This is only a preview of the September 1997 issue of Silicon Chip. You can view 29 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 "Multi-Spark Capacitor Discharge Ignition System":
Items relevant to "Building The 500W Audio Power Amplifier; Pt.2":
Articles in this series:
Items relevant to "PC Card For Controlling Two Stepper Motors":
|
SERVICEMAN'S LOG
The things I do for money
In my never ending quest to earn money, I
never cease to be amazed at what I am asked
to repair. Recently, I ended up repairing a
video games machine on the footpath but
first I faced up to a couple of TV sets.
I must admit that I feel more comfortable working on TV sets and VCRs
but you cannot be too fussy these days
if you expect to earn a living. Computers, computer monitors, microwave
ovens, video cameras and even the
odd video games machine are now all
part of the job as far as I am concerned,
although many organisations prefer to
specialise in just a few areas.
But I still like working on TV sets
(mostly) and so the first two jobs I
tackled when I arrived at the shop this
morning were a couple of Philips TV
sets that had come in late the previ
ous day. The first was a 34cm Philips
GR1-AX chassis (model number
14GR1224/75R) which apart from
making a buzzing noise was otherwise
quite dead. This popular model is now
well known with few surprises left.
But to those not familiar with it, it
can be a little daunting the first time
it is encountered.
If the main fuse is intact like this
one was, there are really four main
conditions possible for the main B+
rail which normally sits at 95V. First,
if it measures only 1.5V across C2660,
the crowbar circuit is probably faulty
and you have to check SCR 6641, zener diodes 6638-6640 and diode 6642.
Second, if it measures only 5V, the
procedure is to check the standby
circuitry from the remote control, the
microprocessor and from transistor
7631.
Third, if it measures 10V, FET 7610
is probably open circuit and D6610,
D6613, R3616 and R3680 also have
to be checked. Finally, if it correctly
measures 95V but the set is still dead,
inductor L5524 is a likely culprit (it
can go open circuit) but it could also
be the flyback transformer (5530)
that’s at fault.
In this case, it was the latter and a
new one quickly cured the problem.
Fig.1: the 11V rail in the Philips 21MK2460 is derived from pin 16 of the
transformer winding via a 1Ω safety resistor and diode 6180.
38 Silicon Chip
I also reworked a few suspect solder
joints on the deflection yoke plug and
socket connections and also on some
of the inductors, transformers, transistors and even IC7020-B. By the way,
the 9V and 12V rails are critical on this
model and so I always replace C2523
(6.8µF) with a 10µF 63VW 105°C type
(and also occasionally C2542).
Finally, a few words ought to be
mentioned about the “Hotel Mode”.
Once again, for those unfamiliar
with this model or who never read
the instruction book, it is fitted with
a childproof lock which prevents the
tuning from working and limits the
maximum volume level. To unlock
this, all you have to do is select program site No.38 and press STORE and
CONTROL+ simultaneously.
Another Philips
The second Philips set was a
late-model 21MK2460 using an Mk.2
chassis. The customer’s complaint
was that there was no picture or sound
and, as far as he was concerned, it was
dead. However, turning up the screen
control soon revealed a blank raster
but that’s about all there was.
This was my first time with this
model although I did have a circuit for
it on my files. My first thought on the
problem was that perhaps it was stuck
in the AV (audio-video) mode but con
necting an external video source made
no difference. The problem seemed to
be situated in the small signal circuits
and as it was affecting both sound and
picture, I checked the low tension 12V
rail supplying them. This measured
OK and so I next turned my attention
the logic circuits and the 5V supply
rail to them.
This time, there was a clue – the 5V
rail was down to just 3V which meant
that I now had something to go on.
By tracing this rail back towards its
source, I quickly discovered that it is
derived from an 11V rail via transistors 7182 and 7184. And this 11V rail
revealed as a very faint hairline fracture in the copper track immediately
adjacent to the pin. This was definitely
an 11 out of 10 on the “dj” (dry joint)
scale.
The rest was an anticlimax – soldering this microscopic crack fixed the
problem once and for all. But what
caused it in the first place? My guess
is that the transformer pin was slight
ly crooked and the fracture occurred
when it was forced into its mounting
hole during assembly.
A job for the Terminator
was also low, measuring just 7V from
diode D6180.
I then measured all the other rails
from transformer 5161 and found
them all to be correct bar this one. The
circuit here is quite simple (see Fig.1).
The top of the transformer winding
terminates on pin 16 and goes to diode 6180 via a 1Ω safety resistor. The
rectified output is then filtered using
a 680µF electrolytic capacitor (C2180)
to produce the +11V rail.
The 1Ω resistor checked out OK and
it made no difference when I substituted C2180 – the rail was still too
low. However, when I disconnected
the 5V rail, the 11V rail came good.
In fact, disconnecting anything from
this rail caused it to rise but I could
not determine whether it was an excessive load problem or an insufficient
supply problem.
I disconnected pin 42 of the microprocessor (IC7200) and all the other
components one by one but they all
restored the 11V rail. They couldn’t
all be faulty, could they? Surely not.
My next thought was that perhaps
diode D6180 was intermittent, so I
tacked another one in parallel with
it to the copper side of the PC board.
This made no difference, so I decided
to short out R3180 by connecting the
new diode’s anode directly to pin 16
of transformer T5161 via a short length
of hookup wire.
When I did this, the 11V rail immediately came good again and the set
burst into life. What’s more, the 5V
rail was also now correct so what was
happening? I removed and replaced
R3180 but it still wouldn’t work properly without the wire link between
R3180 and pin 16 of the transformer.
By now, it was obvious that the
connection between pin 16 and R3180
was not kosher, despite the fact that
it looked perfect. However, when I
checked this path with an ohmmeter
it measured no resistance. Of course,
that only meant that the fault was not
showing up under no-load conditions.
There was a lot of white paint
surrounding pin 16 and so I decided
to scrape this away with a utility
knife. And there at last was the fault,
I was about half-way through the
second Philips set when the local
video store rang and asked me to fix
their Mortal Combat. “Mortal who?”
I responded. “You know, the one-arm
bandit”. This sounded dangerous.
“You mean you want me to fix a one
arm mortal combat bandit? Are you
sure you got the right number?”
“No, No, I mean the video game
machine”.
Ah ha! A cyborg repair – a job for
the Terminator.
Eventually the story unravelled as
follows. The video store, which is a
pretty small shop, was supplementing
its meagre income with a video games
slot machine. This was located near
the entrance and the kids had been
fairly pounding it recently because
first the picture gave lines and then
it went dead.
Because it was a fine day, I decided
to abandon my stuffy workshop as
soon as I had finished the Philips set
and have a go at fixing this. Normally,
I would be quite reticent at taking
on such unfamiliar equipment but I
was feeling unusually optimistic. It
couldn’t be anything too complicated,
could it? I didn’t realise that I would
be fixing the thing on the pavement
but, as it turned out, that was the only
way I could find sufficient space to
gain access to the machine.
The game consisted of a sturdy
wooden case on wheels with a 20-inch
CRT on top and the controls on the
front. The case was reinforced with
steel and multiple locks and chained
to the front of the shop. Good neighbourhood, this, I thought.
Anyway, I eventually removed the
front and rear panels and found that
the parts were generally quite accessible. What’s more, the unit appeared
to be very well-made. On the base of
the cabinet was an expensive looking
September 1997 39
Serviceman’s Log – continued
of things that needed doing.
First, both the protective glass and
the tube surface were covered with
dust and dirt. However, the glass is
easily removed by unclipping two
suitcase-type latches inside the cabinet and hinging back the control
panel. Once this has been done, the
rest is easy and the picture looked a
lot better after I had finished cleaning
the glass surfaces with detergent and
a damp cloth.
On the downside, my cleaning
efforts revealed a further problem
in the form of visible retrace lines.
These were eliminated by reducing
the screen control setting and setting
up the sub-brightness.
All that remained to do was to relock it all up and extract my fee from
the proprietors. This had turned out to
be a relatively straightforward job and
for once my optimism had paid off.
The crook microwave
power supply and this was neatly terminated, with all the voltages clearly
marked. The actual games circuitry
consisted of a large PC board that
looked just like a computer mother
board and, finally, there was the TV
moni
tor board. These boards were
connected by a large wiring loom to
the various controls and to the money
collector.
Power supply checks
The whole machine looked quite
dead so I decided to start by checking
all the different rails in the power
supply. I was surprised to find them
all correct and so took the next step
of looking at the monitor board. The
CRT heaters were on but because of
the street noise, I couldn’t determine
whether the line output stage was
working. However, the B+ voltage on
the line output transistor was correct,
as were the screen voltage on the CRT
40 Silicon Chip
socket and the voltages on the RGB
output transistors.
It all looked pretty good so far, so I
turned my attention to the computer
board. This was clamped to the side
of the cabinet and was awkward to
get at. Anyway, while I was loosening
the clamps, the large edge connector
fell off with an ease that was all too
apparent. I reconnected it and pushed
it home as far as it would go. It could
travel quite some distance and I noticed that the wiring loom connected
to it was heavy and quite tight.
It didn’t take a genius to figure out
that the connector could easily come
loose if someone was giving the case
a good thumping. And that’s just what
young kids are inclined to do when
the game isn’t going too well.
I repowered the machine and it
immediately came to life with sound
and picture. But although the picture
was reasonable, there were a couple
My next job in this day’s potpourri
was a Sanyo Micro-Convection microwave oven, model EM-5710. This
particular unit is now approximately
20 years old and features a stainless
steel oven, a heating element in the
roof and a fan. The problem that it
had first became evident seven years
ago but, on that occasion, the owner
had decided not to opt for a full repair
and instead went for a temporary one.
I first saw the oven back in 1990,
when it was brought in dead. The
8-amp power fuse was open circuit
and the 25W SES globe had failed.
I also noticed at the time that the
convection fan in the roof was loose
due to worn bearings. However, the
customer was adamant that he didn’t
want to spend money on getting this
“minor problem” fixed up.
This time, seven years later, it was
sparking badly in the roof from two
different locations. First, there is a
ceiling partition between the element,
fan and the main oven and this material had, over the years, been caked
with variety of different foods that had
gradually become carbonised. Once
it reaches this state, the microwaves
regard it as unwanted metal and the
area absorbs a lot of energy. As a result,
it gets hot and generates sparks which
creates even more carbon.
The second source of the sparks
was the fan itself. This was lurching
around in its now totally worn out
bearing case and was striking the
stainless steel roof, all the time giving
a shower of sparks. It is quite possible
that minute molten metal particles
were spraying the partition below and
were responsible for starting the fire
there in the first place.
Fairly obviously, the first job was to
clean up the mess and I spent some
time removing the dirt, rust and carbon deposits around where the fan
was hitting the roof. The carbon on
the partition was removed by cutting
it out, just as a surgeon would cut
away a cancer.
Despite the oven’s age, the fan bearing was still available from Sanyo, as
indeed was the belt. After reinstalling
it, the acid test came when the power
was connected and all the sequences
tested. A mug of water was used to ensure that the magnetron had a load to
absorb its energy, while a small fluoro
tube (with its metal ends removed)
from a discarded portable torch was
placed in the water to check for the
presence of microwaves.
When the oven came on, the fluoro
tube glowed brighter than it ever did
in the torch. More importantly, the
sparks had stopped and the rest of
the oven was neutral, even when the
water boiled.
Purple patch
My next story concerns a Compaq
Presario 14SV monitor, vintage 1995,
with the complaint that “first it had
purple lines, now it has no picture”.
This super VGA monitor is made by
Samsung for Compaq and, as is my
policy, I first connected it to a com
puter to confirm the symptoms. This
was just as well as the symptoms I
observed were completely different
to those described by the customer.
I did get a picture of sorts but there
was no horizontal control and I also
noticed that the green LED power
indicator wasn’t working.
I prepared for surgery and dived
in. Removing the back is quite easy
provided that you know about the two
concealed catches at the top. Once
inside, you realise that the bottom
PC board is inaccessible unless you
remove all the metalwork and then
the motherboard.
My gut feeling was that there was
a dry joint somewhere and the clue
would have to be the LED. So once I
got the chassis out, I checked the LED
and the connections all the way back
to the power supply where, to my
relief, I found that there was indeed
a dry joint on 3-terminal regulator
IC202.
On the dry-joint scale of 1 to 10,
this was about a 6. I resoldered the
regulator and then carefully inspected
the rest of the chassis for similar faults.
This revealed no further problems, so
I reassembled the motherboard into
its metalwork and reconnected all the
plugs and sockets. When I switched
it on, the monitor’s picture locked
perfectly and the power indicator
LED shone green and so it was put
aside to soak test before being finally
pronounced cured.
The old Sanyo
My final story this month comes
from K. Sims of Black Forrest, SA). I’ll
let him tell the story in his own words.
I’ve been in the lighting and sound
business for around eight years now.
TV’s certainly aren’t my cup of tea but,
on this occasion, I decided to tackle
a project that would be a bit more of
a challenge.
The set in question was a Sanyo CTP
5604 that was found in the “equipment graveyard” room at a night
club that I have been working for.
Apparently, it had been there for about
two years but no-one knew where it
originally came from. As the manager
was junking all the rubbish from the
offices, I decided to take a look at the
set to see if it could be rescued.
Plugging it in for a quick inspection
revealed a squashed image no more
than about 8cm high across the centre
of the screen and there was no sound.
This seemed to be two rather unrelated faults but as the picture is the
most important of the two I decided
to start there.
Because the set had its back towards
me, I decided to remove the rear cover
and have a quick look inside for any
obvi
ous signs of trouble. As I was
about to switch the set off, I noticed
that the picture had “grown” to the
extent that it now almost filled the
screen. This indicated that the fault
was heat sensitive and that a search of
the deflection board for dry capacitors
would be a good place to start.
After shifting the set to my home
and making a list of the capacitors
that needed replacement I also noticed that 0.5W resistor R452 had
been overheating, to the extent that its
outer insulation (and colour coding)
had burnt off the resistor’s inner core.
PCB POWER
TRANSFORMERS
1VA to 25VA
Manufactured in Australia
Harbuch Electronics Pty Ltd
9/40 Leighton Pl. HORNSBY 2077
Ph (02) 9476-5854 Fx (02) 9476-3231
After a few calls, I was informed that
Sanyo had a techni
cal information
line that you can call for information
on assorted topics regarding repair.
They kindly faxed out a schematic for
the deflection board and informed me
that R452 was actually a 33Ω resistor.
It still tested OK but I replaced it for
good measure.
After replacing half the listed electrolytics, I realised that the worst one
(C444, 4.7µF) was actually a non-polarised type. Eventually, I ended up
replacing this capacitor and the remaining high-value electrolytics with
low-leakage types obtained from a TV
repair centre.
On powering the set up again, I
was rewarded with . . . absolutely
nothing! I could still hear the whistle
from the scanning coils and the tube
heater was a lovely shade of orange but
there was no picture at all. On closer
examination, I then noticed that a 3W
resistor adjacent to R452 had come
adrift, probably due to a dry joint. A
little resoldering fixed that problem
and that did it – the set came good.
It seems that replacing all those
suspect capacitors cured the fault and
the set now delivers a top class picture
and sound. I also spent around an hour
looking for the vertical centring control only to finally find that it wasn’t
a pot but a jumper lead that goes to
one of three connecting posts on the
deflection board. The UHF tuner was
also removed and the tuning mechanism relubricated, as it had frozen
SC
solid from lack of use.
September 1997 41
|