Return to Amusing Stories page
Satellite Workshop 12
A Pace SS9200 gave me an interesting problem. It came in as "dead" but, after the power supply was repaired, the receiver seemed to have lost most of the Astra channels. In fact only a few Astra 1B programmes could be found and these were at the bottom end of the tuning range instead of the top. It seemed that the tuner had slipped by some 600MHz!
Eventually, however, the answer struck me when I noticed power supply interference on the picture. The problem was obvious when I realised that I'd recently fitted a Grundig "Universal" LNB to my dish. The power supply was sending interference up the cable at a frequency which must have been about 22kHz. As a result, the LNB was switching in its "High Band" Local Oscillator of 10.6MHz, thus explaining the 600MHz shift.
Replacing all the front end capacitors in the power supply cured the fault not that the customer would have noticed with his "Blue Cap" LNB but I'm a perfectionist.
Interestingly, I had almost the same fault on another SS9200 yesterday. This time I had to replace C33 (1000 F/25v) to cure the fault which occurred only when horizontal polarisation was selected.
Those of you who have upgraded to a "Universal" LNB beware!
I saw my first dead Fidelity SR950 recently. The cause was the "Topswitch" device which is bolted beneath the power supply board. It's a bit awkward to remove you need a really hot iron because the tab is soldered to the copper land. I could find no reason for the failure and no other component was damaged. It was, perhaps, a case of "infant mortality".
The SR950 has no factory reset and is causing some problems with local installers. To set it up you have to connect it to a dish system then put it into search mode. It scans the entire frequency band to pick up all horizontally polarised channels then all vertical ones, storing them in memory as it goes. Unfortunately, this can take several minutes and it does seem to store some completely blank channels with nothing to be seen but "snow". If a standard 10GHz LNB is in use, the receiver will still attempt to store Astra 1D channels, although many of them are very weak and "sparkly". This has resulted in several perfectly good receivers being returned as "faulty poor pictures" when, in fact, it's the old LNB which is the problem. Another common problem is to find no decoder messages because the video bandwidth has been set to "wide".
Although this is a basic receiver, therefore, it has proved to be troublesome for the less knowledgeable installers who don't seem to understand its operation. In addition, a small number of receivers give decoder "dropout" if the picture goes bright momentarily. Amstrad have issued a fix for receivers which do this but it's a workshop job requiring the use of an oscilloscope.
In common with the other new Amstrad models, the SR950 and SRD700 uses a hybrid tuner which includes most of the video and audio sections as well. First impression is that this module, which uses tiny surface mount devices, is going to be very difficult to repair when it fails (and I've had one fail already). To replace the complete module is going to be just as expensive as spending the time to locate the problem. I think a lot of these will end up being scrapped as "Beyond Economic Repair" over the next few years.
A PRD800 arrived in the workshop with "No E-E" reported. Now I specialise in satellite receivers and decoders. I haven't repaired TV sets since the days when you could swap the ECC81s, polish them and charge the customer an exorbitant "thirty bob" so I didn't understand this term. However, it referred to a lack of terrestrial UHF loop-through in the RF modulator section, according to a friend. I checked and he was right. Satellite pictures were fine but terrestrial pictures were decidedly grainy.
A quick look under the board revealed a nasty rust patch on the base. This looked ominous but, luckily, the only damage was to a single surface mount transistor whose legs had corroded away. Replacing Q10 and cleaning the board around it provided a full cure. I advised the customer to get the TV aerial coax replaced immediately.
The customer complained that some channels were missing. Suspecting a polarisation voltage problem, I checked the receiver on the bench but could find no obvious fault and gave it back to him. The following day he returned complaining that, above a certain channel number, he could receive only two programmes. This turned out to be so. The receiver would not tune higher than Sky Sports.
Closer inspection showed that the top of C172 (2200 F/35v) was bulging. The cause was not hard to find since the diodes feeding it were also looking the worse for wear. In fact one of them was open circuit. For good measure I replaced all fourteen diodes. They are tricky to desolder without peeling tracks off the board so my usual method is to snip the diodes flush with the top surface and desolder the remaining legs beneath the board. For reference, the tuning supply voltage across C172 should be more than 32 volts.
Interestingly, I had another RD480 in the same week with the same faults but this time the symptom was that it would suddenly give a blank screen five minutes from the end of a movie! After cooling down, it would work fine until five minutes from the end of the next movie (well, that was the customer's story!) It looked like a tuner fault but it wasn't.
This receiver came in "dead" but, after fitting the power supply kit and getting it up and running, I discovered that there was no visible on-screen tuning menu unless the LNB was disconnected to remove the signal. After a lot of probing with my oscilloscope, I resorted to the more basic method of measuring the value of each resistor. Eventually I found that R74 was open circuit. It seems that the resistors in this model are very prone to failure of this sort.
This receiver is based on the SRD510 but with a few improvements to the circuit design. The receiver was showing "Your Card Is Invalid" with a valid card inserted.
Exchanging the decoder and card reader boards for working ones did not cure the fault. Eventually I found that R205 (10 ) was open circuit. This resistor is not fitted in the SRD510 it's one of the later "improvements" for reliability. Oh dear! You'll find it near the front left corner of the main board.
Another problem I'm seeing with these "improved reliability" models is failure of C622 (2200 F/10v) on the 5 volt rail. When it goes open-circuit the feedback circuit raises all the secondary voltages to compensate. The result is usually a line of exploded electrolytics across the main board and the power supply. It can fool you because the only one which does not explode is C622 itself! Unfortunately, the fault usually destroys the decoder module.
If you get one of these on the bench, monitor the LNB voltage because the first sign is 60 volts going to the LNB (which smokes, believe me!)
It's all organic farming, here, (judging by the smell!) so the only mad cow is the one married to a local farmer. I could tell she was mad by the way she banged the shop door behind her.
"It's done it again!" she announced as she slammed a Pace SS9200IRD down on the counter. I didn't need to ask what it had done again. This was the sixth Monday in a row on which she had brought in this receiver. The first time it came in I replaced the usual parts including C9, C11, the BUT11A and the fusible resistors. The second time only the BUT11A and the fuse had failed but I fitted a new transformer (T2) and the large electrolytic (C7) for good measure. Since then it had returned punctually each Monday to have the fuse and transistor replaced. I had checked everything but could find no cause.
This time was definitely the last time. I decided to make a "home visit".
That afternoon, with the receiver working once again, I drove to the farm to check the installation. It transpired that the Living room was actually an extension to the old building and the socket used for the satellite receiver was supplied from the phase used in the milking sheds. The other sockets in the house were on a different phase!
The actual fault was occurring, not because of the difference between the phases, but because of the milking schedule and the power surges caused by the equipment. During the week, all milking was done by hand. On Saturdays and Sundays, however, it was done by machine.
The solution was simply to plug the receiver into a different phase. Before leaving, however, I gave the farmer an invoice for my call out and advised him to get an electrician to sort out the wiring for safety sake!
I'm answering quite a lot of enquiries by email. It takes time but it's interesting because it lets me see the faults which others are finding and quite often I can help. The only real problem I have is a shocking memory. If somebody sends me a question one day then refers to my answer a couple of days later, I can't remember the original question! A typical follow-up mail would be:
"Dear Jack, thank you very much for your reply. Now I have another problem. Can you tell me why it causes wavy lines on the other three channels? Kind Regards....."
When this first happened, I spent a fruitless hour searching through previous replies to find out the original question and answer. Now I don't bother. So, if you get no reply (or a very curt "Huh?") you know why!
I answer between five and twenty messages per day. They come from all over Europe and not just in English. It takes time. If you want to tell me about new faults and solutions, I'm very interested to read them. If you have a problem, I am only too pleased to help; but please list all the relevant details and use a separate message for each model. If you list several models and faults in one message it becomes confusing and I have to spend time in separating them.
Please don't use abbreviations. I don't always understand them.