This is only a preview of the October 1996 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 "Send Video Signals Over Twister Pair Cable":
Items relevant to "600W DC-DC Converter For Car Hifi Systems; Pt.1":
Items relevant to "Infrared Stereo Headphone Link; Pt.2":
Articles in this series:
Items relevant to "Build A Multimedia Sound System; Pt.1":
Articles in this series:
Articles in this series:
|
SERVICEMAN'S LOG
To tip or not to tip – a few tips
Ever wondered what happens to equipment
which is written off as not worth repairing?
Typically, it would be stripped for any
worthwhile spares and what was left would go
to the tip. But not always.
Yes, sometimes there is the temptation to try to salvage an item, even
if obviously uneconomical at a commercial level and there is a risk that
the attempt may not be successful.
One may spend many hours searching
for an elusive – possibly intermittent
– fault and not find it. Or, if a fault is
found, it may turn out to be a component which, for one reason or another,
cannot be replaced. But that’s the risk
one has to take.
I encountered two such exercises
recently which illustrate this situation very clearly. One is from my
own bench and one is a colleague’s
experience.
My own story involves what the
makers (Grundig) simply call a receiv-
er – Receiver 3000 (GB) – although it
would be better described as a tuner/
amplifier combination. It consists of an
elaborate stereo amplifier plus an AM/
FM stereo tuner, the latter featuring
press-button tuning, as well as continuous tuning and a “Digital Frequency
Indication Module”.
There are several input sockets to
take external signals of various kinds
and an appropriate switching system
to go with it. It also features both automatic and manual muting systems, and
these operate when switching between
stations or other signal sources.
All-in-all, it is a most attractive unit
and this example was in good physical
condition. So what was the story behind it? It belonged to a colleague and
apparently had had a rather chequered
service history, having previously
belonged to someone else.
But now, as my colleague summed it
up, it didn’t go and he had earmarked
it for the tip. When I expressed regret
that such a nice unit was to meet
such a fate, my colleague responded
instantly: “take it if you want it – you
can send it to the tip as easily as I can”
(he is not given to undue optimism).
And so I took it. I wasn’t sure what I
was going to do with it, assuming I
could fix it. For the moment, it was
mainly a challenge.
A nasty mess
At the first opportunity I put it up
on the bench but it was dead. I pulled
the covers off and this revealed a rather
nasty mess. There were three fuses,
one of which was obviously the mains
fuse while the other two were on the
secondary side of the power transformer. All three were blown. In addition,
there was a swag components which
had been unsoldered. Someone had
really gone to town on it.
Fig.1: the power supply circuitry in the Grundig 3000. Mains fuse Si.I is at extreme right, while fuses Si.1 and Si.2
are to the left of the transformer. T2 and IC1 are at the extreme left.
40 Silicon Chip
I decided the only logical approach
was to put everything back and start
from taws. I re-soldered all the components, then looked at the fuse situation.
The mains fuse, designated as Si.I was
marked 2A and the other two were
designated as Si.1 and Si.2. Si.1 was
marked as 250mA and Si.2 as 1.25A.
(No, there weren’t two Si.1s; the mains
fuse designation used a capital “I”.
Talk about planned confusion!)
I replaced the mains fuse and Si.2
and switched on. Splat! Si.2 blew immediately. I realised then that it would
be fruitless to go on without a service
manual or, at least, a circuit.
I rang Southern Cross Electronics,
the local agents for Grundig, and
asked about a manual. There was some
mucking about here. I had to fax a
request and they replied a week later
quoting $25 for a circuit photostat. I
placed an order and it arrived after
another week.
There were 10 A3 sheets in all.
Six were circuit diagrams taken from
what were originally two large foldout
sheets. The rest were parts lists, etc.
After dispensing lots of sticky tape and
patience, I eventually reconstructed
the foldout sheets, each of which
turned out to be 1.2 metres long!
After all that, I started over
again. I tackled the Si.2
fuse circuit first. Si.2 is
between a 12V secondary
winding and a full-wave
bridge rectifier (GL2)
which generates a 15V rail.
This is then applied to a
voltage regulator (IC1) to
provide a 5V rail.
I suspected IC1 and
I was right but in more
ways than I expected.
First, it was short circuit,
which didn’t surprise
me. What did surprise me
was that it turned out to be
a bodgie component. It was
not a 5V regulator at all but, in
fact, a 12V unit.
Just why this had been changed
and by whom remains a mystery.
It had probably failed because it
was the wrong type, particularly as
it was working directly into a 5V
zener diode (although the zener’s
role is something of a mystery in
itself).
Anyway, I replaced the regulator
with the correct type, fitted another
fuse and switched on. This time
the fuse held and we had a 15V
rail and a regulated 5V rail, with
no signs of distress.
So far, so good. Now to fuse
Si.1. This is part of another supply
rail which is derived from a 63V
transformer winding. This drives
bridge rectifier GL1 which in turn
drives another voltage regulator
based on T2. T2 is not a regulator
within itself, however. It is a Darlington pair, housed in a TO-220
flat pack encapsulation, and takes
its reference from external zener
diode D4. This arrangement provides a 55V rail.
When I switched on there was
an immediate response – R8, a 47Ω
1W resistor in the collector line to T2,
began to overheat. T2 was the obvious
suspect but an ohmmeter check failed
to reveal anything wrong.
Nevertheless, I unsoldered it and
pulled it out. And there was the fault
in full view – the insulating washer
between the heatsink (collector) and
chassis had punctured. And it was
obviously a voltage sensitive breakdown, immune to the low voltage of
the ohmmeter.
I fitted a new washer and tried
again. This time Si.1 held and I had
October 1996 41
Serviceman’s Log – continued
tran
sistors form part of the muting
circuit. When an appropriate voltage
is applied to their bases, they turn on
and mute the tuner signals into the
amplifier. I confirmed this operation
by the simple expedient of shorting
the base of each transistor to chassis in
turn, whereupon I had normal output
from the amplifiers.
So, the problem was really quite
simple – the “muting” voltage (or
possibly some other voltage) was
being applied to the bases of these
transistors and turning them on, even
though the muting switch was off.
All I had to do was find out what was
causing this.
Complicated circuits
a 55V regulated rail. I had rather
hoped that the thing would burst
into life now but it didn’t. Granted,
some of the LED displays and other
lights were now on but the frequency
display was dead.
A healthy buzz
I wasn’t sure of the significance of
this but decided to ignore it for the
moment and concentrate on getting the
sound path working. I pushed a scrap
of bare wire into the various amplifier
input DIN sockets and, eventually, was
rewarded with a healthy buzz from
each of the speakers.
So, the amplifiers were working –
we were making progress. But there
was no sign of life from the tuners.
Initially, I suspected that the frequency
display failure could be a symptom of
a major failure in the tuner section,
which was rather a nasty thought.
But then I noticed something else.
If I turned the volume control fully
up, I could detect faint sound when I
pressed some of the channel selector
buttons. So, was the tuner working
but unable to pass its signals to the
amplifier? After poring over the FM
tuner circuit on the other foldout
sheet, I pinpointed the stereo outputs
as being, initially, at transistors T5 and
T6. From there, the signals went to T8
42 Silicon Chip
and T9 and from there – on the other
sheet – to the switching circuits ahead
of the power amplifiers.
One of the most useful pieces of test
gear I have is a small audio amplifier
which is equipped with a probe. I use
it to trace audio signals and track down
losses and distortion. And this quickly
confirmed my suspicions; there were
strong healthy signals at both T5 and
T6 and also at T8 and T9.
OK, over to the switching circuits.
The tuner stereo signals come into the
switch bank on terminals 12A1 and
12A3 and emerge on terminals 4A1
and 4A3. From there, they go to the
amplifiers via switch position A3/B3
and plug socket SA10.
Only they didn’t. The signals were
present at the outputs of transistors T8
and T9 but not at the 12A1 and 12A3
switch input terminals. This drew
my attention to another part of this
circuit. Although the tuner signals
are routed through the switches to
the amplifiers they also go directly
to another pair of transistors, also
designated as T8 and T9, just to make
it harder.
These two transistors are connected
between the audio lines and chassis in
such a way that, if they are conducting, they pull the audio lines down
to chassis. In greater detail, these
But what had gone before was
merely routine compared with what
lay ahead. It was a real round-theworld-for-sixpence job. These circuits
are drawn using what I call draughtsmen’s cables; long thick black lines
into which individual lines disappear,
identified only by a number. One has
to follow the line until the number is
found, usually on another sheet.
And as likely as not, after a small
digression into a piece of circuitry,
it will go back into the cable on its
way back to the first sheet. Believe
me, it’s easy to go bonkers trying to
trace a circuit like this. Thankfully, I
didn’t go bonkers or at least I don’t
think I did.
I won’t bore readers with all the details of my circuit tracing. In any case,
without the circuit, which is much
too large to reproduce here, any such
description would be meaningless. I
actually lost count of the time I spent
on it and as readers will appreciate,
there is no way one could ever charge
a customer for this work.
In summary, I first tracked down
the manual mute switch (i1/i2) and
backtracked from there to an 8-pole
switch which is used to select the
FM preset channels. Only seven poles
of this switch are used for the actual
channel selection – the eighth pole is
in the muting circuit I had been tracing. And its function is to momentarily
activate the muting function whenever
any of the channel selection buttons
is pressed, thus masking any clicks,
bangs, or crackles, generated in the
process.
It was here that I struck oil – the
switch was faulty. Not only were the
muting contacts jammed closed but
the whole mechan
ism was giving
trouble. In a sense, I had already been
made aware of this. I had noticed
that, when a button was pressed, it
often took several attempts to get it
to lock into position. However, I had
previously put this fault to one side,
as something to be attended to when
the electrical problems were solved.
In fact, it was causing one of those
problems.
As a practical short term solution I
removed pin 1 from the plug assembly connecting to this switch, which
permanently disa
b led the muting
contacts. I could still mute the system
via the aforementioned manual mute
switch and the auto-muting, on weak
stations and between stations, still
functioned correctly.
The digital readout
Putting the switch problem on hold
for the moment, I turned my attention
to the only other remaining problem:
the Digital Frequency Indication
Module.
This is in a small metal box and, on
removing the covers, I was rewarded
with the sight of numerous dry joints
– more than I could be bothered to
count, in fact. How many more less
obvious ones there were I had no
idea. To solve the problem, I finished
up resoldering every joint but it was
worth it. The thing came to life and
worked perfectly.
And that’s how things now stand. I
consider it a pretty good effort, especially as I had done what, apparently,
those before me could not.
So, what about the switch? Should
I repair it? No way; it is made up of
numerous tiny pieces, many of them
under spring tension. Tackle that lot
and there would be bits flying every
where.
What about fitting a new switch?
That’s the logical answer but it’s no
longer available off the shelf and finding one may be difficult. It was most
likely specially designed for this set,
which is probably now about 10 years
old. The agents are currently checking
to see if one can be obtained from the
manufacturer.
And that’s about the best I can hope
for.
Scrounged video recorder
My second story, from a colleague, is
about a device he scored from another
colleague – a National NV-180 portable
video recorder. Because its fault had
proved elusive and so was potentially
expensive, the customer had written it
off and so it had been sitting in a corner
of colleague No.2’s shop for about a
year. But it left him in a quandary. He
wasn’t keen to spend more time on it,
yet felt guilty about sending it to the
tip. So, when my colleague showed an
interest, a deal was struck.
My colleague’s interest was understandable. He has a personal interest in
video cameras and associated portable
recorders. The NV-180 was originally
supplied with the models A1, A2
and similar video cameras. Although
bulky by modern standards, it was
regarded as a major breakthrough in
its day, weighing only 2.3kg without
the battery.
Apart from its portable role, it is an
attractive unit in its own right, featuring a large multi-function digital display, slow motion and variable speed
stop motion. Its accessories include
an AC adaptor, a tuner and a remote
control unit.
Unfortunately, after a year, the original fault details were rather vague. All
that my colleague could find out was
that it was something to do with tape
speed and a possible faulty capstan
motor. As a result, he had to start from
scratch. However, before presenting
his story, a brief review of the transport
control system may help the reader to
follow it more readily.
In considering the playback mode,
it is obvious that the speed of the
drum and the capstan – and therefore
the tape – must be held constant, at
a speed very close to the recording
speed. During recording, the speed
is controlled by the incoming signal
but there is no such reference during
replay; the system is on its own.
In this mode, it is controlled by an
internal reference; eg, a crystal. The
capstan motor itself is equipped with
a pulse generating device, typically a
sensing head (inductive or capacitive)
mounted close to a rotating wheel or
magnet.
The resulting pulses are fed to a
servo system which compares them
with the reference (crystal) frequency. This system then generates error
correction voltages which hold the
speed of the motor constant. A similar
system is used to control the drum
motor speed.
But that is only part of the story. As
well as running at the correct speed,
the system must also be in correct
phase. The drum must be positioned
so that a head, when it meets the tape,
exactly engages the beginning of a
track. And not just any track. If we are
talking about head No.1, then it must
engage a track recorded by head No.1.
It’s a similar story for head No.2.
The way in which this is done is
quite straightforward. When a tape
is being recorded, square-wave reference pulses, derived from the vertical
sync pulses, are recorded every 40ms
(alternate field) on a control track on
the lower edge of the tape. These are
used to provide the aforementioned
phase control and also the switching
between heads.
OK, here’s my colleagues story, as
he tells it.
Donald Duck sound
My mate had been right about there
being something wrong with the
capstan speed; it was fast, much too
fast. As a result, the sound had gone
“Don
ald Duckish” and there were
noise bars running up the screen. But I
didn’t buy the idea of a capstan motor
fault; capstan motors normally either
work or they don’t. Perhaps they might
run slow but I’ve never ever seen one
run fast.
The first thing I did was to give the
machine a good once over mechanically. This involved a routine clean,
belt tension and pinch roller checks,
and a check of the pause and search
functions. I found nothing wrong. I
then checked the main supply rails.
There was 5V at pin 13 of IC2505 and
9V at pin 14 of plug FJ24 – exactly as
marked.
My next step was to check the electrolytic capacitors around the capstan
motor drive, mainly C2532, C2533,
C2534, C2535. These were checked
by simply bridging them with another
unit of the same value but this had
no effect.
It was time to put the CRO to work
and check pulses. Unfortunately, the
compact nature of the device means
that servicing it can be difficult. For
example, I needed to check the Servo/
Power PC board which mounts hard
behind the front panel.
In order to gain access to both sides,
it is necessary to remove the front
panel and then mount the board in
a special jig – Service Connector Jig
(VFK0275) – which sits it at an angle
of 45 degrees, while maintaining all
October 1996 43
Serviceman’s Log – continued
connections. Fortunately, I have such
a jig.
I started by checking for the reference (FG) pulses generated by the
capstan motor and the CRO confirmed
that these were correct. The FG pulse
(FG1) appears at terminal 4 of the
capstan motor block and, via an allover-the-place path, finishes up on
pin 25 of IC2001 (AN3615K), which
is also test point TP2015. I traced the
pulses right through to this test point.
Next, I checked the internal reference (clock) frequency to which the
drum and capstan are locked. This is
a 4.43MHz crystal oscillator which
applies a 1.2Vp-p signal to pin 26 of
IC2001. And as a matter of routine, I
also checked the control pulses from
the Audio Control Erase (ACE) head,
although these are basi
cally phase
rather than speed control pulses.
These were present and checked
through to pin 9 of IC2001.
So, we had FG1 pulses from the
capstan, clock frequency pulses from
the crystal and control pulses from the
control head, all being fed to IC2001.
But for some reason, the capstan motor
was out of control and running free.
What followed was a laborious
check of various voltages and waveforms on the Servo/Power PC board.
This was at times quite difficult but
it eventually lead to pin 4 of IC2001
(test point TP2004) where there should
have been a 4.43MHz 50mV p-p waveform. However, this waveform was
missing; nor was there any voltage on
this pin, shown on the circuit as 3.2V.
The upshot of all this was that I
concluded that the IC was faulty and
ordered a new one on spec. And that
was a big mistake. When it arrived I
found I’d been billed for $93 – yes $93,
for one IC. Move over Mr Kelly.
There was worse to come. It was a
small IC, with closely spaced pins,
and mounting it on the double-sided
PC board was not easy. The job took a
long time – and achieved absolutely
nothing. The fault was there exactly
as before. Words failed me – well, in
print anyway.
The real fault
I had to find the real fault now.
Taking a closer look at the circuit
around the IC, I noted that pin 4
44 Silicon Chip
was internally connected to
two functions: (1) a playback
control amplifier (P.B. CTL
AMP); and (2) a tracking
mono multivibrator (TRACKING MMV), the latter connecting to pin 13. Pin 13 then goes
to the tracking control. It was
supposed to be at 0.6V – or higher
– but was in fact at 0V.
I hadn’t checked this voltage before, due to the difficult
access. Nor had I previously
checked the tracking control.
I checked it now; it wasn’t
working.
I traced the circuit
through to the tracking control (R6562,
100kΩ). This pot is
panel-mount
ed and is
connected via a short
length of 3-conductor
ribbon to connector P205.
And the one which ultimately connects to pin 13
was broken where it joined
the connector. It wasn’t immediately obvious, however, as it
is normally obscured and the
other two conductors held
the ribbon in place.
Of course that was it,
although how it happened is a puz
z le. I
can’t imagine any kind
of user abuse which
would cause it. More
likely, I suspect, the unit
had originally suffered
from a quite different
fault. The serviceman
had fixed this but had
broken the lead in the process. And
the resulting symptoms had proved
too tricky and confusing for the fault
to be traced.
In fact, it is not immediately obvious
just how the tracking control circuit
upset the speed. But, as far as I can
see, the loss of a connection to pin
13 was sufficient to upset the whole
capstan servo function within the IC.
If only I had checked the tracking
control first off.
And that’s my colleague’s story. My
first reaction is to quote another of my
colleagues who, in such situations,
was wont to remark, “that’s a decent
sort of an oops”. Which it was and I’m
glad I didn’t make it. But that’s not to
say that I might not have in similar
circumstances.
The bright side
On the bright side, my colleague
scored a very nice machine for the
price of the IC, plus his labour. Not
bad, really and he does have a spare
IC in his drawer.
But I feel the moral of both stories
is obvious; think very carefully before
you tackle an undertaking like this.
And be prepared for a lot of work – and
SC
the risk of failure.
|