Hacker News Reader

Ok, great answer. I live in Norway btw.

Both the constructive interference and reduced number of transmitter arguments sound reasonable. I'm a bit more skeptical to the power cost, how much more expensive can that really be for FM? Electric power is pretty cheap. Does not really matter what the answer to this is though, given the other arguments.

At the moment I have to say I'm not very impressed with the coverage and quality of DAB in Norway, in my opinion it seems worse than FM. But I'm guessing this will only get better over time.

My main issue with the adoption is the requirement to buy new receivers, it's just so much waste. The adapters they sell for cars are in general just so ugly and a technically inelegant solution (local FM broadcasting, pretty ironic). A funny anecdote I read in the news was that some of these adapters come with a bluetooth handsfree solution, which also gets broadcast locally on FM. This means that other people driving close to you can listen in on your conversations if you use that functionality.

Hi,

Agreed - DAB coverage is seemingly much more spotty than what is claimed by Norkring (the operator) - makes you wonder what models (or antennas, in the case of field tests) they use to produce the maps.

Add to this the annoying property that whereas a FM signal degrades gracefully - as the signal gets poorer, you simply get more hiss, but can still make out what is being said, DAB basically drops off a cliff - it is fine until it - well, isn't, at which point it is pretty much unlistenable.

As for power, there are two factors both tipping in DAB's favour - first of all, you get good area coverage with lots of channels from one transmitter, seeing as the different stations are interleaved in one stream - say, a dozen channels for a realistic example. Having that same dozen channels on FM covering the same area, you'd need a dozen transmitters, all illuminating the same area.

Additionally, an FM tuner requires a signal which is significantly stronger if it is to decode it properly than what a DAB receiver does.

(This benefit is somewhat offset, power-bill wise, by the fact that DAB final amplifiers must be very, very linear in operation (because of the number of carriers) whereas an FM amplifier need not bother about linearity at all - and, hence, can be run at higher efficiency.)

(For a very rough comparison - the DAB transmitter on Tryvannshøgda providing local radio to Oslo boasts less than 5kW output. The Radio Norge FM site next to it? 88kW.

Thanks for the additional information, I completely agree with your description of coverage and the graceful degradation of FM.

DAB has obviously arrived and is here to stay in Norway, but I'm interested to see what the rest of the world does. My guess is that internet connected vehicles will be a both cheaper and better solution for consumers in many regions. Given the amount of data people use on their phones today, it seems likely to me that providing audio streaming of DAB equivalent bandwidth is feasible - even in Norway.

My complaint is that the usability sucks. Early adopters had to throw out their DAB receivers and buy DAB+ receivers. You regularly have to factory reset them because they keep moving the frequencies. Bad UIs for handling 20+ stations. The stations are just niche playlists, no content. I can just stream those ad free instead...

The car adapters are even worse.

> Additionally, with FM you need to make sure no transmitters on the same frequency can cover the same area, as that would lead to interference.

In the UK the BBC transmits their FM channels on the same range across the whole country [0]. In most cases the frequency between transmitters is only slightly different (e.g. 88.5 vs 88.1), so it can be picked up without having to retune the receiver. It requires a bit more thought up front, but it’s not that difficult (compared to other considerations you need to have for tranmitting radio), and it’s a much better UX than having to remember different frequencies for the same station in different cities.

[0] http://downloads.bbc.co.uk/reception/pdfs/FMradiotrans.pdf

I suspect (mind, suspect!) that approach wouldn't work here in Norway because of the (generally) different geography; in most of the UK, the main problem would be that you got too far away from the transmitter site for proper reception, in which case overlapping like you describe would work fine.

Over here, multipath interference is the main problem with FM - signals bouncing off any number of mountainsides before arriving at the receiver, causing a signal transmitted at time T to arrive at [T+t1], [T+t2]...[T+tn] at the receiver, with no way for the receiver to tell them apart as they are all on the same frequency. Garbled mess ensues.

Hence, rather than a few huge transmitter sites, we rely on thousands of smaller sites, the smallest of which used to operate at 200mW, IIRC.

The UK suffers from similar issues, we are not that flat. If you see a map of TX sites like [0] then you will see many small TV and Radio relays covering valleys.

[0]http://tx.mb21.co.uk/mapsys/google/alltx.php

> The reason, however (doubly so in Norway, which is basically as inconvenient as countries come for broadcasting purposes - lots of mountains and hardly any people) is that in DAB, you interleave a lot of channels in one data stream, transmitted from a single transmitter, whereas in FM, you need a dedicated transmitter for each channel.

Would it be possible to broadcast multiple FM channels with one magic sdr transmitter?

-It would be technically possible, but probably not make much financial sense.

Each FM channel would still require as much power as if it was operated separately to give adequate coverage; so rolling [n] channels into one SDR exciter, you'd need a power amplifier with [n] times the output.

And that is only the beginning of your probl... eh, challenges. This monster amplifier would need to let go of one of FM's greatest advantages - the ability to use high-efficiency, non-linear amplifiers for the final output stage. Now that you have multiple carriers, you would need to ensure the amplifier operated very linearly (and, in practice, at much lower efficiency) to keep the various carriers from interfering with each other.

Also, you'd probably be unable to use the same antenna for all channels, unless they were very closely spaced - or were all meant to cover the exact same area - and then you would need to apply extensive, expensive filtering to separate the carriers before feeding them to different antennas. (Which would be much simpler and cheaper to do just after the SDR, since the signal there would be very low power - but then you could use separate power amplifiers and make it even easier on yourself. (At which point you might as well use separate exciters, too - and be back where we are today. :))