This is only a preview of the February 2000 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. Items relevant to "A Digital Voltmeter For Your Car":
Items relevant to "Build A Safety Switch Checker":
Items relevant to "A Sine/Square Wave Oscillator For Your Workbench":
Articles in this series:
Purchase a printed copy of this issue for $10.00. |
SERVICEMAN'S LOG
Projection TV – from many angles
In the course of my service career, I have
come across quite a few projection TV sets.
Normally I shun these because of the logistics
involved in servicing them. In-situ servicing
can be difficult in many cases while the sets
are just too big to easily transport back to the
workshop.
I was once asked to repair a projection TV set in a hotel and being
more naive (and hungrier) than I am
today, I attended the set to find that
although it worked, it had no green.
But the real problem was that it was
switched on and the drinking clients
were waiting to watch a world heavyweight title fight. The hotel manager
had not mentioned this; he had simply
told me that the set was in the lounge
and left me to fix it.
However, as I started to work on it,
one belligerent and somewhat intoxicated customer decided that I was
about to damage the set and felt that
it was incumbent upon him to protect
it. I count myself lucky that I was able
to get out of there without personally
being readjusted – and projected!
Thereafter, I made it a strict policy:
no house calls to pubs – ever. If they
want their sets fixed, they can deliver
them to me at the workshop and handle the delivery costs.
I have also been forced to apply the
same policy to property managers and
other time wasters. The scenario normally goes along these lines. It starts
with a request to pick up keys from
a real estate office (usually in a busy
street in a long No Stopping zone)
and go immediately to flat 27 on the
third floor in an old building (with no
lift) and fix an ancient unnamed TV
set with an unspecified intermittent
fault – straight away. They then want
you to return the keys and submit your
account for payment within 90 days.
Oh yeah! – if I’m lucky. And then
only after them first questioning and
whingeing about the cost.
Many of my colleagues know how
to deal with this – they charge like
wounded bulls. Personally, I prefer
to just politely refuse – it’s not worth
the hassle or my time. If they want
the set fixed, they can bring it to me
and pay when the job is done, just like
everyone else.
However, I did make an exception
recently when Mr Schultz, a well-spoken businessman, asked me to attend
to a rear projec
tion TV set he had
just imported from Germany. It was
a 117cm (46-inch) RP46 Thomson
employing an ICC9 chassis, about
four years old.
Apparently it had been working perfectly in Germany but the picture was
distorted and blurred when workmen
had unpacked it and installed it in
its new location in Australia. A local
company had sent a technician along
and he reported that one of the boards
had been cracked and that the set was
probably a write-off. I was asked to
check it out and give a second opinion
Sets Covered This Month
•
•
•
•
Thomson RP46 projection TV set
Seleco SVT 150 projection TV set
Dual Digital Concept TV4170 TV set
Sony KV-X2931S TV set
on behalf of the insurance company.
Fortunately, I already had experience with this series of sets and have
had similar problems with cracked
convergence boards which are held
along one edge, needing only one
quick jar to put an unacceptable strain
on the mounting. And so I agreed to
call and make a brief examination, to
confirm or refute this diagnosis.
Sure enough, the picture on the
screen had colour but was distorted
in all geometric settings. In addition,
both the static and dynamic convergence settings were way off.
To get to the convergence module,
I had to remove the loudspeaker baffles and find the concealed screws
that allow the front panel to come
off. This done, access to this board is
easy. However, although none of the
controls worked, the board looked
pristine with no sign of a crack.
I replaced the front panel and removed the back. The irritating part
of this is the need to use a Torx anti-tamper screwdriver for some screws
and a 4BA spin-tight for others. That
done, I was able to locate the convergence power supply on the righthand
side. I measured voltage going in but
none coming out and decided that
this was where the problem might
be. I removed this unit and told the
client that I would take it back to the
workshop.
The power supply was a conventional switchmode FET type but the
major problem was trying to match
the component numbers from the
circuits with those on the boards. For
example, the chopper transformer is
shown on the circuit as TR01 but on
the board layout it is marked T7100.
There is a note on the circuit and a
small chart under the heading – and I
quote – “Conv. of europ. Names to US
names.”, where these items are shown.
But it gets worse when identifying
the connecting plugs and sockets. For
example, BS02 becomes J7002 (mark
ed J2 on the board), which connects
FEBRUARY 2000 75
with J8204 on the convergence generator board, which then translates
to J204 and then BE04. It was all very
confusing.
Anyway, there was no sign of the
power supply even oscillating. I went
straight to the start-up resistors RP09
(R9U) 100kΩ and RP18 (R18) 220kΩ
and measured them. The former was
OK but the latter was nearly open
circuit. I replaced it and the whole
thing fired up properly. I returned
with it, refitted it to the set and after
some adjustment of the controls, the
set gave a very good picture.
Mr Schultz was happy on one hand
that his set was now OK – but unhappy on the other hand because he had
to pay for the repair rather than the
insurance company. However, it just
wasn’t possible for him to claim on
his insurance. After all, how could
one faulty resistor be put down to
accidental damage through shipping?
Another projection set
The next projection set I had to
repair was for a club that had bought
a secondhand Seleco, which wasn’t
76 Silicon Chip
working. Unlike the Thomson, this
was a front projection system, the
picture being projected onto a wall
screen.
The set itself was a 1987 model SVT
150 and was worth around $12,000
new. This one had to go directly back
to the workshop and I don’t know why
I took it on, as I didn’t even have a
circuit – I suppose I was just curious
to see what one gets for $12,000.
Removing the covers revealed immediately that it was very corroded.
The only positive aspect was that I
recognised the construction and layout of the boards as being similar to
the Fujitsu General series of colour TV
sets made in Italy at about the same
period. Indeed, the main deflection
board was marked BS950, the same
as for a Fujitsu General FGS281PTXT.
Even the remote control was the same
(30D3).
This was a stroke of luck because it
meant that I now had a circuit.
I started by switching it on and
there was a brief display, then an
“electronic click”, followed by silence
– something had just died. After a
more detailed investigation, I found
the set had been endowed with two
power supplies and the one marked
BS820 had failed.
Fortunately, it was a very conventional supply and I replaced the
chopper transistor (BU508A, T601),
four electrolytic capacitors and two
resistors and fuse F451. Coils TR454
and TR453 (20µH) had been slightly
melted but apart from that had held
up to the strain.
This time, when I switched on, it
tried to start but closed down after a
few seconds with just P30 displayed.
Because this set was very corroded,
I next looked at board BS776 which
I deduced generated the EHT for the
three tubes. After removing it, I went
over it very carefully, testing all likely components I thought might fail.
The only thing I could find here that
was that R615, a 100kΩ resistor, had
gone high.
My broad plan of attack was to
check each board, one at a time,
looking for obvious signs of corrosion
related failure or damage. However,
after spending a great deal of time, I
couldn’t find anything. The only thing
I could work out was that something
was killing the horizontal drive after
a few seconds.
I worked through all the boards until I reached the digital board (BS815).
This is very similar to the BS816 in the
Fujitsu General colour TV sets and indeed other digital TV receivers under
names such as Schneider, Dual, Teac,
Akai ITT and Nokia, to name a few.
These all use the same chip sets but
are not necessarily interchangeable as
they are often encoded for the instruc
tion set used within that model.
The horizontal drive was generated
from the DPU2553 deflec
tion processing unit IC (CI5) which is in turn
controlled by a central processing unit
and its EEPROMs. Though this set
was made in Italy, it was designed in
Germany and there was another clue.
The display constantly showed P31 or
program 31, no matter which button
of the remote control was pressed.
This suggested that it was highly
likely to be the CPU (or CCU as the
Germans call it), or even more likely,
the EEPROMs.
The Fujitsu General showed these
EEPROMs to be MDA2062s, whereas
this set used a sub-board (BS842) soldered into the main board in location
CI3. This sub-board carried a single
8-pin DIL EEPROM NVM3060 instead
of the original 14-pin DILs.
I thought I might be snookered here
in not being able to get the spare parts
but as luck would have it, the local
agents were able to sell me the IC.
When it arrived, the set powered up
and stayed on and there was sound
– but no picture – on all channels. It
took another saga to locate and replace
CI8 DTI2223, the “Digital Transient
Improvement” IC, to finally bring up
the picture.
But that wasn’t the end of the matter.
After it had been on for a while, the
picture broke up into black and white
horizontal bars and the Teletext no
longer operated. With the aid of some
freezer, this fault was traced to CI12,
the Teletext RAM chip (TMS
416415N). Fortunately, I managed to find
a replacement on an old computer
motherboard. After that, I was home
and hosed and the old projection set
performed quite spectacularly.
Dual TV set
Reverting now to conventional TV
sets, my next story concerns a 66cm
Dual Digital Concept TV4170. This
set uses a DTV2 chassis and is made
by Schneider in Germany. It is also
made under the brand names of Teac,
NAD, Nokia, Salora and ITT. The
entire chassis is no bigger than that
of a 34cm portable set and when one
looks inside, the first question one
asks is “where is the rest of the set?”
Apart from the control panel, which
is about 10 x 10cm square, the motherboard is divided into two sections: (1)
the power/deflection board and (2) the
digital/small signal/audio board, that
latter using the ITT digital IC chip set.
As with a few other recent jobs, this
set came to me via the tortured route
of failed repairs from several other
service centres who really didn’t want
to know. However, that’s not surprising really, considering that there are
not many of these sets in the country
and the agency has closed. We would
all prefer to work on easy faults that
earn money rather than complex ones
that don’t.
The set was reported as dead but
strictly speaking, it wasn’t. There was
no sound or picture but the switch-
mode power supply was working and
delivering all the secondary voltage
rails except for the 5V U3 rail which
was low at 3.3V and delivering 1.5A.
There was no horizontal or vertical
drive because this rail fed the “digi
board” where the oscillators are
located.
The problem was whether it was
a load fault or a supply fault. When
the interconnecting St.DT (play) plug
was removed, the 5V rail recovered
to its full value and this suggested
that it was a load problem. And it
seemed that the only way to locate
the fault was to desolder each device
in sequence until the short vanished.
This is rather difficult as the 16 odd
ICs are all hard wired/soldered close
together on the double-sided PC board
– not to mention the tuner, IF, audio,
AV and control panels. In addition,
many of the 40-pin ICs have multiple
connections to the U3 rail, many of
which are not marked. Nevertheless,
I could see no easy alternative and so
I persevered as best I could.
Unfortunately, after spending a lot
of time following this procedure, I had
FEBRUARY 2000 77
signal never arrived at pin
23 of IC701.
This turned out to be a
fault of my own making;
pin 23 had been poorly
soldered when I fitted the
IC socket. Because the IC
is on the component side,
it is very difficult to solder
the pin from that side. All
I could do was melt more
solder onto pin 23’s pad
from the other side until
it finally connected with
the pin.
When this was finally
achieved the whole set
was trans
formed into a
well-behaved, responsive
receiver with an excel
l
-ent picture. Everything
worked, including the
Teletext. The only thing
that didn’t was the picture-in-picture facility
and that was because the
optional module wasn’t
fitted.
A weird fault
made little progress. I could find no
direct shorts, due partly to the amount
of circuitry connected to this rail and
partly because it is almost impossible
to desolder awkwardly placed components on the double-sided boards.
But the work wasn’t entirely wasted. I was beginning to suspect the
Deflection Processing Unit, IC701
(IC1) DPU2553 and decided to remove
it completely and fit a socket for it to
the board. This was a drastic decision
because it isn’t easy to fit a 40-pin IC
socket to a PC board by soldering it on
the component side – especially with
virtually no room. However, I eventually managed to get the job done
and then ordered and fitted a new IC.
This turned out to be a good move,
with the 5V rail making a big recovery – although not quite enough. And
there were partial signs of life, with
EHT and a white line at the top of
the screen.
I next suspected the two EEPROMs
– IC1302 and IC1303 (MDA2062) – but
then decided that IC1301 (CCU7916)
was more likely to be the culprit, as
neither the remote control nor any of
the controls was having any effect.
And I have to confess that I could
only make intelligent guesses as to
which section or IC might be faulty.
78 Silicon Chip
Fortunately, my guess proved to be
correct. When IC1301 (CCU7916) was
replaced, the sound was restored and
all the controls were working.
However, there were still problems
with the picture. The screen was intermittently trying to produce a full
scan but there were huge quantities
of what looked like hum moving up
and down the screen. In addition, the
horizontal deflection was off speed,
as evidenced by picture tearing. The
set was also suffering a great deal of
stress, with a lot of heat being generated on the deflection module.
I switched it off and had a think
about the problem.
The most valuable symptom was
the horizontal deflection system being
off frequency. Why should this be so
on a digital set when the oscillator
is crystal controlled and has AFC
feedback?
Because the frequency was only
slightly off, a logical assumption
was that the problem had to be in the
AFC feedback loop. This is generated
from pin 4 of the horizontal output
transformer TR302 (TR2) and goes
through CR410 (CR10) to pin 6 of the
interconnecting harness plug St.DT.
I followed this with the CRO to the
anode of diode DT02 (D2) but the
My final story this month describes
one of the weirdest faults one could
imagine. It was found in a set by our
local Sony service agent.
The set, a 1990 Sony KV-X2931S
AEB1 chassis, belonged to an elderly
couple and it had given good service for about eight years. Recently,
however, it had started cutting out
intermit
t ently, giving no picture
(black screen) or sound but leaving
the channel number displayed on
the screen.
After a few abortive house calls to
fix it, it ended up on their workshop
bench. Subsequently, a number suspect solder joints were found which
were duly attended to and the set was
put aside for testing. Initially, they
were quite confident that the fault had
been scotched by this wide sweep but
it wasn’t to be.
This was the kind of fault that just
keeps coming back and after a few
days, they were right back to square
one. Over the next couple of weeks,
the set rocked back and forth from the
soak-testing bench to the workbench
and each time some suspect joints
were discovered and resoldered.
Finally, it was considered good
enough to go home where it worked
for about a month before the symp-
toms returned. It then came back to
the workshop where a couple of more
joints were resoldered it worked OK
for a week before going home again.
This see-saw between home and
workshop subsequently went on five
times, with the set always being OK
at the service centre and in trouble
at home. Of course, you can imagine
how the owners were beginning to get
a little tetchy about this.
In the meantime, all hell was breaking loose at the service centre as to
how to fix such an intermittent fault.
Finally, the boss, whose expertise is
normally confined to audio equipment decided to have a go.
To cut a long story short, the attention of all concerned had gradually
been homing in on a particular board
– the J1 board which carries the audio
control, AV input, Y/C input, SCART
video out and E/W correction circuits.
This board is mounted vertically at
the rear supported by a cream plastic rear support bracket (11) and the
whole combination is hinged at the
bottom so that it can fold down for
service. The copper pattern side of
the board faces inwards, against the
plastic support – see Fig.1.
It seemed that moving this assembly
could bring on the fault – sometimes!
It was eventually established that the
set would always work with the back
off, but intermittently fail with it on.
A lot of time was spent examining
the plugs and sockets and hinge connections J1-41 and J1-51, as pushing
the assembly backwards and forwards
could induce the fault.
Finally, the boss confirmed that
if the cream plastic bracket was removed from the J1 board, one could
do anything with the board and the set
still worked. But when everything was
reassembled into the normal positions
and the plastic back of the cabinet was
replaced, the fault could be induced
by pressing on this plastic back.
The point here is that there is a
plastic cover mounted inside the
cabinet back, apparently to protect
the component side of the board. So
pressing the cabinet back can move
the board/bracket combination.
Simple solution
Finally, he reached a dramatic conclusion – it wasn’t moving the board
that caused the fault – it was just
pushing the board against the cream
plastic bracket that could do it!
Fig.1: an exploded view of part of the Sony KV-X2931S AEB1 TV chassis.
The J1 board (12) is supported by the plastic bracket (11). Note the fitting
(13) and the cover inside the cabinet back.
Apparently, this bracket had become conductive and with the back
on, the cover pressed against the
component side of J1 board. This in
turn meant that the copper side of the
board was pressed against the plastic
bracket.
So far, no one has managed to measure the conductivity of the plastic or
determine precisely which circuit
was affected. But the solution was
quite simple – use insulating tape
around the edges of the plastic bracket
where it touched the J1 board. He
had previously cleaned the bracket
with acetone to remove any residual
conducting chemicals but without
success.
This finally achieved a lasting
repair and the set is still working
months down the track.
Of course, a new plastic bracket
may have fixed the problem but why
go to the expense when just a few
centimetres of tape was all that was
needed?
Who would have thought that this
board/bracket assembly could cause
such a fault? Apparently, not all plastics are good insulators as we have
been lead to believe!
Finally, note that this fault could
be very similar to one found in a run
of 1992 Mitsubishi TV sets, models
C3420/C3421. This produced a product recall when a white plastic PC
board frame had to be replaced with
SC
a black one (BFC 3420-01).
Silicon Chip Binders
Heavy board covers with 2-tone green vinyl
covering
Each binder holds up to 14 issues
SILICON CHIP logo printed in gold-coloured
lettering on spine & cover
REAL
VALUE
AT
$12.95
PLUS P
&
P
Price: $A12.95 plus $A5 p&p each (Australia only; not
available elsewhere). Buy five and get them postage free.
Just fill in & mail the handy order form in this issue; or fax
(02) 9979 6503; or ring (02) 9979 5644 & quote your credit
card number.
FEBRUARY 2000 79
|