If the problem was the wires were picking up mains hum and other interference then using the drain wire as a ground wire intead of an ordinary wire will shield the wires from interference. Usually 0V is the ground in a circuit.

 

I said the 0 volt insulated wire because there has been so much discussion about bare wires, ground, shield, and drain. There is only one insulated wire that always has zero volts on it. That is the one to connect the resistor to.

 

haha wow this getting frustrating.
The wire I'm using as ground Is NOT insulated. It is a bare wire that is wrapped in foil (the shield I take it). Due to comments posted it seemed appropriate to use this instead of the insulated wire.
So can I still do the terminator (ha) on this bare wire? And if yes, this simply means throwing a 1.5K resistor across signal and ground? I CANT ruin this sensor, thats why i'm asking so many f**king questions

FYI this is the cable i b usin' :https://www.allelectronics.com/inde...92a403b2f4a37a59384e7ab28f02be1";i:1;s:0:"";}

is this an ok cable? seems so.... :I

 

That cable should work. And yes. the 1.5K from signal to ground.

 

Did you disconnect the insulated wire that was connected to zero volts and now you are just not using it?

 

yes. foolish?

 

Yes, foolish. The insulated wires are for clean signals. The bare wire and foil are for shielding the clean signals. One insulated wire and the foil shield both end up connected to ground, but they serve 2 purposes. External nasties trying to get in travel on the foil shield and bare wire to die at the ground connection. Intentional, clean, supplies and signals hide inside the shield.

 

son of Batch. ok, i'll have to re-do this.
Thanks for the help!!!!!
M

 

You have a very low current in the drain wire so i don't think using the insulated wire and the drain wire as a ground will make any difference.
The drain wire alone makes a good ground connection and a good shield.

 

Have you tried another cable? Is there a particularly nasty source of rf noise in the vicinity? Like motors, flourescent lighting, or digital displays? There must be something. Another source is house wiring that is close beside and parallel to your cable.

Consider this: Al is not the greatest conductor and when it is just a thin film it has a high resistance per foot compared to, say, braided shielding as with coax.

And is your power supply grounded to actual earth ground or is it not? If not then your ground is merely a common reference point at the supply. What would prevent rf from treating your 25 foot cable as an antenna? If this is the case ground your ground to the earth via a cold water pipe or a copper stake in the ground.

Is your computer grounded to earth ground? If not then this cable may be just an antenna gathering noise and feeding it to your computer's so-called ground.

Last thought: It would be interesting if you put the output of your sensor on an oscilloscope and determined the frequency of the noise. It may be white noise and hard to determine the source. On the other hand is it grass riding your dc? In this case it has a frequency and that frequency is the clue to the source.

 

Well, there goes Audioguru telling you to unhook the insulated wire and don't use it because running the signal ground on the same wire as the shield is a good thing to do when there is a known noise problem...

 

I think the noise is picked up by the unshielded "hot" wire. When it is shielded then the noise should stop. The shield is the drain wire and the tin-foil touched by it.

It should be simple to try the insulated wire and drain wire both connected to ground, then try just the drain wire connected to ground.

 

But if this so-called ground is nothing but an antenna? The environment corrupts the cable's so-called ground which in turn corrupts the computer's so-called ground?

 

just to clarify, when I say noise I mean this:
with short cable, the numbers in Max/MSP stay relatively stable. With 25' they are moving rapidly, never staying at -say - 0.24. If i change distance to around 0.34 it swings from around 0.31 - 0.36 etc. etc.
I did a kind of no-no thing via soldering, used solder to join a distance between a wire and terminal. Could that cause issues? Its not a cold solder tho, nice n shiny yes/

 

There is no "unshielded hot wire".

 

The analog output pin of the Sharp IR distance sensor is what I called the "unshielded hot wire". Datasheets from Sharp do not say the output resistance so I don't know the sensitivivity the analog output signal has to interference picked up by its output wire.

The output wire is an antenna that picks up mains hum and other interference.
If the output wire is inside a properly grounded shielded cable then the interference pickup is almost zero.

 

Fortunately, the "unshielded hot wire" is only long enough to get to the buffer op-amp. Hopefully, that is about an inch and the buffer amp is shielded by being in a metal container which is connected to the bare wire/foil conductor. After the buffer amp does its job, the signal wire, or, "shielded hot wire" will travel inside the shielded cable to the arduino inputs where it will be terminated with a 1500 ohm resistor. from the signal wire to the insulated wire with zero volts on it.

Meanwhile, the insulated wire carrying +5 volts needs a new name. How about, "shielded, insulated, 5 volt power supply wire"?

 

Why use a buffer opamp? My turntable that played vinyl records didn't. It just used one signal wire that was surrounded by the shield that connected to the circuit's ground. Its levels from its magnetic cartridge were much lower than this sensor and its shielded cable did not pick up any interference.

Don't forget that this thread started with the shield not being used.

 

A buffer amp lowers the impedance of the line, thus improving its noise immunity. As you said, data sheets do not say the output impedance, so that issue was resolved with a buffer amp, 21 posts ago.

 

The noise occured with the buffer opamp that he did not want to de-solder.
The old opamp is very noisy but I don't think its noise is the problem.