This is only a preview of the May 1988 issue of Silicon Chip. You can view 39 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 "Fit High-Energy Ignition to Your Car":
Items relevant to "Walkaround Throttle for Model Railroads, Pt.2":
Articles in this series:
Articles in this series:
Articles in this series:
|
Double or quits
Yes, this situation could easily toss you. How many
things can go wrong with what should be a
relatively straightforward service job? If you think
you've been there, and know the answer, this story
might well make you think again.
This is not my story, but is from
one of my regular correspondents,
J.L., of Tasmania. More precisely, it
is really two stories about the same
set and, to further confuse the
issue, also contains a story from my
own bench covering a virtually
identical problem. This is how J.L.
tells it.
J.L.'s first story
A customer recently brought in
two identical AW A sets; a rather
rare occurence, but it was too much
to hope that they would both have
the same fault. They didn't.
They were models C608; 48cm
sets fitted with the Mitsubishi "G"
chassis. The fault in the first set
was so . simple that I can't
remember what it was. The second
24,0V,
56
A
one was a different story.
The set was shut down and the
103V rail was up to 120V plus.
There was no voltage on the collector of the horizontal output transistor, nor were there any subsidiary rails operating. Quite obviously, the horizontal stage was
not working.
As I tried to trace the course of
the main rail towards the horizontal output transistor, I wondered
why these sets didn't fail more
often. They are a maze of plug and
socket connections, any one of
which can cause trouble.
From the regulator board the
main HT (103V} rail passes through
an unreliable edge connector into
the main board. It then goes into
and out of a link in the deflection
"'"'1,,'J;
~OOt> CA~AC\T'OP-, ?V,- ACROSS
SHOUL-'t> S-rAV a>OL At-l.- t>A.V•••
SILICON CHIP
toss ya!
yoke plug (DY}, back into the main
board, via plug MF to the horizontal
output transformer (Pin 3}, from pin
2 to PCB-FR via plug EA, through
fuse F591 (1.6A}, then back to the
main board via plug MA. Finally, it
goes via plug MB (pin 1} to the collector of the horizontal output transistor. (See accompanying circuit,
with path picked out in colour}.
I was able to trace the HT voltage
up to the fuse F591, but no further.
The fuse was open circuit, but did
not appear to be blown, just broken.
I hoped it might be a spontaneous
breakdown - but I should have
known better.
I fitted a new fuse and switched
on. There was a rush of sound and I
thought I heard a burst of EHT. But
only for a second. The new fuse failed with only the tiniest flicker.
There was obviously an overload,
but only a small one - and that
meant it was going to be hard to
find.
I started at the horizontal output
transistor. This item unplugs easily
for testing, but in this case it was a
waste of time; the transistor was
perfect. But I changed it anyway.
Rather than blow another fuse, I
connected a 1.5A thermal cutout in
its place and tried again. I had
sound and a trace of EHT for a couple of seconds before the cutout
tripped. Which confirmed that the
horizontal output transistor was intact and trying to function.
The next items on the "output"
side of the fuse are four flyback
tuning capacitors, C531 to 534 inclusive. Two are rated at 2kV each
and the other two at 630V. It's hard
to measure leakage on high voltage
caps with an ordinary multimeter
so my trick is to put them across
240V AC and see if they get warm.
A good capacitor should stay cool
all day and these four passed the
test.
The last item on this board is
--------------·
----------·
J405 ZSC9l6 81< 12sc 1 5tA
V.OUT
'tI
I s,oE·'
I ,...!181
:g-L
.l
~ ~i~
l
---Pr-<. . •-•;~;;, 1':f'"'""""
,w
~
-:.
'
r-'V'\/\r
0~9l
"4VT • lOS
115U
200
.,.
UNIT-HV-RECT
Fig. 1: horizontal output stage for the Mitsubishi C608 "G" chassis. The orange overprint shows the tortuous path
of the main HT rail from the regulator hoard to the horizontal output transistor.
boost diode D531. I checked it
for leakage using 30V from the
bench power supply and could
detect no sign of leakage. It's not a
perfect test, considering the voltage
they have to withstand, but at least
it's a guide.
Tracking back from this board,
via plug MB, I came to diode D571,
a high voltage type that supplies
800V to the picture tube screens.
This rail is bypassed by three 1µF,
450V electros in series. Considering
that low voltage electros are
unreliable beasts, I was prepared
to find one, or all, leaky. I also
suspected the diode.
These capacitors and diode were
as far out of reach as any windscreen wiper motor you have ever
seen. They were tucked away on
the main board, underneath the
"tower" that supports the entire
horizontal output stage, board,
transformer, tripler, etc. The diode
and two capacitors were on top of
the board and one capacitor
underneath.
To make matters worse the
underside of the board had been
covered with some kind of goo
which resembled anti-vibration
compound used in cars. Before I
could unsolder any joints I had to
peel, scrape, and brush this goo
away using Shellite and a stiff bristle brush.
I eventually unsoldered the
underside capacitor and freed the
other two and the diode, but I could
not get these latter out from under
the "tower". This is held to the
chassis with only three screws but
also involves plugs EA, ED, MA,
MB, and MF, plus the focus and
EHT leads to the picture tube.
Of course, after all that, all four
items checked out perfectly.
And those were the last items on
the HT rail. So what was loading
this section of rail? The only
answer seemed to be the horizontal
output transformer
shorted
turns? - or a faulty tripler.
I disconnected the tripler from
the transformer, then switched on
while watching the fuse. It didn't
blow, and there was a very healthy
corona around the bare lead on the
transformer overwind. So it was a
dodgy tripler.
This presented several problems.
First, it was going to be an expensive repair. Second, I had been told
that these units were no longer
available. And, third, the UNIT-HVRECT , as Mitsubishi calls this
assembly, incorporates the focus
resistor chain and a very large
Z00MQ bleed resistor. This makes
any idea of a tripler substitution a
very dicey one.
I asked around among my colleagues whether anyone had ever
fitted a substitute tripler to a "G"
chassis. No one had, but they
couldn't think of any reason why it
wouldn't work. So I resolved to give
it a try.
I separated the rectifier unit
from the focus and bleeder resistor
unit, and found that there was room
to mount a common tripler on the
plastic casing of the focus unit.
Everything went together nicely,
but I deliberately left the bleeder
resistor disconnected at this stage.
Time enough when I knew it was going to work.
Fortunately, I didn't fit the ultor
MAY 1988
57
SERVICEMAN'S LOG
srr
..... ,,.. L.-OC>Ks t\S -m()C.)(p'"' 9™tc:
w,u_
g e:G,.O N\& ~"10"n\E,1\. '"ru"'1"-\ £. .....
cap to the tube. I brought it outside
the cabinet and poked the high
voltage probe into it. Then I switched on. To say that it worked would
be the understatement of the year.
The meter needle slammed hard
against the 30kV stop and a loud
zzzzz came from the ultor cap. I
switched off hurridly.
At this point it looked as though
the bleed resistor was an essential
part of the system. I carefully exposed the wire in the EHT lead, attached the resistor to it, then wiped
a generous layer of silastic over the
joint and left it to cure. I added
another layer the next day and by
the third day I was ready for
another test.
This was just as dramatic as the
first. A bit slower, perhaps, but just
as violent.
Then I did what I should have
done in the first place - looked
more carefully at the schematic
diagram of the "tripler". Had I
done so I would have seen that the
HV-RECT contains only three diodes.
And three diodes don't make a
tripler. I had replaced a voltage
doubler with a tripler. No wonder I
had excess voltage.
So, with no HV-RECT units
available, it looks as though the set
58
SILICON CHIP
will become another junkie, useful
only for whatcan be salvaged from
it. I have yet to tell the customer the
bad news.
My story
Well, that's the first part of J.L.'s
story, and this would seem to be an
appropriate time to take a break
and consider what Yours Truly was
doing during this time. In fact, my
story would have started rather
earlier than J.L.'s because I encountered several long waiting
periods but, unfortunately, neither
of us knew what the other was
doing.
My story is also about a "G"
chassis, but in my case it was a
model C609, 33cm set. And the
symptoms were almost identical
with those observed by J.L. But I
was lucky. According to Murphy's
Law, whichever end of a chain one
elects to start from, it will always
be the one furthest from the fault.
Murhpy knows this and I know this.
But on this occasion Murphy
reasoned that I would reason, that
Murphy would reason ... well you
know what I mean; I beat Murphy
at his own game. I took a punt on
the transformer/tripler end of the
string and picked it in one. It was
the tripler - or what I believed to
be a tripler at the time.
I was lucky in another way too;
nobody had told me anything about
any shortage of triplers and, since
the Mitsubishi Electric AW A spare
parts division was only a 20c untimed phone call away, I simply dialled
the appropriate number. When I
told the bloke at the other end I was
interested in a tripler for a "G"
chassis, he didn't hesitate. "Yes we
can supply a replacement high
voltage unit for a "G" chassis exstock. It'll cost you about $70 plus
tax. I can give you the exact figure
in a moment.' '
I thanked him and said that
would be near enough for now. I
still had to get the customer's OK.
The $70, plus Mr Keating's cut,
would put the price over the $100
mark which, with my service fee,
would make it a fairly expensive
repair .
Naturally, I didn't expect the
customer to be too happy about the
cost, but I didn't expect he would
knock it back. But he did. He
reckoned it wasn't worth it and that
he'd rather put the money towards
a new set. I didn't argue, of course
- it was his decision. I simply told
him that he could pick the set up
any time, no charge.
I put the set in the store room and
more or less forgot about it for
several weeks. Then, meeting the
customer in the street, I reminded
him that I still had his set.
"Yeah, I must collect it sometime.' '
He didn't sound very convincing
and, in fact , this episode, with
variations, was repeated several
times over the next couple of months until, eventually, it became obvious that he had no intention of collecting the set. Well, that suited me
fine ; I had my own plans for it.
More specifically, I needed
another workshop monitor for VCR
servicing and, if I could salvage this
set for $100, it would be ideal.
It's not a tripler
So I placed an order and the new
unit arrived in a couple of days.
And it was only then that I realised
that it was a substitute device,
needing a moderate amount of
mechanical modification to the
chassis to accommodate it. And
~ \UF\S A
1\\00SHT ~
also that it was a doubler rather
than a tripler.
Thus prompted I looked at the
circuit again and it became obvious
that it could not possibly be a
tripler.
Anyway, the new unit was duly
fitted without any serious problems, the set switched on, and up
came a first class picture. It was as
simple as that.
So it was with some surprise that
I read J.L.'s story, set out above,
which came to hand shortly after.
In particular I was surprised to
learn that there was a supposed
shortage of doublers for the "G"
chassis, since I had just bought one.
I realised that it could be a different type, being for a different
size picture tube and, at the first
opportunity, I checked with the Mitsubishi spare parts division. In fact,
it is different, but I was assured
that there were ample stocks of this
model.
I was on the point of writing to
J.L. - in fact I had drafted a letter
- when a sequel to his first story
came to hand. This is how he tells it.
J.L.'s second story
When told the bad news about
his set [that it was unlikely that I
1WSH 0,: SOONt) ANO "!
\EA~O P. ~ OF ~ ' - • •
could obtain the necessary part),
the owner's response was quite
unexpected. He had a special
reason for wanting the set repaired
and was prepared to spend up to
half the cost of a new set to this
end. I was to make any reasonable
effort to get a new part, or a second
hand one as a last resort.
The first thing I did was borrow
the doubler from the customer's
other set and try it in the dud set. It
came good immediately, the picture
tube was in excellent condition, and
a new doubler would make the set
as good as new.
I put in a call to Mitsubishi AW A
spare parts in Sydney and, to my
surprise, was told that, "Yes, 'G'
cha ssis high voltage units are
available" . There would be no
change out of $100, but it was
available, contrary to what I had
been told earlier. I checked with the
customer, then placed the order.
When it arrived I was shocked to
find that it was the wrong type; it
was for the later "K" chassis, not
the "G" chassis. I was about to
throw a tantrum when I picked up
what I had thought was an invoice
and found that it was an instruction
sheet on how to fit this doubler to a
"G" chassis.
There were 11 separate instructions on the sheet, mainly involving
drilling various mounting holes and
re-arranging the focus and earth
wiring. It is a rather awkward arrangement with the focus control
tucked away inside the set, but it
does fit.
Volts but no picture
At switch-on the new doubler
worked like a charm. There was
EHT and sound and I waited for a
picture to appear. But after two
minutes there was no more than a
few coloured specks , and occasional picture highlights.
From these prelimina ry, clues I
deduced that the problem might be
around the video output stages. The
meter showed 140V on the picture
tube cathodes, some 30V higher
than the figure on the circuit
diagram. This same voltage appeared on the output transistors '
collectors and was very close to the
B3 rail voltage which supplies
them, suggesting that they were not
turned on.
The voltages at the base and
emitter of these three sta ges were
also higher than normal but,
significantly, were the same for
base and emitter. The absence of
M AY 1988
59
the normal 0.6V difference between of electrolytics and resistors which
these elements supported the idea
were marginally off, but achieved
that the transistors were not turned
nothing. In desparation I rang a colon.
league who I knew had a good "G"
The base voltages for these three
chassis and arranged to borrow the
stages are derived from the colour video board from it.
difference stages - Q619, 620 &
I was shocked to find that this
621 - on the chroma board. These board was only a little better. There
are driven, in turn, via coupling was now a picture outline and betcapacitors C621, 622 & 623 from
ter highlights, but it was still far
pins 1, 2 & 4 of the chroma IC
from right.
(IC106). The three coupling
capacitors were in good condition,
so the wrong voltage was arising
411:
after the IC. This left Q619, 620 &
621, plus the DC clamper as possible suspects.
The emitter voltages for the three
\
stages are applied via a direct
coupled network that goes back
'
almost to the video detector. Included in this network is the second
'
video amplifier (Q202), the
~ .
\
luminance delay line (D1202), the
video buffer (Q206), and the third
video amplifier (Q203). The signal
then passes via switch S20~ to the
~•--'---"'emitters of the output transistors. ~ ..
Also associated with this net- '----__,.._..._.
►
work is an automatic beam limiter
t W~ Pl\Q\M' ,O '™tM ~
stage, QZ04, which funtions by
if:\tJThUM Wli\£1\J
RXJt\lt)
....
varying the bias on Q203, in
response to a voltage derived from
'M& INVO\CE WAS ~N
pin 6 of the horizontal output
, NS°TRvC1'"cO~ SMEJ:..'"\ ••••·
transformer (T591). This varies the
amplification of Q204, the bias on
the picture tube cathodes, and the
Comparing both boards revealed
diode and a larger capacitor. I
beam current.
only
one
significant
difference:
the
altered
my board accordingly and
Strangely, it was only the emitter
network
R208
and
C203,
between
was
rewarded
with a similar order
voltages on the three video output
of
improvement.
the
video
buff
er
and
the
third
video
transistors which were significantMore importantly, the modificaly wrong. The colour difference out- amp, had been changed in my colleague's
board
to
include
an
extra
tion
had somehow changed the
puts were correct, as were the
voltage
levels around the the beam
voltages on the bases of the three
limiter, Q204. This suggested that it
output transistors. Again, in the
was a beam limiter fault but, also,
TETIA CORNER
video chain, all voltages were close
that
it might not be on the video
to correct and the oscilloscope
Philips K9 chassis
board.
This would account for both
showed that all seemed to be proSymptom: One colour missing
boards seeming to be faulty when
cessing video normally.
completely. Output stage voltages
they had worked previously.
It was only when I checked the
are normal, as are the grid and
Control voltage for the beam
waveforms more carefully that I
cathode voltages on the picture
limiter
is generated across R272
realised there was something ·
tube base board . The G2 voltage
and R274 in parallel, these being
wrong. The level at the delay line
for the missing colour will be low,
connected to pin 6 of the horizontal
input was about normal at 3V p-p,
although there is no measureable
output transformer and diode D595
but at the output of the third video
leakage on the rail.
and
C594 between pin 6 and
amplifier it was les.s than half the
Cure: the beam switch on the conchassis. Diode D272 is suitably
..6.5V p-p shown in the diagrams.
vergence board has developed a
biased to prevent this voltage
leakage path at high voltage. The
reaching the limiter transistor until
easiest repair is to remove the
Lost gain
a certain critical level is exceeded.
switch and bridge the appropriate
I went right over this network, but
Clearly, the video amplifier had
pads. The switch is not essential to
found nothing wrong.
lost a lot of gain. Yet I could find
the operation of the set.
In desperation I tried to brute
nothing wrong. I changed a number
,·~
__ _
_L.J_\~--
,·.. \ '
. '
-
ni~T
60
SILICON CHIP
r
force a change in voltage levels
from the limiter transistor onwards. To turn the third video
amplifier on harder I patched a
22kf! resistor from base to chassis
- and was rewarded with an immediate increase in screen
brightness.
But the picture was still so washed out as to be virtually useless.
This agreed with the very low
amplitude video signal that I had
noted on the CRO earlier, and convinced me that it was the beam
limiter system that was falsely biasing the video stage. There aren't
many components in the beam
limiter circuit. I checked them all
and found nothing wrong.
Finally, I came to a conclusion
that I should have reached hours
before. Near the beginning of this
story I mentioned that the set had
worked perfectly with a borrowed
doubler. Nothing had changed since
then, so it should still work perfectly with an original "G" chassis
doubler. Therefore the problem
must due to the new "K" chassis
doubler.
Perhaps this unit is not a direct
substitute for the original. Perhaps
there is something different about it
which has to be allowed for.
Granted, the modification notes are
purely mechanical, with no mention
of electrical changes. But this
would not be the first time that
essential information has been
omitted from such notes.
The service manual decription of
the beam limiter action suggests
that the key voltage is that at the
junction of R272/274 and diode
D272. The limiter bias on Q204
varies with this voltage, so I decided to change it and see what happened. The easiest way to do this
was to remove one of the two
resistors. This would increase the
resistance and indicate in which
direction the change should go.
With R272 out of circuit the picture became much darker, so this
was the wrong way. R272 is 180kf!
so I replaced it with lO0kf! and
achieved an immediate improvement. I continued reducing R272,
each time getting a brighter picture
until finally, at 33kf!, I had a
perfect picture with a small reserve
in the brightness control range.
It is not my usual practice to re-
design a set while I am repairing it.
I normally acknowledge that the
original designer knew better than I
what value should be assigned to
each component. Yet there are
times, like this, when what looks
like a communications breakdown
leaves me wondering just who is
responsible for advising of changes
in circuit values.
In this case it appears that the
replacement doubler has significantly different characteristics,
requiring a change in external circuit values to achieve correct
operation.
The trouble is, the hours it took
me to solve this problem are not
chargeable to anyone. I have to
bear the cost and hope that a
similar job comes to light in the
future to help me recover some of
the losses.
Conclusion
Well, that's what happened to
J.L., and it's not a pretty story. Not
to put too fine a point on it,
somebody goofed and, in my book, it
wasn't J.L. Remember that J.L. had
been able to test the chassis with a
known good doubler and establish-
ed that there was nothing else at
fault.
So what went wrong? J.L. suggests that Mitsubishi neglected to
include all the information covering
the substitution; ie, the information
was available, but not enclosed.
My bet is that the information
was not enclosed because it didn't
exist. Could it be that someone had
established that the "K" type
doubler could be used with the "G"
chassis, but only on the basis of a
trial in one, or perhaps two, "G"
chassis types?
While all " G" chassis are
basically the same there are minor
differences according to the size
and type of picture tube used. As
well as whole range of picture tube
sizes, some sets use the delta gun
tubes, others the in-line gun version, and EHT and other element
voltages are provided to suit.
So was the model C608 the odd
one out, unable to accommodate a
slightly different EHT voltage
and/or loading created by the
substitute doubler?
I don't know, but I'd like to find
someone who does.
·~
MAY 1988
61
|