Maps like this show the signal levels expected for certain types of reception, and can clearly show any nulls in directions due to radiation pattern restrictions or terrain. Similar maps exist for France, too.
The problem is that no maps show large trees or buildings, which are one of the most common cause of poor reception.
As I`ve said countless times, you`re better off working out your own TV signal strength, it`s likely to be more accurate......
Watch out! The numbers in the files are attenuation, not gain. 0 means no restriction in that direction. Excel doesn't understand how to do this.
To get a usable result, subtract the value from 20 (as -20dB is the largest attenuation in the files).
I get this for the same file: http://twitpic.com/4wq7pj
Compare with a map - if your radiation pattern was right, the transmitter would be broadcasting to a mostly unpopulated area, rather than the town of High Wycombe itself.
There are 2,700 files and not all of them have the same format. Some provide a 3D radiation pattern.
OK cheers for that, easy enough to do in a spreadsheet at least. I haven't worked out how to sensibly plot the 3D ones yet.
Excellent, one down, four thousand to go.
Actually, is it neccessary to do all the MUXES for all the transmitters ? Isn`t there a good chance the radiation pattern for one MUX (at any particular transmitter) will be the same for the others ?
Excellent, one down, four thousand to go.
Actually, is it neccessary to do all the MUXES for all the transmitters ? Isn`t there a good chance the radiation pattern for one MUX (at any particular transmitter) will be the same for the others ?
I don't think so, I've plotted a couple of sites with all 3 PSB muxes on and the data is identical, and 'Pop' has found the same.
Thanks for the polar plot info. 'Pop' - i'll try and digest that later today when I have time. i've got a Smith chart plotted in Excel somewhere, maybe I can use that as the basis of a template.
I don't think so, I've plotted a couple of sites with all 3 PSB muxes on and the data is identical, and 'Pop' has found the same.
Thanks for the polar plot info. 'Pop' - i'll try and digest that later today when I have time. i've got a Smith chart plotted in Excel somewhere, maybe I can use that as the basis of a template.
When you say "I don`t think so", is that :
A There aren`t 4000 more to go.
(or)
B The radiation patterns for all the MUXES at any particular transmitter are not all the same.
Whatever, if anyone can get a polar plot (like the one linked to) for all the transmitters I think that`d be very useful. I`d definitely put a link onto such a website from quite a few of the pages on my website !
The big question is whether this data is relavant since most main antennas have been replaced for DSO, and if I understand correctly the data precedes that event.
When you say "I don`t think so", is that :
A There aren`t 4000 more to go.
(or)
B The radiation patterns for all the MUXES at any particular transmitter are not all the same.
I think he means that from a very limited number of files looked at, the plots at any one site are the same for all muxes. Whether this is true for all sites we can only find out by checking all the files.
It might be quicker for somebody to raise a new FOI request asking for the actual plots, rather than the raw data!
The big question is whether this data is relavant since most main antennas have been replaced for DSO, and if I understand correctly the data precedes that event.
I doubt very much whether the repeater transmitters radiation patterns have changed, the small ones at any rate.
In any case Mr Butterworth says quite specifically which info is post DSO and which is pre DSO.
I really think this info could do with being publicised in a readily accessible form. A good example is the High Wycombe pattern shown earlier. If you lived in Downley you`d think you`d get a good signal off the High Wycombe TX, but it`s attenuated in your direction ! You can often make a good guess which direction the signal is transmitted in if you can see the anennas, but High Wycombe`s seem to be hidden within the shroud at the top.
So after several hours this weekend of programming - and another couple of uploading - I have the complete post-DSO set of plots.
- Most of the main transmitter sites are redacted (missing)
- Many relays in the south and east are also redacted
- Some plots are possibly dubious (King's Lynn really completely omnidirectional? Hmm)
- No vertical (tilt) data was used. For 3D patterns the peak horizontal pattern was used.
I've plotted all the muxes for a transmitter on the same chart, using as few lines as possible - that is, if the pattern for more than one mux was the same, they share a line. For most of the relays there is only one line.
In a few cases the plot never touches the outer edge. This is because the data for that mux never had a 0 dB point.
The images are organized by original parent transmitter (i.e. by first three digits of site ID). Some relays were reattributed in the past, or have been reattributed for DSO.
The pre-switchover data is in a different format and I've run out of enthusiasm for the moment. Again, most of the main transmitters are redacted so it might not be that useful anyway.
So after several hours this weekend of programming - and another couple of uploading - I have the complete post-DSO set of plots.
- Most of the main transmitter sites are redacted (missing)
- Many relays in the south and east are also redacted
- Some plots are possibly dubious (King's Lynn really completely omnidirectional? Hmm)
- No vertical (tilt) data was used. For 3D patterns the peak horizontal pattern was used.
I've plotted all the muxes for a transmitter on the same chart, using as few lines as possible - that is, if the pattern for more than one mux was the same, they share a line. For most of the relays there is only one line.
In a few cases the plot never touches the outer edge. This is because the data for that mux never had a 0 dB point.
The images are organized by original parent transmitter (i.e. by first three digits of site ID). Some relays were reattributed in the past, or have been reattributed for DSO.
The pre-switchover data is in a different format and I've run out of enthusiasm for the moment. Again, most of the main transmitters are redacted so it might not be that useful anyway.
Excellent work.
If I put a load of links to these, can we just confirm that the page addresses will stay the same, for years that is !
As an aside, I think it`s absolutely disgraceful that the radiation patterns for the main transmitters are missing. What possible reason could they have for that decision ?
I don`t suppose anyone has the time and inclination to fight that decision ? ! ? I`ll back them !
If I put a load of links to these, can we just confirm that the page addresses will stay the same, for years that is !
Justin, maybeyour site and/or Mike Brown's mb21.co.uk site could be granted the copyright and be willing to house these very useful radiation pictures ?
As an aside, I think it`s absolutely disgraceful that the radiation patterns for the main transmitters are missing. What possible reason could they have for that decision ?
I don`t suppose anyone has the time and inclination to fight that decision ? ! ? I`ll back them !
I lost interest in these once I realised that Sudbury was missing. However I was in correspondence with someone in Ofcom some months ago after submitting a FOI request. He gave me basic pattern details for Sudbury as words (PSBs omni on the old analogue antennas, COMs with a notch towards Calais/Cap Gris Nez from the Sudbury B tower). He did say, however, that Ofcom were intending to publish antenna patterns, and I got the impression it would be later this year. But that may be wishful thinking. Believe it when I see it is probably the operative phrase here.
If I've understood it all correctly http://kjs1982.me.uk/article/dvbt_nottingham.php should show the Nottingham transmitter, both pre and post DSO - the post-DSO similarities with kruador's imply i'm on the right lines at least.
Would have been easier if jpGraph could put 0' at the top of a polar chart...
The Derby diagram confirms what I've plotted on a map using the DUK postcode database. I live within 2 miles of the Derby transmitter, but do not get a signal. The DUK database actually indicates that if you are just 1/4 mile to the SW of the transmitter you don't get a signal (even though you can clearly see the mast), neither does Mickleover and that the suburb of Mackworth is cut into two as a result of the directional beam. West of the old college site near Radbourne Lane Water Tower is just outside the beam, East of the old college site is within the beam. The sharp cut-off to the SW is because it co-channels with Repton - its church spire can be clearly seen from the ridge that the Derby relay is sat on. Sadly Derby will co-channel with Nottingham's COM muxes from October.
It would be interesting to have a map superimposed on the diagrams (don't worry this is not a request for anyone to do so). Does any one know the scales involved if a map was superimposed on such diagrams, e.g I am assuming the round sides is the horizon, so the higher the mast the further away the horizon is?
Nottingham is also very interesting, because it shows the beam doesn't get out fully in a WSW direction toward my location, but I am getting a post-DSO signal on BBC-A and D3&4 on my loft aerial. The COM muxes are too weak to be reliable at the moment (in a transitional stage until later this year).
Justin, maybeyour site and/or Mike Brown's mb21.co.uk site could be granted the copyright and be willing to house these very useful radiation pictures ?
100% agree - isn't this what 'freedom of information' is all about ?
Lars
Hello Lars, not heard from you for a bit !
I`m not actually that IT savvy, so putting all those diagrams on my site would have to be done manually, I don`t even want to think about it....... I know that MB21 are thinking about something, though Mike Dimmick`s page does actually have them on anyway.
It`s a great pity the main TXs aren`t on, because that negates much of the usefulness of them, after all, most people are actually on main TXs. That said, the smaller the TX the more chance that its radiation pattern is not omnidirectional. In fact a contact at Ofcom told me that post DSO nearly all main TXs will have omnidirectional radiation patterns, but there`s a significant difference between all, and nearly all !
The Derby diagram confirms what I've plotted on a map using the DUK postcode database. I live within 2 miles of the Derby transmitter, but do not get a signal. The DUK database actually indicates that if you are just 1/4 mile to the SW of the transmitter you don't get a signal (even though you can clearly see the mast), neither does Mickleover and that the suburb of Mackworth is cut into two as a result of the directional beam. West of the old college site near Radbourne Lane Water Tower is just outside the beam, East of the old college site is within the beam. The sharp cut-off to the SW is because it co-channels with Repton - its church spire can be clearly seen from the ridge that the Derby relay is sat on. Sadly Derby will co-channel with Nottingham's COM muxes from October.
It would be interesting to have a map superimposed on the diagrams (don't worry this is not a request for anyone to do so). Does any one know the scales involved if a map was superimposed on such diagrams, e.g I am assuming the round sides is the horizon, so the higher the mast the further away the horizon is?
Nottingham is also very interesting, because it shows the beam doesn't get out fully in a WSW direction toward my location, but I am getting a post-DSO signal on BBC-A and D3&4 on my loft aerial. The COM muxes are too weak to be reliable at the moment (in a transitional stage until later this year).
The outer ring is simply maximum signal from the TX.
Inner rings are attenuated levels of signal.
In fact a contact at Ofcom told me that post DSO nearly all main TXs will have omnidirectional radiation patterns, but there`s a significant difference between all, and nearly all !
Here's what 'omnidirectional' means for one of the big panel arrays, at Rumster Forest (10kW):
This is eight tiers of four panels, using channels 21, 24, 30, 59 and 62 (for some reason, BBC A on C27 is missing from the data). Note that my software does not take account of the COM muxes being 3 dB down on the PSBs. The signal can be as much as 4 dB down at some angles compared to maximum (i.e. 40% of the power).
It's pretty much what I would have expected from looking at the RFS data sheet (PDF). Note that this is likely to be two four-tier arrays stacked on top of one another, rather than one eight-tier.
The variation by frequency is easier to see with the PHP-5S data sheet. Some sites (e.g. Hannington) are getting this as their new aerials, and some with only some of the faces populated (e.g. Dover only on south-west, north-west and north-east faces, the south-east and south faces not populated). Other sites, particularly the tubular masts, got PHP-12S arrays, while still other sites had a different manufacturer altogether (e.g. Crystal Palace). The reserve aerials are usually PHP-12S for 'omnidirectional' sites, because they can wrap around a lattice mast, though I think some might still be skew-fire.
Apparently Arqiva are having the arrays custom-built to their specifications, hence why it took such a long time to replace the burnt-out antenna at Oxford.
I suspect 'cogging' - where alternate rows of panels, or alternate halves, are at different angles - also has an effect on the radiation pattern of the overall antenna (otherwise, why do it?) Black Hill is cogged, for example.
I have now also uploaded the 25 pre-switchover patterns. Three of these are odd:
There were data files for 9 channels at Hemel Hempstead and 12 channels at Malvern. The extra channels were used before the equalization programme in 2000! While Hemel still uses three of the same channels, the newer plots have a comment "New 8 lambda antenna carrying all DTT" so I'm assuming the old plots for those channels were overwritten, and have only plotted the three that fell out of use.
The third oddity is that there is a pattern for Sheffield on C35. As far as I can tell, Sheffield has never broadcast anything on C35. I've uploaded it anyway.
I have not mapped channel to multiplex on these plots as this data set doesn't include it.
Sorry to re-open a very old thread, but would anyone be able to send me a link to, or the actual file of antenna patterns mentioned in this thread below? Unfortunately the UK Free link no longer works. I know some clever people have gone to great lengths to convert them to nice pictures, but for my application I want the numbers!
What that does show is that a number of Freeview Lite transmitters, e.g. Arfon, still have a full set of 8 channels allocated to them - or at least they did when the document was published.
Comments
The problem is that no maps show large trees or buildings, which are one of the most common cause of poor reception.
As I`ve said countless times, you`re better off working out your own TV signal strength, it`s likely to be more accurate......
Excellent, one down, four thousand to go.
Actually, is it neccessary to do all the MUXES for all the transmitters ? Isn`t there a good chance the radiation pattern for one MUX (at any particular transmitter) will be the same for the others ?
I don't think so, I've plotted a couple of sites with all 3 PSB muxes on and the data is identical, and 'Pop' has found the same.
Thanks for the polar plot info. 'Pop' - i'll try and digest that later today when I have time. i've got a Smith chart plotted in Excel somewhere, maybe I can use that as the basis of a template.
When you say "I don`t think so", is that :
A There aren`t 4000 more to go.
(or)
B The radiation patterns for all the MUXES at any particular transmitter are not all the same.
Whatever, if anyone can get a polar plot (like the one linked to) for all the transmitters I think that`d be very useful. I`d definitely put a link onto such a website from quite a few of the pages on my website !
I think he means that from a very limited number of files looked at, the plots at any one site are the same for all muxes. Whether this is true for all sites we can only find out by checking all the files.
It might be quicker for somebody to raise a new FOI request asking for the actual plots, rather than the raw data!
Pop
Pop
http://imageshack.us/photo/my-images/546/10102msf5420070830.jpg/
PS - Its HERTFORD
Very good that.
I doubt very much whether the repeater transmitters radiation patterns have changed, the small ones at any rate.
In any case Mr Butterworth says quite specifically which info is post DSO and which is pre DSO.
I really think this info could do with being publicised in a readily accessible form. A good example is the High Wycombe pattern shown earlier. If you lived in Downley you`d think you`d get a good signal off the High Wycombe TX, but it`s attenuated in your direction ! You can often make a good guess which direction the signal is transmitted in if you can see the anennas, but High Wycombe`s seem to be hidden within the shroud at the top.
Oooohh better than my effort, nice one.
- Most of the main transmitter sites are redacted (missing)
- Many relays in the south and east are also redacted
- Some plots are possibly dubious (King's Lynn really completely omnidirectional? Hmm)
- No vertical (tilt) data was used. For 3D patterns the peak horizontal pattern was used.
I've plotted all the muxes for a transmitter on the same chart, using as few lines as possible - that is, if the pattern for more than one mux was the same, they share a line. For most of the relays there is only one line.
In a few cases the plot never touches the outer edge. This is because the data for that mux never had a 0 dB point.
The images are organized by original parent transmitter (i.e. by first three digits of site ID). Some relays were reattributed in the past, or have been reattributed for DSO.
The pre-switchover data is in a different format and I've run out of enthusiasm for the moment. Again, most of the main transmitters are redacted so it might not be that useful anyway.
Nicely done!
Pop
Excellent work.
If I put a load of links to these, can we just confirm that the page addresses will stay the same, for years that is !
As an aside, I think it`s absolutely disgraceful that the radiation patterns for the main transmitters are missing. What possible reason could they have for that decision ?
I don`t suppose anyone has the time and inclination to fight that decision ? ! ? I`ll back them !
May I suggest to put the NGR, the UHF channel and the ERP of each mux on each picture.
http://stakeholders.ofcom.org.uk/broadcasting/guidance/tech-guidance/dsodetails/
Justin, maybe your site and/or Mike Brown's mb21.co.uk site could be granted the copyright and be willing to house these very useful radiation pictures ?
100% agree - isn't this what 'freedom of information' is all about ?
Lars
I lost interest in these once I realised that Sudbury was missing. However I was in correspondence with someone in Ofcom some months ago after submitting a FOI request. He gave me basic pattern details for Sudbury as words (PSBs omni on the old analogue antennas, COMs with a notch towards Calais/Cap Gris Nez from the Sudbury B tower). He did say, however, that Ofcom were intending to publish antenna patterns, and I got the impression it would be later this year. But that may be wishful thinking. Believe it when I see it is probably the operative phrase here.
Would have been easier if jpGraph could put 0' at the top of a polar chart...
The Derby diagram confirms what I've plotted on a map using the DUK postcode database. I live within 2 miles of the Derby transmitter, but do not get a signal. The DUK database actually indicates that if you are just 1/4 mile to the SW of the transmitter you don't get a signal (even though you can clearly see the mast), neither does Mickleover and that the suburb of Mackworth is cut into two as a result of the directional beam. West of the old college site near Radbourne Lane Water Tower is just outside the beam, East of the old college site is within the beam. The sharp cut-off to the SW is because it co-channels with Repton - its church spire can be clearly seen from the ridge that the Derby relay is sat on. Sadly Derby will co-channel with Nottingham's COM muxes from October.
It would be interesting to have a map superimposed on the diagrams (don't worry this is not a request for anyone to do so). Does any one know the scales involved if a map was superimposed on such diagrams, e.g I am assuming the round sides is the horizon, so the higher the mast the further away the horizon is?
Nottingham is also very interesting, because it shows the beam doesn't get out fully in a WSW direction toward my location, but I am getting a post-DSO signal on BBC-A and D3&4 on my loft aerial. The COM muxes are too weak to be reliable at the moment (in a transitional stage until later this year).
Hello Lars, not heard from you for a bit !
I`m not actually that IT savvy, so putting all those diagrams on my site would have to be done manually, I don`t even want to think about it....... I know that MB21 are thinking about something, though Mike Dimmick`s page does actually have them on anyway.
It`s a great pity the main TXs aren`t on, because that negates much of the usefulness of them, after all, most people are actually on main TXs. That said, the smaller the TX the more chance that its radiation pattern is not omnidirectional. In fact a contact at Ofcom told me that post DSO nearly all main TXs will have omnidirectional radiation patterns, but there`s a significant difference between all, and nearly all !
The outer ring is simply maximum signal from the TX.
Inner rings are attenuated levels of signal.
Often one can work out a radiation pattern for oneself by examining the potential co-channel issues on a table,
Here's what 'omnidirectional' means for one of the big panel arrays, at Rumster Forest (10kW):
[urlhttp://s1211.photobucket.com/albums/cc432/MikeDimmickTx/Radiation Patterns/STV North Grampian/Rumster Forest group/?action=view&current=14800_RUMSTER_FOREST.png[/url]
This is eight tiers of four panels, using channels 21, 24, 30, 59 and 62 (for some reason, BBC A on C27 is missing from the data). Note that my software does not take account of the COM muxes being 3 dB down on the PSBs. The signal can be as much as 4 dB down at some angles compared to maximum (i.e. 40% of the power).
It's pretty much what I would have expected from looking at the RFS data sheet (PDF). Note that this is likely to be two four-tier arrays stacked on top of one another, rather than one eight-tier.
The variation by frequency is easier to see with the PHP-5S data sheet. Some sites (e.g. Hannington) are getting this as their new aerials, and some with only some of the faces populated (e.g. Dover only on south-west, north-west and north-east faces, the south-east and south faces not populated). Other sites, particularly the tubular masts, got PHP-12S arrays, while still other sites had a different manufacturer altogether (e.g. Crystal Palace). The reserve aerials are usually PHP-12S for 'omnidirectional' sites, because they can wrap around a lattice mast, though I think some might still be skew-fire.
Apparently Arqiva are having the arrays custom-built to their specifications, hence why it took such a long time to replace the burnt-out antenna at Oxford.
I suspect 'cogging' - where alternate rows of panels, or alternate halves, are at different angles - also has an effect on the radiation pattern of the overall antenna (otherwise, why do it?) Black Hill is cogged, for example.
I have now also uploaded the 25 pre-switchover patterns. Three of these are odd:
There were data files for 9 channels at Hemel Hempstead and 12 channels at Malvern. The extra channels were used before the equalization programme in 2000! While Hemel still uses three of the same channels, the newer plots have a comment "New 8 lambda antenna carrying all DTT" so I'm assuming the old plots for those channels were overwritten, and have only plotted the three that fell out of use.
The third oddity is that there is a pattern for Sheffield on C35. As far as I can tell, Sheffield has never broadcast anything on C35. I've uploaded it anyway.
I have not mapped channel to multiplex on these plots as this data set doesn't include it.
Cheers.
http://www.ofcom.org.uk/static/broadcasting/Web_Int%20Coord%20800MHz%20DTT%20Clr%20Tech%20and%20Nominal%20Templates_22May13.pdf
TC