Does wire length effect performance?

Sparkfun recently created this amazing musical staircase with an Arduino. http://www.youtube.com/watch?v=-h2intYYghk

In it they use some really long wires, I'm guessing at least 20 -25 feet long. These wires go to photoresistors inside the railing. My question is, at the low voltages used with an Arduino, how long of a wire can you use without effecting performance?

These wires go to photoresistors inside the railing. My question is, at the low voltages used with an Arduino, how long of a wire can you use without effecting performance?

I can think of a few variables:
How long the wires are
How thick the wires are
How much interference might there be near these wires.
How much current the supply is and how many sensors you are powering off this 5v line.

How long are you thinking of and then suggestions can be made. There is not really a definitive answer to your question. :wink:

Mowcius

My question is, at the low voltages used with an Arduino, how long of a wire can you use without effecting performance?

One place where you can start is calculate the resistance of the wire length you intend to use and see what the resistance and voltage drop will be.

It is not only the physical parameters of the connection like resistance but also the capacitance and susceptibility to interference.

Basically you question has no definitive answer as it depends on lots of things you don't know and can't really find out. Not least of which is your criteria of:-

without effecting performance

That is a whole other can of worms.

You'll also need to avoid kinks / tight bends else you will get reflection, and bear in mind that long unshielded wire can act as an antenna too.

You'll also need to avoid kinks / tight bends else you will get reflection,

No not at these sorts of frequencies you won't. That only applies to step pulses not analogue levels.

True, but the OP didn’t say what he was intending to use the lengths for?

Just for sake of argument take the Sparkfun example and say they used 22 gauge wire hooking up 24 photoresistors. The longest wire is say 35 feet long and then the photoresisors are at 1 foot intervals are place at 1 foor intervals.

22 gauge wire

Copper wire? Solid strand? Pure?

Just for sake of argument take the Sparkfun example and say they used 22 gauge wire hooking up 24 photoresistors. The longest wire is say 35 feet long and then the photoresisors are at 1 foot intervals are place at 1 foor intervals.

So what are you asking?

Read my lips:-

Basically your question has no definitive answer

You can't add enough detail and even if you could it would be subject to changing with temperature, levels of interference the mass of people standing close by and a whole load of other things.

Making the circuit as shown in the video will not work reliably in the long term no matter how nice the video showed it working. It will be subject to false triggering, not triggering and changing thresholds as the ambient light levels change. You are already over the limit for reliability.

Sites like below can provide info on wire resistances. Standard telephone cat3 wire is inexpensive and would probably work. Don't be put off by some of the responses of why your project won't work, as they make somewhat stupid assumptions to validate their argument. ;)

http://www.cirris.com/testing/resistance/wire.html

as they make somewhat stupid assumptions to validate their argument

Given some of the posts you have been writing today, ;D ;D ;D ;D ;D

You can’t add enough detail and even if you could it would be subject to changing with temperature, levels of interference the mass of people standing close by and a whole load of other things.

Making the circuit as shown in the video will not work reliably in the long term no matter how nice the video showed it working. It will be subject to false triggering, not triggering and changing thresholds as the ambient light levels change. You are already over the limit for reliability.

  1. please define your concept of “reliably”.
  2. what is “long term”? Five seconds, 5 days, 5 years, 5,000 years?

“You can’t add enough detail…” the pot calling the kettle black! :wink:

Quote: as they make somewhat stupid assumptions to validate their argument

Given some of the posts you have been writing today,

Oooo..., Grumpy barking like he might be the bit dog!

Seems to be a strange day here on the old forum. :wink:

This topic reminds me of an old quote from Abraham Lincoln:

He was asked how long should a mans legs be. He responded “long enough to reach the ground”

Lefty

Well, it is not only about wire lose (called I squared R loss) it is also about delay. A significantly long line will have a specific delay dependent on several things you can get from nano seconds to micro seconds really easy and now I am guessing but I suspect they were looking for this delay.. because as we all know there is a lot easier way to get resistance should that be the desired effect.

DILLON

what is "long term"? Five seconds, 5 days, 5 years, 5,000 years?

There is a daily cycle you have to worry about. Then a monthly cycle. If you get over that then it is probably OK for about 6 months.

The problem is as I have said before that there are so many variables you can't answer this question. No one can. There are too many variables and it is impossible to see how they all interact.

In engineering terms putting 20' of wire in front of a high impedance analogue input is too much. It will not work reliable. That means for every person in every location, for all weather conditions for ever. If you were trying to see this as a piece of professional kit then it wouldn't meet any standard you threw at at. Yes you can do it once for a video as a demonstration but what happens is that you will try it and then come back on the board saying that it sometimes resets your arduino or that it sometimes fails to trigger or a host of other intermittent stuff.

I have used photo transistors at the end of similar length wires reliably, but at several have said, there are a lot of variables that effect the equation. Suck it and see.

I have a 100k current limiting resistor on my phototransistor, the resistance of the wire (maybe 5 ohms) isn't a factor worth considering.

Its capacitance and its effectiveness as an an antenna in the environment it runs through is a different matter.......