7 April 1994
Norman McLeod
When I went to the Radio Academy's debate on 105 - 108 MHz I burbled on about lattices, enough to cause Mark Thomas to interject, but probably not enough to get across the point I was trying to make to those not technically inclined...
I should have picked something simpler to talk about, like Brian West's attitude to pirate radio. Lattices need a bit of explaining - certainly more than you get from the Radio Authority's exposition in the consultative document.
They're also very definitely a subject where a picture is worth at least a thousand words. I have serious production difficulties with pictures for inclusion in this magazine. I would do more technical talk and less anoraking if only I could draw diagrams easily, but on a A5 page and lacking suitable software it's so much easier to tackle subjects adequate to text only. This week is the exception. I hope that the opposite page will help.
It's reproduced from a paper I wrote in Wireless World, July 1983, called Planning for Plenty, which in turn arose out of work I had done in 1982 for the GLC on new radio services for London. It shows a minature lattice which could have been overlaid on London to produce around 160 very small neighbourhood stations from seven unused channels. (Unused, that is, apart from the pirates.)
There are four key features of a planning lattice:
(a) the number of frequencies available;
(b) the service radius of each station, and by inference the power;
(c) the separation between stations on the same frequency, and
(d) the separation between adjacent stations (on different frequencies).
In the case of the London lattice, the parameters were:
(a) 7 frequencies;
(b) 2 km service radius, around 2 - 10W erp;
(c) 10 km co-channel spacing;
(d) 3.6 km between one station and the next.
Each little station would have been not unlike some smaller RSL's in its scope, and given the population density in London most latticees would have had potential audiences between 50,000 and 100,000 people, which is just about what is currently believed to be the lower limit of viability for a permanent service.
I chose seven channels because with this number in a lattice you begin to get slight overlaps between the service areas of physically adjacent stations. If you imagine a one-channel lattice, consisting, say, only of the channel 4 stations, you can see that only a small percentage of the land area is covered with service. As you increase the number of channels and fit them into the lattice, you reach a stage where the service areas meet each other, and 100% of the land area has service on at least one frequency.
Happily one could fit seven channels neatly across 105-108 with comfortable 400 kHz spacing.
A possible set of channels, together with the IF offset frequencies which must be avoided, might look like the table below.
Channel frequency - IF offset
Ch 1: 105.2 MHz - 94.5 MHz
Ch 2: 105.6 MHz - 94.9 MHz
Ch 3: 106.0 MHz - 95.3 MHz
Ch 4: 106.4 MHz - 95.7 MHz
Ch 5: 106.8 MHz - 96.1 MHz
Ch 6: 107.2 MHz - 96.5 MHz
Ch 7: 107.6 MHz - 96.9 MHz
What is this 'IF offset' about? If Radio 4 is running on 94.5 MHz, then almost every FM receiver will have an oscillator running 10.7 MHz above that frequency (105.2 MHz) when tuned to R4. This is to convert Radio 4 to the IF, or intermediate frequency, which is always 10.7 MHz. Mixing two frequencies together produces the sum and the difference between them.
To cut a fairly complicated story short, there is a risk that if two frequencies 10.7 MHz apart are radiated in the same area, and particularly from the same mast, then interference will be caused by each to the other and quite possibly other services as well. There is no REAL interaction between them - it's all done by the receiver, but it's a fact of life and something which has to be recognised.
So we may have to offset some of our channels in areas where the lower frequencies are in use. This is not a factor peculiar to lattice structures - it affects ANY planning in the 105 - 108 MHz band. Below is the London lattice as illustrated in Wireless World. More next week...
Any readers wishing to comment on what I have said are welcome to call me on 01 273 684 172 or send e-mail to normac@fastnet.co.uk
Copyright NJ McLeod 1995