2
\$\begingroup\$

I have a 1-wire network using cat5 cable. There is a bus running along the roofspace of the building and the sensors are on stubs off this bus. The bus along the roofspace is about 25 metres long and the stubs are between 1 and 5 metres long.

I understand that a linear bus is better than one having stubs or a star configuration. It occurs to me that without having to replace any cabling I could rewire each sensor to be directly on the bus. By using one of the twisted pairs in the Cat5 cable to run the DQ and Vdd down to the sensor and then use another pair (in the same cable) to run the DQ and Vdd back up again the bus could continue without having made a spur.

enter image description here

Would running the out and back wires inside the same sheath be OK? Would it improve the network or would it cause interference, make the total cable length even longer and be worse than just having spurs on the bus?

UPDATE:

The bus is driven using a Maxim DS2482 at standard 1-wire speed. The devices are not using parasitic power.

I am about to add some more sensors onto the bus. I am wondering if there would be any benefit in changing the way it is wired since I will soon have an opportunity to do so.

I would not be able to replace the existing cables, hence the idea of running two legs of the bus inside the same physical sheath to create a linear bus. Does this sound like a good idea worth pursuing further or is it pointless?

\$\endgroup\$
  • \$\begingroup\$ what speed are the signals running at? what are the edge rates? if the signals are slow enough you can have all the stubs you want with no ill effects! read Howard Johnson's book... \$\endgroup\$ – hwengmgr Mar 16 '15 at 10:38
0
\$\begingroup\$

I think that capacitance is going to be the enemy in your setup. Accordingly, I think that the option that you are mentioning instead of using stubs will cause more problems.

The best way to figure this out is to either measure the capacitance of your existing cable or look at the datasheet for the cable. Then calculate how much the capacitance changes when you re-wire to eliminate the stubs. You can then use that capacitance value along with your pull-up resistor to see if you are going to run into problems.

\$\endgroup\$
  • \$\begingroup\$ So, basically what you are saying is the extra capacitance from the longer cable runs is likely to outweigh the advantage gained by losing the reflections from the multiple end points? \$\endgroup\$ – Paul D Mar 16 '15 at 20:43
  • \$\begingroup\$ Yes, that's what I think. But the best way to find out is to actually measure the capacitance (or calculate it) and then figure out what your rise time is going to be. \$\endgroup\$ – Dwayne Reid Mar 18 '15 at 5:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.