How does CBUS like unconnected segments.

[Charlie Crackle]

Running a job that will have some CBUS ITEMS connected at a later time.

This will leave some "Arms" on the network with nothing on the end.

2 options.

1) Run the "Arms" but not connect both ends till equipment goes on the end. (No desirable as 1 end is hard to access and easy to terminate now (ie end that connects to cbus network)

2) Connect one end to the cbus network and the other end unterminated but pairs insulated (to prevent shorts) (Desirable) and later connect cbus equipment.


What effect will these "arms" 20 meters of cable (5 of them) have on the network ?? will it make it less reliable ?? reflections ?? EMI

[allgo]

It wont make any differnece if you dont connect the last cbus items. Voltage will be present at the intended termination, therefore any volt drop can be checked. Current wont flow as there is no load is connected.
Similar to standard lights, with the switch off the light wont work. Switch on with the light bulb removed, voltage is present but as nothing is connected no power is used.

allgo

[rhamer]

I don't entirely agree with your comments Allgo.

It may well be true that in Charlies case there will be no problem, but we are not talking pure DC here.

C-Bus uses a balanced transmission line where the impedance of the cable is important.

Now I believe the C-Bus data rates are quite low, (probably intentially for this reason) and as such the cable characteristics will be less critical but there will be a point where hanging unterminated cable in parallel accross the network will kill it.

I don't know how much unterminated cable would be required to stop it working but I would mock it up just to be sure.

Then again, I could be wrong :)

Regards

Rohan

[Colin Smith]

Would a manual "load" help with this situation? (Resistor)
This could be removed when the input / output units are finally connected?

Just a thought.

Colin Smith

[ashleigh]

You should be OK, but the advice to mock it up is good.

If you can get the point at which the interminated segments join the main bus to be fairly easily accessible then you could just drop the cable in and connect it later.

[Don]

C-Bus was originally designed as a "Bus", that is, for any installation, providing that some basic rules were followed, the only high frequency effect of the cable that needed to be accounted for was total cable capacitance. The wiring was intended to follow as much as possible the same restrictions that sparkies are familiar with, such as "avoid short circuits", and "the signal can't navigate too well across an air gap", etc. I still think this was a good decision, as it allowed C-Bus to exploit the characteristics of a bus architecture with its protocol. The result was acceptable communication speed, good EMC charateristics, and low cost. The new Wireless C-Bus units can't exploit the same characteristics, and consequently have taken quite a while to develop, to solve many new problems.

The propagation time of signals down the bus has been taken into account, as have reflections, so it is possible to leave unterminated 'stubs' without affecting the overall network performance.

The only exception so far encountered in my experience with C-Bus (10 years!!), is that some unreliable communications were discovered in networks which were constructed as a 'ring' , with units connected at intervals. I am not sure if the porblems have been clearly associated with the ring configuration, but in any case, this topology is not encouraged any more. This might be worth the experiment.

Don