Coax as waveguide

No, what you get is what they have and use. There /are/ cables with fibre and power conductors, but these are not widely used for long cables as they are thicker and/or weaker.

Cheers

So this is the loss for the whole 10 km cable ?

The thermal noise density at room temperature is -174 dBm/Hz, so at

-14 dBm. Add to this the required SNR (depending on modulation method used).

At least due to voltage strength issues, the cable should easily be able to handle a 1 kW (+60 dBm) transmitter power, which would translate to 74 SNR, which should be enough even for 4096QAM so 1 Mbit/s would still be possible. A larger bandwidth and/or lower transmitter power would definitively be possible.

If you need to deliver the transmitter DC power from the opposite end and assuming 2 x 70 ohm loop resistance and 1500 Vdc feed, the short circuit current would be 10 A. At 5 A you could deliver 750 V (3750 W) down the line, which would be more than enough for generating 1000 W RF power.

With more realistic figures, say 200 kHz bandwidth, for which I would expect the loss be 155 dB, the reference level for noise is +34 dBm or

So if you are now getting 250 kbit/s, you are already quite close to the physical limits.

at

with

Well you can get very low losses at some frequencies and waveguides not normally referred to as such. How much end point equipment can you deal with?

?-)

I think it must be fairly close, as the company who does this are ahead of the rest, and there's money to be made. The transmitter power can't be very high, as packaging constraints mean there's no easy way of getting rid of the heat and the ambient may be 180'C.

Cheers