Return to Amusing Stories page
Satellite Workshop 18
I had my first look at a Prima when a local installer brought one in for repair. "It's dead" he announced and scarpered. I did wonder why he didn't simply return it to Pace under warranty but it was too late to ask. I soon found out why: several cracks radiated outwards from the PCB mounting pillar, close to the power supply. Luckily, none of the cracks had reached the live side of the circuit so I decided that a safe repair was feasible. I did not attempt to solder fine wire directly to the damaged tracks but, instead, soldered "Kynar" wire (teflon-insulated single-strand) across solder pads nearest to the breaks. I used blobs of glue to secure the wires to the board ("Tak-Pack" is best). This is an essential operation from the safety aspect because a loose wire might conceivably touch part of the live circuit.
When the installer returned he seemed surprised that the receiver now worked. "Spaghetti junction underneath?" he asked. "Wires everywhere. Will it be reliable?" Since I had reinforced the board with epoxy resin, cured with the aid of a hair dryer, and I was happy with my neatly soldered "spaghetti", I assured him that it would be perfect. However, I spoke too soon. Within half an hour he burst angrily into the workshop and slapped the receiver down on the bench. "Won't respond to the remote!"
I connected the receiver and tested it with my own handset. No problem. On questioning, he admitted that he was using an MSS200 handset which, although similar in appearance, uses a different code set. The Prima handset can be distinguished by the designation "RC-10" at the bottom right corner of the keypad.
Remote control handsets have many uses: apart from dislodging the cat, stirring tea and throwing at the spouse, they are often used to change channel! The small, uncomfortable-looking man, who trailed in after a contrastingly large and domineering lady, clearly fell into the "spouse" category.
"'E 'it it wi' 'is 'ead" explained the lady, demonstrating the point by tapping a handset on the poor guy's bonce. "Now it's gone dead. We'll be back for it in an hour, shall we?" she commanded.
I decided that this job was urgent and pushed aside a FilmNet decoder to make room on the bench. The handset was for the Pace PRD range. I checked the batteries which measured a little over 3 volts in series. There was no infra red output when the buttons were pressed. Pace handsets are not the easiest to take apart and many people either wreck them or give up. My method is to warm them up with a hair dryer which makes the plastic less brittle. I twist the casing with both hands until I can see the split line start to open. Then I insert a blunt penknife blade and gently ease it along, unclipping the two plastic shells as I move it. Don't be afraid to twist the handset hard. The circuit board will not suffer and the plastic won't break if you have made it nice and hot. This method works with Amstrad handsets, too.
The usual problem is that the 455kHz ceramic resonator legs snap. Simply solder two wires into the board and solder the legs of a new resonator to the wires. Glue it down, unless you like repeat business! Occasionally the LED fails instead. I test it by connecting 12 volts across it via a 1k resistor. An infra-red sensitive key fob is all that's needed to see if it works.
The lady owner returned after 59 minutes, snatched the handset from the counter and demanded a fiver from the small man whilst testing his cranial reflexes with the newly repaired weapon.
"'Ow much?" she demanded.
"Erm, a f' fiver will do nicely." I stammered. Not to appear too downtrodden I qualified that with "Special discount, of course."
No sooner had she gone than the milkman appeared with a handset. This was for an Amstrad SRD510 and extremely sticky to the touch.
"Dropped it in the milk, did you?" I inquired.
"Coke" he countered, waving a bottle of well known fizzy drink in front of my face. "Spilt half the blooming bottle, darn it. Low sugar, too."
I dismantled the handset, using the hair dryer to soften the plastic. Inside was brown goo. I used a toothbrush and Isopropanol to clean the silicone rubber keypad and the conductive contacts on the circuit board. Then I scrubbed the plastic housing shells with warm water and detergent. The hair dryer quickly removed the moisture from all the parts and when I reassembled the handset it worked perfectly.
"Just a tenner," I told the milkman on his return. (He's smaller than I am. Anyway, he'll get some of it back as his Christmas bonus!)
A dealer brought an SRD500 to my workshop last week. It bore the label "LIVE" but no other information. The dealer was reluctant to explain, muttered something that sounded like "kit" and scurried back to his van. Being a cautious sort, I did not plug it in but removed the cover instead. Just as well; the white ceramic radial-lead surge-limiting resistor had been replaced with an axial-lead type. The live wire had been pressing against the steel top cover. I resolved to ask the dealer if his public liability insurance was up to date.
My brother is manager of a local trout farm. He brought me a PRD800 which belonged to one of his lads. Apparently it would decode "Sky One" but no other channel. The card had tested all right in a different receiver. Suspecting a problem with the contrast setting, I plugged the receiver in. It didn't light up at all but there was a nasty smell of fish. Fearing the worst, I removed the cover. The receiver was perfectly clean inside so I reconnected the power and tried again. This time the display lit up dimly. I could change channel but there was no picture or sound. The dim display and fishy smell gave me a clue. I measured the microprocessor supply and found it to be only 3 volts! Sure enough, C15 had leaked, hence the smell and the low dc measurement. Replacing C15 restored the receiver to normal and the card produced clear pictures on all encrypted Sky channels.
Some customers won't settle for medium-rare; they like their satellite receivers to be well done. The MSS1000 which arrived by carrier from Birmingham was one of these. I knew it had been cooked when I saw that the rubber feet had disappeared, leaving a sticky mess behind. The circuit board had gone completely black in places and my first move was to replace every single electrolytic capacitor in the power supply section. This cured the "dead" fault (actually it was "tripping" at high speed). I left the receiver on test and, after an hour, the decoder messages disappeared. Liberal application of freezer aerosol pinpointed the culprit as the PTV111 sync separator chip. However, replacement (twice!) failed to eradicate the fault. After some thought, I replaced all the electrolytics around the I.C. and, as usual, it was the very last one a 1 F. Next time, I really must remember to replace the last one first!
From time to time I still see Ferguson SRV1 receivers (the Pace SS9000 clone). The large, buxom lady who brought in the latest one obviously meant business as she thumped it down on the counter and informed me:
"I just collected this from that silly sod, Wossname, up Church Street, you know, anyway he said the power thingy's dead but he couldn't fix it and he's had it six weeks so I've got it back now and I want it tomorrow otherwise HE's going to go bonkers 'cos of the football, you know."
I took advantage of this pause for breath to inform the large, buxom lady that it would certainly be ready tomorrow, yes definitely, madam, without fail and no, we don't close for lunch.
An hour later I was beginning to regret the rash promise and starting to understand why Wossname, up Church Street, had been unable to fix it. Something was Wrong with it and I couldn't figure out what. I'd replaced every single component in the power supply, including the horrible little surface mount ones, and still it wouldn't work. All the dc voltages on U23 were correct but it simply refused to oscillate.
With the aid of a cup of tea, I compared the board with a scrap one. About to give up, I suddenly spotted the problem! Pins 14 and 15 on U23 are supposed to be joined via a common pad but, on the offending SRV1, somebody had run a knife between them, neatly dividing the copper pad. Clearly Wossname, up Church Street, had assumed a short circuit was present and had removed it. So skilful was his work, however, that it needed an eyeglass to confirm it. A blob of solder cured the fault at once.
I've mentioned this model previously because it tends to suffer from memory corruption if the power supply fails and also because it seems prone to dry joints on the infra red sensor pins. A variety of symptoms is produced but most common is that the LNB voltage can not be changed or else odd things happen in the menu settings. Martin Pickering of SatCure has asked me to point out that his new "Satellite Repair Manual edition IV" incorrectly lists a factory reset code which is actually for an earlier model. Unfortunately, the STU824 actually has NO reset code so the only way to reset it is to download the information from an identical model or to replace the EEprom which is a 24C16.
picture supplied on disc
It seems that every person in Scotland owns one of these; possibly because they were produced there. Unfortunately, they do not seem to tolerate the moderate warmth of Glasgow, and the plastic cases become brittle as a result. The Scotsman who dispatched one to me clearly didn't understand that "packing" means more than a Sainsbury carrier bag with my address glued to it. The receiver was well and truly smashed. Luckily, I still have a few "unrepairables" in the workshop and I was able to use the mouldings from these. The initial fault symptom was that the screen was black, apart from the top few centimetres which were white. This looked like a power supply fault so I replaced all of the large electrolytics.
The improvement was noticeable but the picture was still obscured by hum bars and interference lines. I removed all fourteen 1N4003 diodes by the usual method of cutting each lead flush with the board then poking the remainder through with my soldering iron. This is the only way to avoid damage to tracks and pads. With new diodes neatly formed and soldered in place, the picture was almost perfect. However, the sound was extremely crackly and a loud hiss was apparent in the background.
I replaced the middle crystal, the U2829 FM demodulator IC and all the electrolytics around it. The sound was much improved but still a little sibilant. Since I'd already spent an hour on the job, I decided to call it quits. I've heard worse sound, straight from the factory.