Alice at the greengrocers gave me a nice smile when I walked
in.
"Am I glad to see you," she beamed. I wondered if it was my lucky day
but my hopes were quickly dashed.
"I've got lines all over."
"Sounds nasty. Seen the doctor?"
"No, silly, on the Sky." I looked out of the window and Alice went
red and looked flustered.
"On MY Sky at home," she enlightened me. I gave her a puzzled look
and smiled condescendingly, while backing towards the door.
"She means the statellite," (sic) explained her husband, winking at
me.
I nodded knowingly and offered to take a look.
Back at the workshop I could see a whole mess of lines across the
picture so my first move was to fit all the low ESR capacitors from
RELKIT 1. In addition, I used my "Genie" ESR meter to check all the
10uF capacitors around the "Nicky" chip, U9. They all gave a high
impedance reading so I changed the lot. Usually, this cures such
symptoms but, although the picture was much better, there were still
soft white horizontal lines moving up and then down the picture. I
determined that the TV Scart picture was now perfect and the lines
affected only the RF output and only on horizontally polarised
channels when the LNB was receiving 17 volts.
This rather threw me "off the scent" for a time, as I checked the 17
volt supply for noise (less than 10 millivolts) and measured the ESR
of all capacitors around the power supply. Eventually I realised that
the LNB was taking more current from the 17 volt supply than from the
13 volt supply (for vertical polarisation) and the effect was to
change the frequency of the power supply and to increase the ripple
on the 5 volt supply to the RF modulator. Goodness knows why it did
this but I replaced capacitor C26 (10uF) with 1000uF/16v and that
fixed the problem.
Since then, I've noticed that the fault can be even worse on later
models where the warm 12 volt regulator "REG 1" sits next to C26.
This capacitor is now included in "RELKIT 1" by SatCure.
A local installer called to ask for advice. His customer's MSS300 had
suddenly developed a blue screen with "no signal" message emblazoned
across it. He had no way to measure LNB voltage (!) so I suggested he
connect the LNB to input 2 and select that in the on-screen menu. He
was immediately rewarded with pictures but whined at me
nevertheless.
"I don't have to reprogram every channel to input 2, do I?"
"Looks like it," I replied encouragingly. But he moaned a lot so I
suggested he bring it to the workshop. I suspected "Q25" had failed
and was surprised to find LNB voltage was present. Obviously not
"Q5". I pressed "F" then "store" to turn off the blue screen
generator. The TV screened showed "snow" on every channel.
Replacing the tuner had no effect and voltage measurements around the
tuner showed that everything was working perfectly. So why no
pictures? It occurred to me that power supply interference might be
switching my "Universal" LNB to "high band" so I set the receiver
into frequency scan mode. Sure enough, pictures appeared at the top
end of the tuning range!
However, a quick prod with my oscilloscope showed no ripple on the
LNB voltage supply so that idea was wrong. I checked the installation
menu. Input 1 had "10.750 GHz" selected for the LNB! As soon as I
changed it to "9.750" normal operation was restored. I wasn't sure
who won the "Idiot's Prize" for this one. The installer for not
checking it, me for not realising sooner or the customer who must
have altered the setting.
This receiver arrived with a label that said "Sky News scrambles
after 30 seconds." Indeed it did, and the other encrypted channels
did, too. I had never seen this symptom before but I had my
suspicions that it was a decoder switching problem, rather than a
decoding problem per se. I replaced the PTV-2 chip with one from a
scrap receiver and this solved the problem.
Decoder problems can be terribly difficult to diagnose without the
correct equipment and I was probably lucky with this one, although
experience plus a basic knowledge of how the signals are routed
always helps.
I had some bathroom scales weighing heavily on my mind the other day.
Well, actually, it was an MSS100 that *looks* like bathroom
scales.
Arthur is in charge of the railway crossing - a tedious job since we
see about one goods train a month. Underneath his cabin is a heap of
scrap iron and, cleverly disguised amongst it is a rusty satellite
dish. A piece of rust-coloured coaxial cable carries the signal to
his cabin where the MSS100 produces pictures on a tiny portable TV.
I'm not sure if the authorities know about this distraction and I'm
reluctant to ask.
Anyway, Arthur brought me his receiver and explained that, suddenly,
the only picture he could see was a blue screen with the words "no
signal." He had had Joe round to check the dish and that was
fine.
The MSS100 remained on test all morning but did not show a fault. I
needed the bench space so I dismantled the receiver, using my pen
knife to release those nasty plastic clips underneath. I resoldered
the notorious "dry joint" on diode D18 and screwed the unit back
together. That should have done the trick.
Unfortunately, it didn't. Arthur was back next day with the same
problem. He had borrowed an Amstrad SRD400 and that worked perfectly,
but he wanted to watch Channel 5 TV and the older model would not
tune low enough so the MSS100 was now "urgent".
Once more I left the receiver on "soak test". It sat for five hours
before the blue screen appeared. I turned off the blue screen
generator by pressing "F" then "store" and was rewarded with a
totally blank screen. Freezer spray and hair dryer failed to bring
back the picture but I could hear normal sound which altered as I
changed channels. It seemed to me that the tuning and audio sections
were fine but the video signal was being interrupted. This was
probably a decoder fault but it just might be a problem with the sync
separator. I replaced the PTV111, the 1uF electrolytic and the 503kHz
ceramic resonator next to it. These are common failure items in the
MSS models but I was guessing, really.
I left the MSS100 on test over night and Arthur collected the
receiver the next day. He hasn't been back since then so the repair
must have been successful. He'll probably come and pay me one
day.
Could you include a sketch of Arthur's railway signal box with a
satellite dish amongst the junk below?
People are beginning to ask me questions about digital receivers. In
general I can't help because I don't do digital repairs. There are
several reasons for this. 1. Most units are still under warranty so
people don't bring them to me.
2. Pace is offering training courses for repairing their Sky Digibox
receivers but the investment in equipment is quite high for a "one
man" business such as mine. In addition, the diagnostic software
copyright belongs to BskyB (I understand) and can not be released.
This fact means that fault-finding is much more difficult.
3. The Digibox hardware is also being made by other manufacturers who
are not (at present) offering training courses.
4. The Digibox is so heavily subsidised that the customers perceive
it as a 50 box rather than a 300 box. Consequently, they expect a
very cheap repair - despite the fact that a single memory chip could
cost up to 40 !
5. Several other Digital receivers are "grey imports" and no service
information or parts are available.
Of course, I could "dig in" without a circuit as I have in the past.
However, although power supply repairs would be feasible, very little
else would be.
The basic problem with a digital receiver is that there is no video
or audio path to follow. The signal comes out of the tuner, vanishes
inside a very fast computer board, which carries only multiple binary
signal paths, then emerges through a process of digital-to-analogue
conversion as video and audio. Nobody repairs computers at component
level any more. It's not cost effective when you can buy a complete
mother board for 50 or less.