This is only a preview of the July 1989 issue of Silicon Chip. You can view 44 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 "Experimental Mains Hum Sniffers":
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
|
You
need a pal to deal with PAL
When the PAL colour system was first introduced
we all burnt a lot of midnight oil trying to bridge
the gap from NTSC theory to PAL theory. While
most of us felt we had succeeded, time and day-today practical work have taken their toll. Now,
when this theory might help, our memory fails us.
Nothing very exciting happened
on my own bench this month, so the
following story comes from my colleague J. L. in Tasmania. It tells how
he encountered the need for some
PAL theory. Here's J. L.'s story:
There are some television sets
that I don't like servicing. One such
set is the Sharp model C1831X (and
I can hear a lot of my colleagues
saying "Hear, Hear!"). It's true
that many of us have had trouble
with this chassis yet others have
had no trouble at all.
It's almost as if the set has a mind
of its own and doesn't like anyone
wearing glasses, or a bald head, or
false teeth. On the other hand,
chaps with beards seem to have no
trouble at all with this set, so maybe
we should all grow long hair!
There are a number of common
problems in the vertical section of
the C1831X which are fairly easy to
cure, so long as you are careful. Unfortunately, as you slide the chassis
out of the cabinet, it is all too easy
to allow the contacts on the underside of the vertical oscillator chip to
touch the metal rails on which the
chassis moves. It doesn't seem to
matter whether the power is on or
off; you don't get a second chance
and the chips are quite expensive.
Another problem is with the
wire-wrap interboard connections.
You can only flex a single core wire
so many times and if you are chas-
ing a difficult fault it is easy to bend
the wires once too often.
They tend to break inside the insulation and you don't know until
you switch on and some part of the
set no longer works. I heard of one
case where a vital earth link broke
and somehow allowed the 200V rail
to go to chassis through the 12V
rail. It did dreadful things to every
IC and transistor in the set!
So with this sort of background to
the C1831X, and the fact that my
glasses steam up every time I see
one, I am quite happy to concentrate on other models and leave this
particular Sharp model to those
who enjoy head-butting brick walls.
Oh boy, a C1831X
But sometimes you just can't
refuse to do a job and the story that
follows tells about one such exer-
,_.T\·HS 1\MG: SH~ C.A
MC:., NOi" -ro C.OME. & FI~ i\-\~E-r, SU\ i'O ,A~E n- AWA'-{ FOR '""C:R .....
52
SILICON CHIP
.d
~ ~ ~-
iHE. 'PlCtURe:.. WPS V ~
DARK & SH-€ C.OUL..'D ONL.'1
WlrrCH rr ITT" \\l\GH-r: ...
cise. When a customer gives you an
old set for junk there's a strong
temptation to get it going again, if
only to see if it can be done. And if
it's a C1831X, then you're stuck
with it!
This particular C1831X belonged
to a well-liked, elderly customer
and I had attended to it a couple of
times in the past. The faults were
never very serious but I was always
apprehensive whenever I heard her
on the phone.
This time she called me, not to
come and fix the set but to take it
away for her. It seems that a small
insurance policy had matured and
rather than bank the money and
complicate her pension status, she
bought a new set to replace the aging Sharp.
It had been playing up a bit
before it was retired. The picture
was very dark and the owner could
only watch it at night. She guessed
that the picture tube had gone but
that I might get some useful parts
out of the rest of it. Hence the call.
I wanted a Sharp C1831X like the
legendary "hole in the head" but
out of respect for a valued
customer, and at the first convenient opportunity, I called to collect
it.
The last time I had seen the
Sharp was some six months earlier
and it had then displayed quite a
good picture. I wondered if it really
had dropped its bundle so suddenly
or was there a more logical reason
for the dark picture.
Back in the workshop I fired it up
and saw what I had suspected all
along. The picture was truly low in
brightness but the chroma was
more or less normal. It had to be an
open circuit bypass capacitor or
feed resistor in the supply rail to
the video output transistors.
This kind of fault is common in
circuits where the video output supply rail is bypassed with an electrolytic capacitor, particularly low
value types of about 10µF or less.
Low value electros are notoriously
unreliable after a few years but
higher values can also fail.
It didn't take me long to confirm
that the video output transistor collector voltages were very low indeed. This seemed to discount the
capacitor theory because usually,
when the bypass capacitor is open,
a multimeter shows a rail voltage at
somewhere around 50 to 75% of
the normal voltage.
During the flyback blanking
period the rail rises to quite a high
level because there is minimum
load on the supply and no
capacitance to absorb the supply
pulse. Once blanking comes off, the
video output transistors draw on
every scrap of available energy and
the rail voltage plummets.
A multimeter averages these two
levels and shows a rail voltage considerably higher than the true
voltage during the line period. (Picture content also has a big effect on
the rail voltage in the fault condition. A white screen loads the rail
heavily, while a black screen draws
little current and the rail can rise to
almost normal value).
In this case the voltage was very
much lower than the condition
described above, and also more
consistent. This suggested that it
was more likely that the pulse
which generates the rail voltage
was not even reaching the capacitor.
And so it proved to be. R952, a
4 70 1W resistor feeding the
capacitor (C104, 47µF), was reading
several hundred kilohms and the
surprising thing is that there was
any voltage at all on the output collectors. And C104 was faultless .
Hanover bars
Replacing the resistor soon had
the brightness back to normal and
the set was displaying quite a
reasonable picture. Except for one
thing - severe Hanover bars.
These bars are an obscure fault
that occurs only in the PAL colour
system. In PAL the phase of the
transmitted chroma signal is alternated line by line and this is achieved by alternating the phase of the
R-Y signal. So, if the phase of the
red signal on line 1 is, say, 103°,
then the same signal on the next
line is shifted by 180° to - 77° (using the accepted chroma-phase
scale), and so on down the picture.
On playback, an identification
signal is used to determine which of
the lines is in the 103 ° phase and
this line is displayed as is. The next
line (at - 77°) has to be inverted,
back to 103°. The circuitry used to
perform this function is commonly
referred to as the PAL switch. If
this switching fails, the picture
shows one line normal red, the next
line (180° out) as cyan (bluish) and
so on. The effect was first noted in
Hanover, Germany, where PAL was
developed, hence the name. It's
sometimes called a venetian blind
effect.
TV TEST EQUIPMENT
YOU NEED NOW!!!
SHORTED TURNS TESTER
(Built-in Meter)
For checking EHT transformers including
Split Diode Type, Yoke Windings and
Drive Transformers.
$78.00 + $3.00 p&p
HI-VOLT AGE PROBE
(Built-in Meter)
Reads Positive or Negative 0-50kV
Designed for checking TVs, microwave
ovens and any high voltage equipment.
$84.00 + $5.00 p&p
TV TUNER REPAIRS
Repair & Exchange (Philips ELC 2060)
$17.00 + p&p
Repair (Philips UV 461)
$18.00 + p&p
Repair & Exchange (Philips Turret)
$18.00 + p&p
Repair (AWA MITS Varicap)
$18.00 + p&p
Repair (Pye LM031 Turret)
$18.00 + p&p
P&P $2.00
Cheque, Money Order, Bankcard or
Mastercard:
Y.V.TuNERs)
216 Canterbury Road, Revesby,
NSW 2212.
Phone (02) 77 4 1154
]UL Y
1989
53
I fed in a colour bar pattern from
the generator and it showed quite
distinct bars in the cyan, green,
magenta and red sections of the
pattern. The yellow bar was unmarked, a significant point that
escaped my notice initially.
Any set can show Hanover bars
if the delay line settings are out of
adjustment so the first thing I did
was to go over all the adjustments
as defined in the manual. This made
no difference at all so I turned to
the ident circuitry and looked for a
more obscure answer.
The PAL switching is carried out
in the demodulator chip, IC804, and
is driven by a multivibrator running
at half line frequency. If the
multivibrator starts up in the wrong
phase, the reds and greens in the
picture will be displayed in their
complementary colours. To avoid
this, an ident circuit looks at the instantaneous phase of the colour
burst and develops a 7.8kHz pulse
which steers the multivibrator into
the correct phase.
It was this circuitry that I looked
at next. The manual gives instructions for adjusting the amplitude
and phase of the multivibrator con54
SILICON CHIP
trol circuit but I am sure that there
is an error in them, although I
haven't heard of any errata for this
model.
On page 18 the amplitude alignment calls for adjustment of T807
for maximum response as in
waveform "B" which turns out to
be a square wave with a curious
bump on the top. Then on page 19
the next step uses the same T807
and is supposed to match the
response curve "A", an H/ 2 sine
wave.
Even if waveforms "A" and "B"
are swapped to make them agree
with the instructions, the result is
still ambiguous. The second adjustment is supposed to position the
bump in the centre of the square
wave. But this reduces the
amplitude of waveform "A". If "A"
is adjusted to maximum, as instructed, then the bump is shifted
right off the end of the square
wave.
In the end I ignored the first instruction and settled for the centralised bump. This proved that the
ident circuit was working but did
nothing to alleviate the Hanover
bars.
Once I had satisfied myself that
all the circuitry was working correctly, I began to suspect some subtle malfunction in one of the components - something that would
give the right measurement but not
process signals in the proper
manner.
So I started to change components. I had an old C1831X
chassis that was known to be have
been a good one so I lifted various
parts and dropped them into the
problem chassis. ICs, coils and
transformers, transistors and even
the chroma delay line all went into
the villain, but nothing made any
difference.
By this time I was pretty desperate. The set had cost me nothing
in cash but had piled up a considerable debt in hours. The pity of
it all was that the set promised to
show a very good picture, if only I
could get rid of those (expletive)
Hanover bars. And by this time I
had another customer who wanted
to pay cash for a good secondhand
44cm TV set; eg, a Sharp Cl831X.
During all the preceding measuring and testing, I had frequently
looked at the video waveforms on
the output collectors. The blue
waveform was the classic on-off
shape and only proved that the blue
channel was responding exactly as
it should. The red and green on the
other hand, were showing a curious
sloping waveform that should have
alerted me.
One problem that comes from
working on your own set is that the
job is done in short bursts, when
customer pressure eases off. So you
often find yourself repeating
something done several days
earlier, and the time so wasted
begins to pile up. Also, working so
close to the problem for days and
days tends to blind one to obvious
clues.
And so it was that I had to turn
away the cash customer and put
the old Sharp on the forget-about-it
shelf. I had spent as much time on it
as I was prepared to and the set
would now have to be junked. It had
cost me nothing in cash but 15
hours doesn't come cheaply even if
it's my own time.
A few weeks later, I was talking
to a colleague about the set and its
TETIA TV TIP
Hitachi PA3-A Chassis (CEP288
etc)
Symptom: l.,ow height. If height
readjusted when the set is warm, it
will still be low for a while after
switch-on. Recovers height in two
to five minutes. 20 volt rail varies
from 1 6 to 1 8 volts during warmup
but does not reach 20V.
Cure: C753, a 1 0µ.F 25V electro
defective. Capacity is very low
Hanover bars. He is a college instructor and knows colour TV
theory very intimately but he gets
few chances to actually service a
faulty set. He asked if he could have
a look at it and I readily agreed.
A few days later we set it up on
his bench and the first thing he
noticed was that the yellow bar was
clear of the fault. This was no great
shakes to me but to him it suggested
that the ident was working at the
beginning of each line but fading as
the line progressed.
And when he looked at the video
output waveforms, the sloping
characteristic, which I had noted
but dismissed, confirmed this interpretation; the ident signal was losing its grip as each line scanned
across the screen.
·
when cold and increases only
slightly as the set warms up. This
is the input capacitor to the 20V
rail filter and its loss changes the
filter to inductive input with consequent lower output.
TETIA TV Tip is supplied by the
Tasmanian branch of The Electronic Technicians' Institute of
Australia. Contact Jim Lawler, 16
Adina St, Geilston Bay, Tasmania.
There are no demodulator chip
waveforms in the manual so one
has to fall back on theory to determine what should be happening.
The demodulator has to be supplied
with chroma, a 4.43MHz reference
signal, and a 7.BkHz ident signal. If
all of these are present then the
chip should produce the appropriate outputs.
I had checked the presence of
chroma and the 4.43MHz reference
and had found what I thought was
the ident signal. But I was wrong.
That ain't right
My friend confirmed the first two
signals as present and correct but
when he looked at the ident, on pin
11 of IC804, he said "Oy, that ain't
right!" He was looking at a series of
WE. C.OUL.t>N~
) f'IIIJC> PN'lONE-
~~\~'t'e;.R
~P\N ~OU
~~
-~
e)
~!_l-
2~~~/\<
A'RE'I'm GLA1) 11-\e:RE:
'SOME:
?eopt.£. S'R\GH-rER "ffi-~N N\£.,
O™ER vJ I-SE 1\-\5 WO'RL--.D
WOU'-l> '8~ NE.C...K 1)€et> \N
C> lJ O TE.L.-L-\.f 'S () ••.
short pulses, first positive going
then negative going, at 7.BkHz.
These were what I had assumed to
be the ident pulses but he knew that
these short pulses could not properly control the phase reversal switch
inside the chip.
In fact, what is needed is a full
square wave with a solid, steady
voltage to hold the switch in
whichever position it should be. The
short pulse could only flip the
switch but couldn't hold it in the
right phase. Hence the clean yellow
bar, on the left of the screen. This
corresponded in time to the first
few microseconds after the switch
had changed over, before it drifted
to an indeterminate position for the
rest of the line.
The square wave is also, in some
way, responsible for setting the DC
output levels of the R-Y and G-Y
signals. In the fault condition, the
pulse was enough to establish the
right levels at the start of each line
but then they drifted to some arbitrary level before the next pulse
reset them. This explained the sloping waveforms I had noted in the
red and green channels.
In the end, the fault was so easy
to find that it was ridiculous. It was
C845, a 0.47 µ.F electro between the
multivibrator output and the demodulator chip. On the capacitance
bridge it read 270pF, far too little to
pass a solid square wave. (It was
apparently differentiating the
original square pulses, converting
them to spikes. Ed).
As I said earlier, you can get too
close to the job and miss the obvious
clues. Then again, it helps to be the
full bottle on theory when the
manual offers no help. I thought the
pulses only had to flip the switch,
not hold it in position as well. I'm
glad there are some people brighter
than me, otherwise the world would
be neck deep in dud tellys !
Well J. L., I've learned a lot from
that and I'm sure our readers have
too. I can't recall having encountered the Sharp 1831X, at
least not in any complicated way,
but I'll be on guard if I ever have to
slide one out of the cabinet. I'll
watch those wire wrap leads too.
As for the Hanover bars - that
was really nasty. No wonder it fazed you. (Ouch! - sorry about that).~
JULY 1989
55
|