TV aerial?

Heres an interesting bit off the MB21 site...

The Clyro mast overlooks the friendly border town of Hay-on Wye, best known for its book shops.

Whilst Clyro is a relay Wenvoe, it does so via the chain of Ebbw Vale, Abergavenny & then Brecon. Therefore being co-channel with its original parent station is not a disadvantage, because direct reception of Wenvoe is impossible at the relay station location. The twin installation of choice in this location was usually an MBM48A or XG21A receiving Ridge Hill with a masthead amplifier, diplexed with a smaller group B aimed at Clyro. Most residents of this "English" area were keener in having Ch4 and five from Ridge Hill than S4C, although in some locations Clyro is the only usable programme source. The extra cost of a diplexer and small group B aerial makes very little difference when you consider the cost of masting and brackets for an XG21A, amplifier and PSU, so twin installations are common.

Incidentally, at ground level around the site of the Clyro relay it is possible to receive Ridge Hill very well on an unamplified 18 element group A aerial. Ridge Hill would provide a much stronger signal source than Brecon at the relay, and some of the locals would be happier with English programming.

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Perhaps that other aerial might not be aimed as well as it should;?..

A quick look around Hay shows a mix of reception systems quite a few satellite systems few long "ish" aerials on Ridge hill but you can see the Clyro mast from quite a few locations like this one here..

In fact there seems to be that side of town a shortage of aerials perhaps the signals good enough for loft or indoor ones!

Regarding my other post on this either you do some complex additions with filters and diplexer's on the existing system or change the existing aerial for a group B on the Clyro station or go freesat!...

It's called "The Law Of Inanimate Hostility", or, if you're Irish, "Murphy's Law".

You probably made good a loose c>

Quite.

Well, if it does go on a sulk again, at least we know it's probably the STB, since the aerial clearly does work.

Even if there is nothing but crap on. As ever.

The problem is that an unpowered amp does not just become a straight through connection with no gain. Without power, it will act as an attenuator, and actually reduce the signal available, and in addition possibly distort it as well.

Ah!, can't do anything about that!..

However can you determine why this is now working at all? And what caused it to behave like that?.

It would now be interesting to see just which mux'es it is receiving thats the English one or the Welsh one or possibly both?.

It may also well be that the box is very marginal on the signals it is receiving, longer term an aerial sort out is I reckon in order..

Perhaps that is just what it needed - a bit more signal loss ;-)

I think that STB is flakey mate.

Mind you had a nastyy scare when me server started making screeching noises in the disk drive area.

pulled it down., wouldn't boot.

Opened it up, shook out the spiders blew out the one and only fan..and sheesh. it finally came to life..

Must have been overheating. First warm day of the year.

Anyway that proves that your aerial rig is in principle OK. But I think that the STB is going west.

Still thats cheap enough...

It might well be, but nothing is guaranteed. You might want to get a TV USB dongle anyway since they are pretty cheap and quite useful to have. Sooner or later it will come in useful - they can't get much cheaper...

You have dismantled and reconnected things several times and eventually hit upon a situation where whichever dry joint/oxide coated metal fault(s) there were are now all temporarily making a good contact.

You can be reasonably sure that in accordance with Murphy's Law it will fail again just as some programme you have been looking forward to seeing for weeks begins to start.

It was ever thus. Although on TDTV with a full multiplex you do have a selection of repeated old comedy programmes somewhere. Out of curiosity do you see any of the Welsh signals in addition to the Ridge Hill MUX?

What I suspect may be happening is that a dry joint is rectifying the incoming signal which although it preserves the fundamental frequency adds a lot of harmonic mush across the band. The earliest ever wireless reception technology the crystal set relied on a fine cats wisker wire probing a piece of galena to exploit this rectification effect.

Curious fault you had there that allows setup to find the channels but the decoder not to be able to make any sense of them.

Well Martin, that's where MY knowledge base (software and signal processing) really DOES exceed yours. Because the distinction between MUX channel information, and actual MUX channel data, does not exist until after demodulation.

And the demodulations process is identical for both.

Or rather, what comes out of the demodulator is a single digital stream that contains all the programs on that MUX and all the program data on that MUX interleaved. That's the data stream that comes out of a PC DVB dongle down the USB lead.

IN simple form, the digital part of the set or set top box, or the host pc on a dongle,. will say 'tune to frequency X and modulation type Y' and will expect to see (or if tuning, try to see if) a coherenet error free data stream is coming out. If it is, the the digital part will split that by packet header onto program/channel information or MPEG streams.

That is: the de-multiplexing is purely digital. All issues of RF have long since ceased to be relevant. RF is concerned only with finding the right MUX in the first place, and de-convoluting it in weird and wonderful ways into that primary raw digital stream.

(I say only, because from a digital perspective the RF/demodulation side is simply one small necessary step to get what that digital engineer wants. A nice sold digital data stream :-))

Normally no one cares to understand the insides of the process, but in this case its very relevant. Because it enables me to say

Now of course the OP and most people have considered this to be an RF issue. Working with their knowledge the problem is:

- new installation of *previously working* STB in new location, with

And that's where understanding a bit MORE about how the *boxes* work allows a different interpretation. In fact the STB has coincidentally gone sideways, partially and it seems not permanently, but there and there alone is the only fault that on CLOSE inspection actually fits the facts as reported.

Maybe it got bashed about in the move and there is a dry joint in it. Maybe its got dodgy FLASH memory in it, and the new retuning failed to rewrite it with valid program data, (in fact that is an explanation that certainly fits the facts as reported), and maybe temperature, or a subsequent erase/rewrite allowed the memory to actually get re-written well enough to 'work'. Or in fact by dint of retuning serveral times its being written to a new part of the flash, now.

In the end it doesn't matter, because at user level it boils down to identifying and replacing the faulty part, and unless you are going to get out a soldering reflow worksation, that means getting a new STB 'just in case'

No, it appears not.

Whilst I take on what you say NP any decoder/demodulator has to have a signal that it can decode properly. Whilst the DTV system apart from its compromised bit rates on some channels is a problem we can do nothing about, it is very robust but the decoder has to have a sufficient signal and a sufficiently clean enough one to work on .. and what's important is that signal is constant and won't fall over under adverse reception conditions.

It would be very interesting to put a spectrum analyser on that aerial system to see just what the decoder is "seeing". It now seems that the OP is only getting the Ridge Hill Muxes which might be a good thing:).

Although he's not an avid TV viewer that TX does carry the all the Muxes HD ones too and the English programming, well the Clyro TX is over the border in Wales;!

However having a very local TX even out of band and not that powerful quite close by that will very often put large signals into the amp and whilst I'm not that familiar with that one and how it behaves, its very possible that it won't take a lot to tip that balance and then you'll have unpredictable things happening in the TV spectrum.

Whilst the indicators might say 100% good I wouldn't trust them to indicate what's actually happening all of the time. I think the OP's best bet is to put a diplexer for Group A and B in line before the amp terminate the unused B group input with a 75 ohm resistor and that will filter off the local stronger signals unless this might have already been done but of that we don't know without closer examination.

It might well be that changing the STB might affect an improvement possibly it might with differing better software cope with marginal signals better from what I've seen of them like most consumer grade TV equipment's they are made down to a very tight cost base and its a wonder they work and stay that way!.. Which they often don't!..

I'd on an engineering basis would do more to correct the input signal first;!...

My point was, that the ability to scan and reliably detect the stations, was evidence that all that part was in fact working.

No Tony. I think you miss the fundamental point.

There is no RF issue that could degrade the program content independent of the actual program identification and program guide data.

Its a bit like saying 'I can see the text on a website, but the pictures are garbled, therefore my internet connection is flaky'.

The point is he tried again and it all 'just worked' and picture quality is excellent (or as excellent as DVB ever is) .

Which is again consistent with there being nothing at all wrong with the RF side of things, and something deeply suspect about the box.

You know as well as I do that beyond a certain threshold a better signal into a digital tuner won't make any difference to the quality at all.

Digital demodulators do not fail gradually and gracefully: the band where signal is 'just enough for garbled pictures' is very narrow. Probably less than 3dB in S/N.

IN short it is normally 'perfect' or 'nothing' and that apples to the whole mux, the program guides, the station ID's in it as well as the actual content.

I have never ever had the situation where one channel on a MUX worked and another did not. IN fact the more I delve into the extraordinary bollocks that is DVB modulation, the more respect I have for the people who designed it. It is true spread spectrum stuff. From one bit to the next bit you are looking at a completely different phase shift on a completely different frequemncy. Unless you know the secret, its 8Mhz of pure noise.

Out of which comes a *single* serial data stream. At around 50Mbps for current DVB modulation. And the point is that it is just *one* single serial stream. That is what comes down the 480Mbps USB 2.0 wire from a PC DVB dongle. Not decoded programs. That is done in the computer itself. That is why you can record a program on the same MUX as one you are watching. In fact you can record a whole MUX if you want.

There is no separate 'channel ID' or 'program info' serial stream. That is all 'done in software'. Either that serial stream is OK, or its corrupted, If its corrupted you wont make any sense of ANY of it.

If you plug a hard disk into a USB port, and some files are corrupted, but you could read the directory OK, you wouldn't immediately suspect the USB cable would you? Unles you are Russ Abbot..

Same here. In networking terms the corruptions and the issues are above the raw data transport layer. Ergo it is whilst all good stuff, not actually germane to the problem the OP was having.

In short with digital you achieve a complete disconnect between stations, picture quality and so on with respect to the RF side of things. The RF is there to do one job, and one job alone. Allow reception of a single 50Mbps serial data stream from a single terrestrial multiplex. And either you are getting good serial data or you are not in general getting any serial data at all. If the demodulators cant error correct, it gives up. You cannmot corrupt 'one part' of that serial stream without in general corrupting ALL of it.

Even that can happen... ;-)

That's not been my experience... I have seen plenty of borderline installations where a channel scan would yield a list of the expected channels in each mux, and yet some or all of those were either unwatchably broken up, or simply resulted in the receiver displaying a blank screen or "no signal" message when selecting them.

I even had a situation here prior to DSO where I got near perfect reception on some channels, and stuttering picture frozen garbage on another... now there is nothing too surprising in that you might think, not all muxes were being broadcast at the same power level, and performance at different frequencies may not be consistent for any given reception system. However the interesting point here was that the channel being lost was in the *same* mux as others that were being viewed with near perfect results.

Now unless you are a subscriber to the Russ Andrews school of fancy cable miracle works, you might be surprised to hear that the fix in this case was to replace a section of old downlead from the aerial to the distribution amp. That the cable had a notable level of loss (10dB in about 7m) is not too surprising, that it could in effect "notch" one logical channel from a single multiplex was far more surprising.

Taking the "digital either works or it doesn't" heuristic at face value would suggest this kind of behaviour is not possible. However if you have a look at the actual framing spec[1] you will note that the content protection (both coding and interleaving) is arranged in two levels within the transport stream, and can be applied differently to each of the logical mpeg streams contained in the mux. This allows operators to trade off the level of ruggedness of protection against bandwidth use and stream quality. Also as with any binary representation of a complicated frame format, being able to extract one short subset of data from multiple copies of a larger set in the presence of a high error rate is far easier than extracting all of it. So it ought not be so surprising that you can get the channel information from a data stream where the error rate is too high to get useful video from it.

Even the choice of mpeg program parameters prior to transmission will have an effect on the subjective effect of errors for the viewer. Use of longer GOP counts will result in unrecoverable errors corrupting a greater number of video frames than the same error rate with a lower GOP (again with a trade off against compression level)

While quite possibly true, but that does not mean it can't happen, since I have seen exactly this happen.

Indeed there is some very clever stuff going on.

If there were only one level of coding used then this would probably be true.

Table A in the above doc also makes for interesting reading where they discuss the achievable Quasi Error Free bit rates for the different coding and modulation schemes and the carrier to noise figures required to achieve them. Its also interesting just how effective the error correction is since they start with a very high post viterbi error rate of 2x10^-4, and yet expect near enough error free operation post Reed-solomon decoding.

Again even this is possible... if your basic transport is yielding a 1 in 10^6 BER, its entirely possible that a couple of KB of directory listing will be retrievable without error on most attempts to read it, and yet a 200MB file would not be. Its just down to probability.

Note I am not saying that a dodgy set top box was categorically not the issue here - I agree it is entirely possible. However, I would not go so far as to rule our RF related issues either, and I feel your confidence that the STB was the only possible explanation is misplaced.

This I don't agree with - since with layered coding schemes you can decode a highly errored raw stream into a better but still errored inner coded stream. If that coded stream is composed of multiple logical separate programmes, each with with potentially *different* "strengths" of coding, it seems quite reasonable to expect that after the final level of decoding, some will of those logical channels be significantly more usable than others.

At _that_ time...

Yes you are making the point I was!. This needs to be _consistent_ reception. Yes of course if the received signal is above certain threshold's then it matters not how strong it is as long as the bitstream can be decoded, without the error correction going overboard, then a stronger signal will not make for better pix!..

No I'm not saying that at all..

But I rather think that where you are for a lot of the time, except those fine high pressure days that bring on the CCI;!, you'll see a rather clean, even amplitude across the band signal. I rather think you won't have seen what is possibly happening under the conditions like we have here where we are trying to receive a rather weak distant signal which all other things being equal will work fine as long as the thresholds as alluded to are met.

However we have a local interferer that may well be causing problems in the overall TV frequency band and prolly beyond. The amp on the masthead will in all probability be doing that job well but its large signal handling capabilities won't be that good, not at that price range!. OK its fine for what it was intended to do but they don't cope that well with larger signals that drive them into intermodulation. This is why I alluded to the analyser to see _what_ the receiver is being presented with. It may well be that some of the time this will be acceptable and the "box" will be able to decode, but sometimes not.

I have seen this elsewhere where there are RF carrier channels that are there one moment and almost gone the next, phase addition and cancellation caused by multipath effects causing very lopsided amplitude variation across an 8 MHz channel various MUX channels there, others not. DVB is very good in this respect but it can't cope with signals that are corrupted beyond that point that they can't be decoded and in this instance they may well be OK for short or even long duration's.

When in this overloaded state the results are very unpredictable indeed there will be spurious carriers and products over the signals that are in the range of interest and they will be of varying decode "able" quality. Hence the advice to the OP to filter off those unwanted signals that are present if this hasn't already been done somewhere.

Furthermore these interfering signals may well have thresholds of their own too. The amplitude of them is likely to be very variable as they are off the "rear" of the existing aerial and are quite likely arriving at that as reflections off other "clutter" such as hillsides, buildings and trees etc they may well be OK one moment and not the next sufficient in amplitude to drive the masthead amp beyond its comfortable limits and then not again.

Yes you may well be right it could be a flaky box and come to that conclusion, but if that box is being handed something that it can't cope with from time to time then it may well be getting the blame for another fault thats not to do with it!.

We have had quite a bit of experience with digital microwave systems using OFDM modulation and that will cope with a lot but we have a problematic link where some bloody trees have grown a bit and now its fine one minute and not the next the thruput capacity going from 100 % to less than 5 in very short timescales. Course chuck in some rain as well and its even more fun to watch what's going on;!...

Nope, I could never bullshit as well as that !..

No course not but its a very good indicator that all is not that well in the new location!.