Im building a multiple temp sensor using DS18B20 all on the one wire. Do I need just one pull up resistor for all sensors or one per each sensor?
One 4.7k pullup ought to work. However how long is your OneWire bus?
thanks i had planned on using one, just wanted to be certain. Total length of wiring would be no more than 1.5 meters for all 13 sensors.
Should be OK - I've done 4m run of 20 sensors, though I used 10k at each end I think.
Don't forget decoupling caps for each sensor (I presume you are not using parasite-power mode).
Thanks for the advice on the de-coupling caps, I don't know what you mean buy parasite power supply?
Craig_Minca:
Thanks for the advice on the de-coupling caps, I don't know what you mean buy parasite power supply?
Parasite-power means that you are powering the sensors over the data line (2-wire mode)
If you are using 3 wires, you are not using parasite-power-mode.
I guess you didn't study the datasheet for the sensor very much - then you would have stumbled upon the "Parasite-power" statement 
// Per.
Awesome, great stuff, thanks. Yes.......you are right I didnt read the datasheet. 
I have a 1-wire run that's about 100 ft long with 10 sensors on it. It runs outside to a weather collection installation. I use a power injector to feed 12v and 5V into the CAT5 cable as well as use 2.7K pullup. I initially used 4.7K with this setup but had pretty regular garbled conversations. The higher strength pull-up solved the issue, though I would be reluctant to go much lower than 2.7K.
hey guys, just stumbled on this thread and was hoping someone could elaborate on the decoupling caps for each sensor... my project will have 2 sensors on maybe 1.5 meter run and the plan was to wire them together in parasitic mode.
Very new to all of this, very limited knowledge and vocabulary... sorry about that. My understanding is that I can use a single pull up resistor for both sensors, but I'm not sure about the caps? and is 0.1 uf appropriate? I have read the data sheet, but as I said, new at this and not all concepts gel immediately
I have attached a pic of what I thought the wiring diagram would look like, with grounds and vdds connected together...
If anyone could tell me if I've misunderstood how these should go together, I would very much appreciate it!
Thanks,
Joe
In general every logic chip should be decoupled. 0.1uF is a good compromise value if you don't
know any better.
In parasite power mode you cannot decouple at all, which is one of the reasons its less reliable
and more prone to interference (but uses only two wires, not three).
Thanks MarkT, I think I got it now
MarkT:
Should be OK - I've done 4m run of 20 sensors, though I used 10k at each end I think.
Don't forget decoupling caps for each sensor (I presume you are not using parasite-power mode).
A few years late on this one.
Good recommendation MarkT. I assume this means the capacitor is connected relatively close to the sensor on the end of the cable.
For those particular ds18b20 sensors (water proof ones with the metal tube) - they're going to be a bit of a problem maybe, right? I like the idea of decoupling caps though.
Southpark:
For those particular ds18b20 sensors (water proof ones with the metal tube) - they're going to be a bit of a problem maybe, right? I like the idea of decoupling caps though.
The documented problem with the ones with the metal tube is that they are frequently not waterproof!
You need to cover the interface between the open end of the metal sleeve and the cable with a collar of self-saucing heatshrink.
Paul__B:
The documented problem with the ones with the metal tube is that they are frequently not waterproof!You need to cover the interface between the open end of the metal sleeve and the cable with a collar of self-saucing heatshrink.
Oh geez. Thanks for letting me know about that Paul__B! I have some of these but had never yet used them. Disappointing effort from those people behind that particular assembly.
Thanks again Paul__B!