Maximum distance of wired sensor

[lepass7]

Hello guys, i would like to know which is the maximum distance of the wire which transfers the signal of a sensor without any buffers. For example i have an lm35 (temperature sensor) how far can i have my sensor from my pc? (i dont want to use rf or any other wireless device).
Does anyone tried this, or should i try this on my own?
Thank you in advance.

P.S: I believe the answer varies and depend of the sensor, of the power supply, and of the sensor reader.

 

[davenn]

it would also depend on the cable used the output voltage range from the chip is relatively small and a long cable is going to pick up a lot of noise
a twisted pair inside a screen would be a good idea. The datasheet does give some info on feeding twisted pair cables. But no info on max range.
Have a play and see what you find out and report back

cheers
Dave

 

[Boncuk]

Hi lepass7,

I have used temperature sensors connected via a three wire cable (actually four twisted wires of 0.8mm diameter copper wire with shielding, e.g. telephone cable (similar to CAT5 cable)) more than 250m in industrial installations without losses or noise on the incoming signal to a microcontoller's A/D converter.

The current requirement of the A/D converter is decisive for accuracy.

If you want to make sure that there is absolutely no loss you should use a current output (0..20mA or 4..20mA) and use a burden resistor on the receiver side to gain the desired voltage level.

A current output may have unlimited distance between transducer and receiver if the maximum current of 20mA can be maintained.

Boncuk

 

[alec_t]

A sensor with a frequency output would also be usable over large distances with little error.

 

[ericgibbs]

hi,
The OP is clearly asking for guidance with a LM35, not frequency, not 20mA current loops.???

What is the distance/length that you want the cable to be.???

Use the remote connection method as shown in the datasheet.

 

[lepass_7]

Thank you all for your answers.
@davenn: i agree i certainly must use a twisted pair
@Boncuk: so its possible thank you!
@alec_t: i guess i could do that if Boncuk didnt try it
@ericgibbs: yes i ll use that. But the datasheet talk about remote but not the maximum (as davenn said)
So, thank you and i will let you know the results...

 

[Mr RB]

The noise won't matter if you filter with a large cap at the receiving end. That's what the cheap thermometer products do when they have a 10 metre lead for an outside sensor. And they avergae the temp over about a 10 second period, which also removes the noise.

Because the LM35 has power supplied the sensor output doesn't have much current at all, so wiring resistance is not that critical.

 

[lepass_7]

hello again,
I tried this with a 10 meter CAT5 and i think is working, the wiring is shown below:

 

[Mr RB]

R? should go at the out end of the cable, not the in end. The RC can be larger than that, try 33k into 47uF (pref a tantalum).

 

[ericgibbs]

R? should go at the out end of the cable, not the in end. The RC can be larger than that, try 33k into 47uF (pref a tantalum).

hi Roman.
The d/s recommends that the R1 be placed at the LM35 end in order to 'isolate' the LM35 from the cable cap loading

 

[Mr RB]

Fair enough. Then a resistor at each end is the way to go, as an RC is needed after the cable to best filter cable noise.