RE: Re: X10 - how ?

["Kilgore, Andrew" <amkilgore@xxxxxxx>]

> However in some
> cases the addressing may not be needed as a housecode remains
'current' so
> you can send B6 ON and then 7 ON and the second transmission will turn
B7
ON
> and only requires an additional 560ms.

As I understand it, the 2 commands are of the form:
- B 6	(addresses B6)
- B On 	(sends an on command to all units on housecode B that are
"listening")

Therefore, if you later want to turn B7 on you would have to send:
- B 7
- B On

The shortcut comes from being able to address multiple units on the same
housecode before sending the command:

- B 6
- B 7
- B On

Does this sound right...?

Thanks,
Andy.

-----Original Message-----
From: Kevin Hawkins [mailto:lists@xxxxxxx]
Sent: 04 December 2003 13:46
To: ukha_d@xxxxxxx
Subject: RE: [ukha_d] Re: X10 - how ?



> -----Original Message-----
> From: Alistair Watkins
> Sent: 04 December 2003 12:03

> I think x10 on the powerline travels at 300bps, pretty
> slow even with the tiny amount of data sent.. this is why
> it takes about .5 sec to turn on an x10 device after you
> send a command over the powerlines. AFAIR.


It even worse than this as I understand it...

The speed of X10 is dictated by the mains frequency - so in fact US 60Hz
X10
transmissions are slightly faster. This is because the 'bits' that form the
data are transmitted at the zero voltage crossing point of the ac mains.
Each X10 transmission is 11 cycles long and normally a gap of 3 cycles has
to be inserted between transmissions.  Each cycle carries 1 bit of
information so the baud rate is 50 baud. The positive and negative halves
of
the cycle carry complementary information for verification purposes.
Additionally each transmission has to be sent twice (for error checking) so
a single transmission occupies (11+3)*2 = 28 cycles  and this is 560ms  -
just over half a sec - and this assumes a perfect (no
interference/collisions) world. So the 'effective' data throughput is now
just under 20 baud :-(   C-Bus for example is about 250 times faster
although this ignores the protocol overhead which probably reduces this by
10% or so.

Simplistically there are two types of X10 transmission - those that define
the address and those that define the action. So to turn B6 on you send an
address command followed by an ON command. Thus to turn B6 on can take 2 x
560ms = 1.12 seconds in just the X10 transmission alone.   However in some
cases the addressing may not be needed as a housecode remains 'current' so
you can send B6 ON and then 7 ON and the second transmission will turn B7
ON
and only requires an additional 560ms.

If you have a lot of devices on different housecodes , for example some
people allocate a housecode to a room,  then the latency if you succesively
switched on lights in different rooms would be greater than if they were
all
on the same housecode, assuming an intelligent transmission of data.  You
may be able to use an 'all lights on' to a housecode though to better
effect.

So AIUI a minimum of .56 seconds to turn something X10 ONOFF and often
twice this at 1.12 seconds.


Think this is right...  Correct me otherwise (Li ??)

Kevin




UKHA 2004: 15th and 16th May 2004

http://www.automatedhome.co.uk
Post message: ukha_d@xxxxxxx
Subscribe:  ukha_d-subscribe@xxxxxxx
Unsubscribe:  ukha_d-unsubscribe@xxxxxxx
List owner:  ukha_d-owner@xxxxxxx

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/