Well here's a good read:
http://en.wikipedia.org/wiki/Power_line_communication

The problems are:
1.  Power loads should not absorb the signal, because the transmitter 
can't put out the signal with the 20 amps it takes to run all the 
appliances.  This is tricky because you can't tell every TV set and 
blender what you want it to do.  Filters are good but you'd have to 
place one on each load.  X10 apparently found a niche where it's not so bad.
2.  Power loads should not create their own noise in the required band.  
Brushed motors create a lot of broadband noise.  Again, filters, but 
it's difficult to filter every load.  Error tolerance would be nice.
3.  Power lines are not designed to transmit high freq.  They have no 
shield or controlled impedance, leading to reflections, standing waves, 
lots of radiation out and external noise coming in.
4.  Limiting the scope of the signal.  All these power lines are 
connected.  You don't want your network signals to broadcast to all the 
houses running off the same pole transformer.  X10 has separate 
addresses to try to avoid that.  But that won't resolve two transmitters 
broadcasting at the same time which could frequently be the case with 
voice or ethernet or whatever.

Danny

Peter wrote:
> Xiaofan Chen <xiaofanc <at> gmail.com> writes:
>   
>> I do not know much about X-10. But I think ASi bus can be
>>     
>
> X10 is a powerline protocol meant for home automation. It is not particularly
> brilliant. There are other powerline signalling protocols that are much better
> and have good support from third parties (like Philips carrier current modem
> chips and so on). Carrier current and wireless have the same problem: noise
> immunity is insufficient for anything but reasonably quiet office environment or
> home use. FO seems to be the best solution currently for industrial and code
> compilant outdoor data wiring (the FO has no dc conductive path so a lot of
> sloppiness in grounding can be tolerated and crosstalk is inexistent between
> parallel cables). FO wiring cost is below $1/meter for 4 core outdoor FO and is
> hard to beat, and I can see it becoming common even in indoor wiring where rack
> to rack connection is already FO in some cases (PBX, measuring interfaces to PC
> etc).
>
> Peter
>
>
>
>   

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist