Return to Amusing Stories page
Satellite Workshop 21
Archie is a Scot, or so he tells everyone. However, he has a London accent and my opinion is that it's just an excuse to wear a skirt! Mind you, he does like to watch those questionable films after midnight.
It was for this reason that he paid me a visit with his Amstrad SRD650 receiver. The internal D2Mac decoder was, he said, "playing up." Apparently it had announced "check the card" for a few days before expiring completely.
A quick check of the power supply revealed that the large electrolytic, C609, was still charged up to 300 volts. I discharged it with a 10 resistor - this is preferable to using the palm of your hand, believe me! The power supply is similar in design to that used in the SRD510 series but with some essential differences. C609 is a 150 F and the two high voltage resistors, R602 and R603, are 100k . The manufacturer never did quite catch on to the fact that these break down by arcing rather than by excessive dissipation. I always replace them with 350 volt resistors rated at 0.75 Watts. After replacing thousands in SRD510s, I've never had one fail. (A shame - I could have made a fortune with the repeat business!) The correct type is supplied in the SRD650 kit from most suppliers (try Economic Devices or SatCure).
With new resistors fitted, the receiver appeared to work perfectly, however it still announced "Check the card" when a valid card was put into the slot. A voltmeter check showed that the 5 volt supply "Vcc" was not reaching the card contacts. The fault was caused by an open-circuit 10 resistor at the front left corner of the main board. This is a current-limiting resistor and seems to fail when non-approved smartcards are used.
It has been a D2Mac month. Rumours abounded that certain satellite channels were going to disappear to the "Thor" satellite at 0.8 W and some would vanish altogether. To watch programmes from "Thor" you would need a rather large dish in the UK. (3 metres, I am told, although 1.5m is just about large enough in London if the receiver is good). The BBD-901 D2Mac decoder was imported into the UK under the "FilmNet" badge and, as such, had special program firmware built into the Eprom which caused it to require the viewing of a FilmNet channel at least once a week. Consequently, with the loss of this channel from Astra, many decoder owners were left with equipment that stubbornly refused to decode any channel.
The answer is quite simple since the decoder simply needs to see a certain code inside the memory chip in order to make it work happily again. I can't publish the code because it is undoubtedly copyright but most repairers are aware of it by now. A User Instruction booklet containing lots of useful notes including Service Menu information and notes about suitable handsets for this decoder is available for 9.95 inclusive from SatCure, PO Box 12, Sandbach, CW11 1XA. The handset that I use is a Philips part number RC6932/01 Order code 4822-218-21137 from Willow Vale. Although this is rather expensive for something you might use twice a year in the workshop, it has the bonus of operating the STU824 satellite receiver as well.
A smart looking lady wearing tweeds and a pork pie hat marched into my repair shop. I could tell from this uniform that she was a Landowner and the two dogs were a dead giveaway, as was the four wheel drive vehicle parked outside.
Her particular problem was a Pace receiver which whistled. I doubt if many people dare to whistle in her presence so the PRD was taking a big risk. She left it with me and instructed me to "phone when ready."
I'm seeing a lot of PRD800/900 receivers with a whistling power supply. This is often coupled with output voltages that are slightly too high. Invariably the culprit is C5 (22 F/35v/105 C) but I replace C6 and C7 as a precaution. Sometimes, however, the voltages remain high, with the 5 volt supply reading 5.5 volts which is really over the limit. I've changed every component in the supply without success so, recently, I've started to change R12 to 15k . R12 is a surface mount 22k resistor which sits in parallel with R11 (2k0 ) and sets the output voltages accurately. 5.25 volts appears to be the norm for the 5 volt rail. It's all a bit of a compromise. If you set the 5 volt supply too low, the 13 volt supply falls to 12 volts or less and gives problems with some makes of LNB if the cable run is long.
In this particular case the capacitors effected a complete cure. Since the receiver was an earlier version with no 9.75GHz menu option, I phoned the Lady to ask if she would like me to upgrade it. This involved either replacing the microprocessor to give the 9.75 menu option or simply uploading the channel frequencies, to tune each one 250 MHz higher, using the Pace Link system on my computer. Being the rich and affluent type, she chose the latter, cheaper, option. I kept the price down by supplying an Enhanced LNB rather than a Universal and gave her the number of a local installer who uses very basic coax. I warned her that the whole lot would have to be replaced for digital MPEG-2 but she sniffed and told me to mind my own business!
I still get a lot of enquiries in my e-mail about faults in PRD receivers, most of which I have previously covered in these pages. "My receiver whistles" or "my receiver flashes on and off" is the most common. Replace C5, C7 and C8 with high temperature capacitors, keep the unit cool, and the fault will not occur again.
"I get flashing lines or a blank screen from the RF output but the Scart output is all right." This is usually Q105, a BC846B, which is in the centre of the underside of the board. Pace recommend removing and discarding R559 close to it in order to improve reliability. This resistor was added to the design at the request of an OEM seller but its effect is to allow too much current to flow through Q105. This fault occurs only on those later boards which have component idents on the underside. Earlier versions do not suffer from the problem.
Now that this model is a few years old I am seeing a lot of degraded electrolytics on the secondary side of the power supply. C278 will cause a noise that I call "next door's vacuum cleaner." Other symptoms vary and you should check for ripple on the supplies or, preferably, simply replace all the electrolytics in front of the transformer. Be sure to fit the correct parts as used by Pace.
This receiver also turns up as a BT SVS200 or under the JVC or Alba badge. It was cleverly designed for quite good performance and ease of manufacture but is difficult to repair when it goes wrong. However, a "dead" receiver usually needs just a fuse so I was rubbing my hands with glee when a young man marched into my workshop and announced that his bedroom receiver did not light up.
"Go and play outside, sonny, and I'll fix it in a twinkle. It'll be about a tenner. Where's your Mam?"
The spotty child ran his sleeve across his nose and informed me, in a surprisingly deep voice, that he was fifteen and didn't require his Mam. (For all you southerners, the translation is "Mum", by the way.) He strode outside and disappeared up the gennel. (Dialect for "ginnel" or passageway.)
Unfortunately, the fuse was intact. The power supply voltages were present but all were too low. Luckily, I have a copy of "The Repair Manual edition 4". "Replace TL431 adjustable zener and the CNY17-2 optocoupler".
With much swearing and use of the desoldering station, I extracted the offending semiconductors from their nasty plated-through holes. New replacements soldered in place produced -- nothing! The fault was still there.
Meanwhile, the spotty youth had already twice poked his nose around the door and asked me, cheekily, "How long is a 'twinkle'? I wanna go 'ome fer me dinner." I was determined not to become frustrated. I have a friend in the repair trade who tells me that he usually "solves" such problem repairs by placing them carefully on the floor, jumping on them with both feet, then ringing round his colleagues (me, usually) in a desperate attempt to find a replacement!
A glance (actually about half an hour) at the circuit diagram finally convinced me that the culprit had to be Q53 (a BC546B). Using all three hands to hold the magnifying glass, solder and iron, I succeeded in replacing this flea-size object. Wonder of wonders, it worked!
The spotty youth paid up his tenner with bad grace. I stuck my tongue out at him. Two hours' work for this. Huh!