"Kristoff Bonne" wrote in message
...
Gegroet,

Just interested. Does anybody know if the AM-decoding in a "normal price"
SW-receiver is done in hardware or in software (ASIC/DSP)?

I'm not aware of any radios at any price that do AM decoding by DSP. I
suspect they exist in some form - after all, that is what one would expect
from Software Defined Radio technology. What about these new DRM radios - do
they also do AM and FM demodulation via DSP?

I might have missed something but AFAIK you cannot decode SSBc with a
standard AM-radio, can you?

Yes. I might have the acronym wrong - I mean SSB with reduced carrier, not
suppressed. The Canadian time signal transmissions from CHU on 3330 and 7335
kHz use this mode - easily received on the cheapest SW radios.

DRM has two advantages of this:
- it's a digital system. This means it includes features like auxilairy
content, "Alternative frequency" information, station identification,
multiple streams (e.g. two audio-streams in one DRM transport-stream),
variable bandwidth from 4.5 to 20 Khz, variable modulation-sceme and
error-correction mechanism based on channel quality, easier support for
time- and frequency diversity, single-frequency networking.

I'm aware of this but it's incompatible with the huge installed base......
I'm not sure about your "easier support for" statement - it's true for time
diversity but I can't quite imagine how and why one would need this for an
analog transmission mode so it seems a specious point. Frequency diversity
and SFN is done with AM. One's sense of hearing does a wonderful job of
"error correction" as part of an analog system - the digital system can't
work without its own. Seems like many of the purported advantages are born
out of necessity.

Auxiliary content, AF, station ID's, multiple streams and various bandwidths
and bitrates in DAB surely did not excite the public in this country. Now,
satellite radio with terrestrial fill-ins may be a different matter, and
cable and satellite DTH audio services with images and text certainly are.
The takeup has been pretty good.

It's easier to integrate into combined DAB/DRM chipsets which allows for
an one "integrated" chipset for both systems. It also allows
cross-platform services; e.g. does the "AF"-service work between DAB, DRM
and FM/RDS.

This may be of interest in those markets that Eureka 147 DAB was adopted
with fair market success (where is that other than the UK?) and maybe where
FM/RDS is likewise deployed but both require that the broadcaster has
multiple transmitters carrying the same program in multiple modes for the
listener to realize the benefit of cross-platform AF service. This may be
more theoretical than practical. The time delay for lockup to DRM or DAB is
so large that AF service using either of them would require dual receivers
so that the alternate frequency is already tuned and locked before the
current
frequency goes unlocked, rather than blindly chasing the AF list.



- it's one technology for LW/MW, for SW, for 11 meter "local" broadcasting
and, with the extension of DRM+, for band I and band II.

So could SSBc and maybe NBFM


This creates an enormous market for this, and -therefor- cheaper chipsets.

Likewise for any mode common to all bands/markets...


- And it allows you to lower the transmission-power (and hence the
electrivity-bill).

My observations of DRM on shortwave has been that high transmitter power is
still needed - maybe less than for DSBAM. SSBc also saves power.

Interest idea. Just interested to know how 10 Khz SSBc would react to
selective fading. As you said in the beginning, the problem is when the
carrier is gone. Wouldn't you have the same problem as with a
DSBAM-receiver?

When selective fade knocks out the carrier, you lose lock but not the
modulation, with a "true" synchronous AM detector. There will be an error
distortion, as in asynchronous ECSS, which could be very small for some
considerable time. DRM is not immune to selective fading - the cause is
multipath which causes jitter or spreading of the digital stream and when it
is big enough - no decoding and then wait for re-lock after the jitter has
fallen below the acceptable threshold.

BTW. AFAIK (correct me if I am wrong) a DSBAM-decoder will completely
chocke on such a signal, so I kind-of miss the reason why you say this is
compatible with a convensional DSBAM-radio.

An envelope detector will demodulate as long as there is carrier. There is
increasing distortion as the carrier is reduced relative to sideband below
some level. There would have to be a tradeoff between power savings and
receiver distortion.

Well, how would such a system react in a situation of two half co-channel
stations. (i.e. one station at 9.900 - 9.910 Khz, and another channel at
9.905 - 9.915 Khz).

(What is the exact jargon-word for this? co-channel? semi co-channel?)

With DSP IF brickwall filtering as in the Dream and SDRadio softwares, I
think it should work very well indeed by setting the filter to less than
5kHz - for the lower of the pair if we assume USB mode for both. The higher
would experience crosstalk from the first for program content whose audio
spectrum extends well above 5 kHz. Conventional AM radios could also be
tuned to minimise interference from the adjacent channel.

How does DRM react to overlapping spectrum from an adjacent or co-DRM? Not
likely very well given its poor performance in the presence of DSBAM or SSB
overlapping its spectrum. DRM SFN's may be another matter but jitter due to
path differences will cause problems somewhere - maybe synchronous SSBc
transmitters could form equally effective SFN's that work with the huge
installed base....

Cheers, Kristoffe. You lead a good discussion.

Tom