the freezer were the
optocoupler (D311), chopper transistor Q331 and electrolytic capacitors
C327, C328 and C330. I changed them
all and reworked the soldering on the
entire board but, after three days, it
played up again.
Next, I replaced A301 (a JUO168
thick film IC which functioned as an
error amplifier), along with resistors
R306 and R307, the latter used to bias
Q311. It made no difference and I was
now only left with a few components
that hadn’t been changed.
I was fairly satisfied that it wasn’t a
dry joint or a hairline fracture in the
board and I was also fairly sure that
the secondary (cold side) of the chopper transformer (T301) was OK. Of
August 1997 61
Serviceman’s Log – continued
course, I hadn’t checked the chopper
transformer itself, mainly because I
had regarded it as an unlikely culprit.
However, as I was now running out
of ideas, I removed it and carefully
examined it, especially where the
leads are wrapped around the posts.
I could find nothing untoward, so
I resoldered all the wire wraps and
replaced the transformer. And that
appears to have fixed the problem.
Despite a further week of intensive
testing, there was no sign of the fault
and I can only conclude that one of
the solder joints associated with the
transformer must have been at fault,
this despite the fact that they all appeared to be OK.
The only other possible culprits
are D332, D333 and C335 but none
of these ever responded to tapping or
freezing. I admit that it’s rather an unsatisfactory conclusion but I believe
that the problem has either been permanently fixed or, at least, postponed
for quite a while. Why can’t they make
TV sets easier to service?
Of course, the Johnstons will never
62 Silicon Chip
appreciate how much effort I put in to
overcome this obscure problem. But
as Sanyo put it, that’s life!
The crook VCR
Mr Nasty brought his Samsung VCR
in just after Christmas, complaining
that it wouldn’t play. This was a
VB-306 Winner mid-drive unit and
when I removed the covers and inserted a cassette, it was pretty obvious
where the fault lay. The loading arms
wouldn’t move at all and it could only
fast forward or rewind.
Unfortunately, there is no access
underneath the deck and the only way
to get to the mechanism is to remove
it, which is what I proceeded to do.
This involves removing countless
screws from the top and bottom, along
with the front escutcheon, before the
deck can be unplugged from the main
PC board.
Removing the loading motor assembly on the underside of the deck
reveals a large rack gear which engages a master cam. And you didn’t need
to be a genius to figure out what was
wrong. The first tooth of the rack was
missing and the teeth on the cam gear
were all damaged. Obviously some
force had been applied here to cause
this. The gears were relatively cheap
but, after adding in the freight and my
labour to remove and replace them,
the customer was looking at a bill of
about $100.
I may have thought that this was
good value but not so Mr Nasty. Instead, he flatly contradicted me when
I called him with the news and almost
implied that I was being dishonest and
attempting to overcharge him.
I left the machine disassembled
so that he could see the problem for
himself when he called to pick it up.
He was even more displeased with
this and I received neither thanks for
my diagnosis nor any payment for
the time I had spent on the machine.
I was quite surprised at his rather
disagreeable attitude and thought
that that would be the end of it. I
had more or less forgotten about
the incident when suddenly, after
about six months, his wife brought
the machine in and asked for it to be
repaired for the figure I had quoted.
(Don’t you just love some of these
guys? They back themselves into a
corner by being obnoxious and then
hide behind their wives after they’ve
thought better of it).
Taken aback by this sudden about
face, I carefully examined the parts
to see if anyone else had had a go at
the machine after it had left my shop.
However, everything appeared to be
as I’d left it and so I reluctantly agreed
to take the job on. I don’t like dealing
with customers who have been unreasonable in the past but I reasoned
that it would be a straightforward
job and I would be able to recoup my
previous losses.
Removing the old rack (or “slide
main” as they call it) isn’t difficult but,
when installing the new one, one has
to align eight points simultaneously
to ensure the correct timing. The two
difficult ones are underneath near
where it engages the gear cam drive
(not shown in the service manual).
The next point to watch for is the
“Gear E/J Eject” drive which is loose
and must be aligned so that slot #1
matches tooth #1 on the ejector rack
at the top of the deck and tooth #1 on
the gear master cam underneath. After
that, it’s plain sailing and you simply
reassemble the parts in the reverse
order that they were removed.
Anyway, it all worked perfectly
once it was all back together again.
Obviously, the original gears had been
damaged by someone forcing a tape
in or out, though Mrs Nasty subsequently denied this when she called
to pick up the unit. Instead, she was
more interested in finding out what
sort of guarantee I gave. I told her that
I guaranteed the parts supplied and
the work done for 90 days but only for
the same fault and provided that the
equipment was not abused, as was so
obvious in this case.
I don’t think that this advice sank
in (or, more likely, she chose to ignore
it) because she brought the machine
back two weeks later, complaining
that it didn’t work again. I stopped
work, connected it up in front of her,
inserted a tape and pressed play.
There was sound but no picture; just
snow. I ran a tape cleaner but it made
no difference, so I removed the covers
and gently wiped the heads using a
lint-free cloth dipped in oil-free acetone. It left black marks on the cloth!
I replayed the tape and the picture
was now perfect. I reassembled it in
front of Mrs Nasty and explained what
had happened but, like her husband
before her, she declined to pay for the
work done to rectify their abuse of
the machine and disappeared with it
without so much as a word of thanks.
I don’t need customers like that and
I certainly won’t be doing any more
work for them.
A pig in a poke
Steve runs a secondhand furniture
shop not far away and one day he
brought in this large Samsung stereo
TV he’d bought at auction for $300.
Of course, it wasn’t working and he
wanted me to fix it for him. I know
him well enough to tell him that he
was mad to buy such a pig in a poke
and gave him a quick rundown on
some of the costs involved if certain
parts like the picture tube were cactus
(eg, anything up to $1000 for a large
screen – and this was large). He didn’t
turn a hair, being the eternal optimist
he is, and agreed to pay to have it
diagnosed and costed.
When I removed the back and found
a large plastic bag full of parts, I liked
his chances even less. As I quickly
discovered, this bag contained a
number of parts that were missing
from the deflection board. The flyback
transformer had also been unsoldered
and was lying free inside the cabinet.
Where was I to start? I removed
the parts from the bag, sorted them
out and found out where they had
come from before performing some
basic resistive checks with a multimeter. Most of the parts were either
completely short or open circuit and
had either come from the line output
stage or the power supply. The line
output transistor (2SC1880) was short
circuit, as was chop
per transistor
Q801 (BUV48).
Because the flyback transformer
had been removed, I initially suspected that it was also faulty. However, a
few basic checks revealed that it was
probably OK, so I replaced it. I also replaced the chopper transistor (Q801),
the line output transistor (Q401) and
any other parts that were faulty.
Restoring the power supply was
obviously the next objective. A few
checks revealed that fuse F801 and
resistors R807 and R817 were all open
circuit. The latter are designated on
the circuit as 0.27W types but had
been replaced with 0.22W 5W wire
wound resistors. I quickly fixed that
by fitting the correct fusible types.
The power supply is a conventional
switchmode type and is somewhat
similar to the ones used in Akai and
Nokia TVs. The difference is that it
has two regulator circuits, with VR801
controlling the primary oscillator
for +130V in the standby condition
and VR802 controlling a secondary
oscillator and feedback circuit which
ensures that the 130V rail remains
constant in the power on condition.
Both ICs in the power supply (IC801
and IC802) had previously been replaced, along with a few other components. To be on the safe side, I replaced all the small capacitors (C817,
C813, C814, C838 & C803) and then
checked all the remaining diodes and
transistors using a multimeter. I then
removed both the deflection board
and the small signal board from the
August 1997 63
set to protect them from any further
damage should something be badly
amiss with the power supply voltages.
Before applying power, I decided to
take a few more precautions. First, I
connected a dummy load consisting
of a 100W 240V globe and a parallel
voltmeter to the cathode of D814 (the
+130V rail). Second, I shorted the base
and emitter leads of the line output
transistor to prevent the line output
stage from firing up. And third, I
connected the AC input via a Variac,
with a 200W globe in series to limit
the maximum current to a safe value.
Now for the big test. I applied power
and slowly advanced the Variac, all
the time keeping my finger near the
on/off switch. And at 130V AC, the
oscillator fired up, the 100W dummy
load began to glow and the meter on
D814 read +130V.
Delighted at this progress, I slowly wound the Variac up to the full
240V. Everything remained intact
and so I switched off and removed
the base-emitter short from the line
output transistor after first confirming
that there was 130V on its collector.
Now for the acid test – would the
line output stage fire up properly?
I switched out the 200W globe and
64 Silicon Chip
wound the Variac up again but nothing happened. The CRO revealed
nothing on either the base or the collector of Q402 (the horizontal driver
transistor), so I traced the circuit back
from the collector through T401, R402
and D406, towards the +14V rail.
According to the circuit I had,
D406’s anode is connected directly
to the +14V rail at the anode of D816.
There is even a link position on the
board, marked J109, to allow for this
connection but there was no link in
position. And, what’s more, this link
had apparently never been fitted, so
what was going on?
In fact, it appears that the circuit
is in error. In practice, D406’s anode
is linked via J430 and J813 to the
emitter of Q805. And this transistor is
controlled by Q803 which, in turn, is
controlled by Q804. When Q804 turns
on, Q803 also turns on and this does
two things: first, it switches the +14V
rail through to IC802 and second, it
turns on Q805 which switches the
+14V rail through to D406 and subsequently to the collector of Q402.
So why wasn’t the +14V rail being
switched through by transistors Q803
and Q805. Answer – because unplugging the small-signal board had
removed the base drive to Q804. This
drive signal is normally supplied from
the main board via pin 2 of connector
CNP801.
As a quick test, I switched my
multimeter on the x1 ohms range and
connected the red lead to the chassis
and the black lead to the anode of
D835 (in series with Q804’s base).
The 2.4V across the test leads from
the multimeter’s internal battery was
more that enough to bias Q804 fully
on and the EHT section burst into life.
The meter on the 130V rail now
showed that it was too high but readjusting VR802 soon corrected this. By
now, I was optimistic that the set was
a “goer” so I removed all the safety
gear I had connected, reinstalled all
the boards and switched on. Eureka!
– up came the sound and we had a
perfect picture.
I rechecked and adjusted the two B+
pots before leaving the set to soak test.
It was still going strong after about a
week and I felt confident enough to
ring Steve and tell him to collect it.
So, in the end, Steve’s confidence
was well founded and he had certainly got himself a bargain – this time!
But I don’t generally advise people to
acquire TV sets in this manner. SC
|