This is only a preview of the July 1988 issue of Silicon Chip. You can view 42 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:
Items relevant to "Studio 200 Stereo Control Unit":
Items relevant to "Build the Discolight":
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
|
Sounding out a video recorder
For the most part, TV equipment sound sections
tend to be taken for granted; they don't give a great
deal of trouble and we don't think much about
them. Which is a far cry from the pre-TV days
when restoring sound was the main exercise.
But modern sound circuits do fail and when they do,
they can be quite tricky.
machine on that occasion. I had it in
the workshop for several days and
put it through many recording and
replay cycles, without the slightest
sign of any problem. The only clue I
had was a faulty tape brought in with
the machine and this carried a
background of the previous sound
track at a reduced but still annoying
level.
Nothing I could do would produce
the fault. And since there seemed to
be an urgent need for the school to
have it back, I simply gave it a
routine service, explained the situation and advised them to give it
another try and bring it back if or
when the fault was more predictable.
And that was the last I heard of
the device until a couple of weeks
ago. I can only assume that it
This story concerns a National
NV-300 video tape recorder. It
belongs to a local private school and
had first come to me with the sound
fault some two years previously. The
complaint then was that it would not
always erase the previous sound
when a new recording was made,
although there seemed to be some
doubt in the mind of the staff member
as to how serious it was. Some users
complained that there was no erase
action at all, while others described
the effect as a faint background of
the previous sound.
The NV-300 is one of the older top
loading machines but has been a
very reliable model. It is also easier
to work on than some more modern
designs.
I didn't have much luck with the
•
R4038
5~0
TO VIDEO
AUDIO OJT
l4001
GNO
. 470uH
BPGOO;
®c,026
Q033
-
____
i
I
-
TP-i001PLAY .
0.5 VI0.5msec . div .
:
'/
FE H(AO
P6005- l
GNO
P6001-1
__-_
, /,/
.
r
'
TP<:001 AEC.
0.5VI0.5 msec. d1v.
Fig.1: bias/erase oscillator circuit for the National NV300 video recorder.
Although a simple arrangement it developed a very tricky fault. Note the
missing chassis symbol on pin 8.
34
SILICON CHIP
operated normally or, at least, acceptably during that time. Then it was
back, brought in this time by another
staff member who wasn't even
aware that there had been a
previous problem. This time the complaint was rather more specific.
Granted the failure was still unpredictable but at least when it occurred now, it was complete; there
was no attempt to erase the previous
signals.
I set the machine up at one end of
the bench, connected to a monitor
and made a half-hour recording over
a previous recording. The machine
behaved perfectly. So it was going to
be one of those faults.
I went through this exercise, at approximately half-hour intervals, over
the next couple of days, without even
a hint of trouble. Then just as I was
beginning to wonder whether there
was some environmental factor involved, the fault appeared. The
previous sound was there at full
blast and, just as important, there
was only a shadow of the new sound
which should have been recorded;
just some brief distorted bursts of the
louder passages.
Well, at least I now knew that the
fault was real enough.
I had not delved very deeply into
the machine at this stage, simply,
removing the top cover but avoiding
moving anything else for fear of
disturbing the fault and causing it to
lie doggo. But now I had to go on the
attack. I began by removing the bottom cover which provides access to
the copper side of the main board.
Access to the component side of
the board involves removing the front
panel and unclipping the sub-boards
carrying the clock display and
various operating buttons which are
on flexible leads. About six screws
secure the main board and removing
these allows the board to be swung
up and propped open.
~f\~'t \
, •• 1-r
SU..Or-.l6S TO A 1-OCAl- PR\VAT& SC~OOl-- • .,.
The section I was seeking was the
audio circuitry and in particular, the
bias and erase oscillator, the circuit
of which is reproduced here (Fig .1 ). It
was fairly obvious that the system
was suffering from loss of erase
signal but remembering that there
had been little of the new signal
recorded, it appeared that it was suffering loss of bias signal as well.
And since both signals come from
the same oscillator, it seemed
reasonable to suppose that this was
where the trouble lay, rather than, in
a lead or plug feeding the erase head.
Circuit details
As can be seen from the circuit,
the oscillator is a simple arrangement; one transistor, two resistors,
three capacitors, a choke and a tapped winding forming one side of a
transformer, T4001. I decided to put
this section under observation and
the easiest place to connect the CRO
was at pin 1 of plug BP6005, which
connects to pin 7 of the secondary of
T4001.
This tapping supplies the "FE
HEAD", or full erase head. This was
not the section of interest but it was a
convenient monitoring point. The
audio erase head is supplied from pin
9, while pin 5 supplies the bias for
the audio recording head, this being
adjustable via variable resistor
R4048. One other point to note is that
there appears to be an omission from
the circuit, in that pin 8 is shown
floating, whereas it is a chassis connection and a vital one.
With the CRO hooked up, I set the
machine going again in the record
mode and observed that there was
approximately 45V p-p at the
monitoring point, at about 63kHz.
And as before, the machine performed perfectly for the next few days.
· However, I adopted a routine of
recording for periods of about 30
minutes, then shutting the machine
down long enough for it to cool,
before making another test. My
hunch was that the fault was most
likely to occur at start-up from cold,
rather than while running.
And in fact, it proved to be. Suddenly, at one start-up, I had no
oscillator signal at pin 7 and a quick
check with the CRO at other points
gave the same result. But the mere
act of making those measurements
triggered the circuit back into oscilla-
tion. It came up part of the way first,
dithered for a second or so and then
rose to normal level.
From then on it behaved more or
less normally but having been
alerted, I observed the pattern more
closely. Sure enough, it was exhibiting a dithering characteristic
whereby the amplitude would vary
over a range of about five volts. Then
suddenly, it began dropping lower
and lower, down to about 10 volts,
then failed completely.
·
Well, that seemed straightforward
enough. All I had to do was find out
why the oscillator was intermittent
and with so few components involved, I didn't imagine it was going to be
particularly difficult.
My first suspect was the transistor, Q4014. I pulled it out and
replaced it but it wasn't long before
the machine began dithering again
and I knew that I had drawn a blank.
That seemed to leave only the two
resistors, R4049 and R4050, and the
three capacitors, C4010, C4025 and
C4024. The resistors were fairly
easily checked, in situ and came up
spot on. What was more, they showed no indication of any intermittent
characteristics in spite of some
vigorous pushing and prodding.
The capacitors were not so easily
checked and I was on the point of
pulling them out and replacing them
when it occurred to me to try some
freezer spray on them. This had no
effect on C4010 or C4025 but brought
a swift reaction from C4024, the
10,-iF electrolytic; oscillation stopped
immediately.
So, I was getting somewhere at
last. I pulled the capacitor out and
replaced it but this produced a
strange result. This capacitor no
longer responded to the freezer but
the original fault was still there!
That put me almost back to square
one. I pulled out the other two
capacitors, tested them, found them
to be well within tolerance but
replaced them anyway. This achieved exactly nothing; the fault still
persisted.
Rescued by a drunk
By now, there wasn't much else
left to suspect; just the transformer
in fact and then only by default. The
snag was that I had no such device in
stock, this being the first time one
JULY 1988
.35
issued some two or three years
previously and how I missed out on it
is rather a mystery. The gist of it was
that, in the event of unreliable operation of this oscillator, the 4.70
resistor in the emitter circuit of
Q4014 (R4049) should be removed.
In fact, the suggestion was that this
resistor should be removed anyway,
as a routine mod whenever this
model was encountered.
SERVICEMAN'S LOG
A new theory
had ever been suspect. But I did have
another NV-300 machine on hand an unfortunate victim of a wild
Saturday night party during which
someone had upset a tankard of beer
over it. It's been in the workshop for
months and I've been working on it
during odd slack periods; the only
basis on which any kind of
economical repair is possible.
In fact, I am close to saving its life
and that should make another story.
In the meantime I was happy enough
to rob the transformer from the
drunken machine and try it in the
sober one. And that was it. The problem vanished and has not been seen
since, in spite of prolonged bench
testing and several follow up calls
after it went back into service.
But that's not the end of the story.
For one thing it left several questions
unanswered. What was wrong with
the coil anyway? Careful testing failed to reveal any obvious faults; certainly nothing in the way of abnormal
or varying winding resistance, such
as one might expect.
It is a small device, apparently of
pot core construction, and one
thought was that it might have suffered from cracked or chipped ferrite material, a condition which can
seriously upset the behaviour of the
associated windings. Well, that was
one to go on with.
36
SILICON CHIP
And what about the freezer upsetting one capacitor but not the one
that replaced it? I was still turning
these questions over in my mind
when I encountered one of my colleagues at a social gathering and
during the inevitable shop talk, I
related this story.
He was suitably impressed but
commented that he seemed to recall
some kind of modification sheet dealing with cranky bias oscillators in
that model. He promised to try to find
it and let me know.
He rang me the next day to say
that he had found it. It had been
TETIA TV TIPS
HMV 12613 (Braddon)
Symptom: No distinct red and
somewhat dirty greens. Colour
bars look vaguely normal but
careful inspection shows red is
really a muddy brown . The
monochrome picture is perfect.
Cure: R541 (4700 0.5W) open
circuit. This resistor supplies Vee
to IC503, one of two reference
oscillator chips in this unusual circuit. In this fault, IC502 (the 8-Y
oscillator) probably supplies some
drive to the red and green circuits
but it's phase is wrong, hence the
weak and bad colour.
Well, that threw a whole new light
on the situation. After mulling it all
over for a few hours, I finally came
up with what I feel is the most likely
explanation.
My theory is that the design of the
original oscillator circuit was a bit
dicey; that is, while it would maintain
oscillation under ideal conditions it
was only just making it, and was on
the verge of dropping out at any time.
It was the reference to the emitter
resistor which gave me the clue. The
use of an unbypassed resistor in the
emitter circuit of a transistor
oscillator - or the cathode circuit of
a valve oscillator in olden days was a favourite trick to improve the
oscillator waveform.
Because it is unbypassed, it is a
simple way of providing a degree of
negative feedback, thus holding
down the strength of oscillation and
reducing the tendency for the transistor (or valve) to be driven to cut-off
and saturation, two conditions which
seriously distort the waveform.
And it can be very effective, the
only snag being that it is a compromise arrangement; the negative
feedback provided by the emitter
resistor is fighting the positive feedback which is fundamental to the
oscillator circuit. So, if the idea is
carried too far, the whole arrangement can turn cranky.
So, was the original oscillator coil
faulty or not? It's hard to say. All I
know is that it would not work in a
circuit which was otherwise fault
free. It's possible, of course, that the
"fault" in the suspect coil may have
been nothing more than a normal
spread within otherwise acceptable
tolerances. Along with other component spreads it could easily have
been the last straw which sank the
camel's hump (to mangle a wellknown phrase).
UHF REMOTE
KEY SWITCH
(EA, JAN.87)
This proven and reliable unit is
*guaranteed* to work. It has
countless applications, 40-50 metre
range, multiple outputs, and is available
at a fraction of the cost of the other
equivalent unit. Beware of "short form"
kits that include little more than the
PCB!
<at>;;;~1~ ...
••• IN Tlt£ tl\E.AN"r\W\E., I WAS HAPP'{
"'rO "ROB "n\e:. Ttlt.1-\1\lSFORIV\e:R
FROM 11\£ ~UNKeN MACHl>J£ •••
As for the original machine, I'm
keeping my fingers crossed. But if I
can contrive to get my hands on it
again, for any reason, I'll whip that
resistor out, just to make sure. In the
meantime, I suggest you make a note
of that mod which, I understand, also
applies to the model NV370.
Let's hear it from J.L.
To change the scene, here is a
story from my regular contributor,
J.L. of Tasmania. It is one which is
not only intriguing technically but
also emphasises that when all else
fails, it is important to be able to
analyse how a circuit is supposed to
work. This is how J.L. tells it.
The set was a 63cm National,
model TC2652, which was completely dead. However, the owner had
observed that it made a brief noise
when switched on. When I tried it I
found that there was a five second
burst of normal sound before the set
shut down.
This kind of overload is often caused by a faulty tripler and disconnecting it will let the set run normally.
Unfortunately, this chassis uses a
diode split output transformer rather
than a tripler, which rules out this
trick.
A CRO check of the horizontal output transistor, Q502, showed that it
started up OK but died as the
waveforms approached full value.
And there was not even a flicker of
EHT, even while the transistor was
running, briefly, at about two thirds
its normal voltages.
I wired a 60W lamp in series with
CHOPPER
T801
0801
+111V
Our kit includes all the parts needed to
make a working link. Even the transmitter battery is included!
ON SPECIAL UNTIL END OF
THIS YEAR AT ONLY
$41.95
Plus $2.50 pack & post
For a complete Remote Controlled Car
Alarm, combine the UHF Remote Key
with our Ultrasonic Car Burglar Alarm
(SILICON CHIP, May 1988) or our
Shock Detecting Car Alarm Module
(ETI, March 1988); or perhaps with
both of them. Interconnecting diagrams
available from us. For more information,
see our advert on page 96.
* If upon completing any one of our kits
you cannot get it to operate correctly,
we will repair it for a maximum fee of
$1 0. 00. This offer applies only to properly constructed individual kits which
haven't been overloaded in incorrect
supply voltages.
OATLEY ELECTRONICS
ERROR
AMPLIAER
0803
8809
Fig.2: simplified circuit of the National TC2652 power supply, as drawn
by J.L. T801 and Q801 form a ringing choke oscillator, with regulation
provided via Q803 and Q802.
5 Lansdowne Pde, Oatley West,
NSW 2223.
Phone: (02) 579 4985.
Bankcard, Mastercard and Visacard
accepted with phone orders.
Send mail orders to PO Box 89,
Oatley, NSW 2223.
JULY
1988
37
e
1'HE..
(\\\f\\111
se., w~s
C.OM?L.E."i"EL"'( t>EA'O•••.
the collector of Q502 and ran the set
for about five minutes. By that time
the output transformer was quite
warm and giving off a typical sour
"brown" smell. I ordered a new
transformer.
This duly arrived and was fitted
but resulted in a set which was even
deader than before; there was not
even the five second burst of sound.
As before, the set would run with the
lamp in series with the output transistor but the 111 V HT rail was up
to 120V. Without the lamp the
voltage shot up to 120V before the
power supply shut down. By feeding
the set from a Variac I found . that
120V HT resulted from only 100V
Philips Tuners
ECL 2060 $17.00
Exchange or repair.
Three months warranty .
Send cheque or money order plus
$2.00 pack and post anywhere in
Australia.
J.V. Tuners,
216 Canterbury Road,
Revesby, Sydney, NSW 2212.
Telephone (02) 774 1154.
38
SILICON CHIP
AC input. Clearly the power supply
was not regulating.
This power supply is a self
oscillating chopper circuit, which
controls its ouput voltage by varying
the chopper "on" time via an error
amplifier. The trouble was that the
three transistors and the reference
zener diode involved all checked OK.
Also, the HT ( + B Adjust) control
worked in a seemingly normal way,
varying the voltage by about 20V a not unreasonable range.
At this point I put on my thinking
cap and reviewed what I knew about
the problem.
(1). The set worked OK when fed with
low AC input. This restricted the
fault to the power supply circuits.
(2). The chopper transistor Q801,
regulator transistor Q802, error
amplifier Q803 and reference diode
D811 were all OK on static test and
gave the proper responses under
load.
(3). The HT ( + B) adjusting trimpot
worked over a logical range but at a
higher than normal level.
Correct HT output should occur
with the chopper operating at about
50% duty cycle but in this case it
was on for close to 90% of the time.
This suggested that the error
amplifier was delivering the wrong
information.
There are not a lot of parts in this
circuit but I found its operation difficult to follow. In fact, I could make
no sense of it at all until I redrew
the circuit in simplified form, omitting every component that was not
vital to its operation. Fig.2 shows my
simplified diagram. As far as I can
see, it works as follows:
Q801, T801, D806 and their
associated components form a ringing choke oscillator, with voltage
regulation via Q802 and Q803. Imagine that Q802 is turned on. This
will tie the base of Q801 to its own
emitter and thus turn it off.
In normal circumstances, Q802 is
held in dynamic conduction by error
amplifier Q803 using zener D811 as
a reference. When the supply rail
rises above a preset level, Q803 and
Q802 turn on and Q801 turns off.
So much for my theory. I already
knew that all the transistors and
diodes were OK. I also knew that
C809 and a number of other
capacitors were, at least, not
shorted. They might have been open
but that would not produce these
symptoms. At this point I should have
been able to find the fault with just a
little logic. With hindsight, I can see
it sticking out a mile.
In fact, I decided to try an experiment. I removed Q802 and switched
on. There was absolutely no difference in behaviour whether the
transistor was in or out of circuit.
Since it wasn't Q802 it had to be the
only other component in this part of
the circuit, R808. Sure enough it was
open circuit and a new resistor
restored normal operation.
One final point. Why did I get some
output adjustment when the
regulator was patently not working?
This was the result of Q803 varying
its resistance, as part of a divider
chain also involving R803/820, R809
and R810. It was only a superficial
action but was enough to mislead me
in the early stages.
Well, that's J.L.'s story and I think
we should thank him for the trouble
he has taken to set down his analysis
of the system and provide a
simplified circuit. While this is not
the kind of approach we can afford
to adopt with every tricky set that
lands on the bench, it is one we
should be prepared to adopt whenall
else fails.
~
|