Dirt cheap way to detect change in resistance over tiny changes in wire length?

Captain Bob

Joined Nov 9, 2022
7
As someone who has no idea of what I'm getting myself into, I am relying on the expertise of the members of this forum to stop me before I waste a lot of time on something that is impossible to accomplish. Here is what I am trying to accomplish:

Two parallel wires, no more than eighteen inches long, connected to an ohmmeter. Circuit is completed by a conductive something (let's call it the Train) that rides on the two wires like a train on tracks. Is there any way the ohmmeter could detect the change in resistance if the Train moves as little as 2mm? And just to make this harder, the system must get its power from common batteries (AA, AAA, or 9-Volt, etc).

What I'm primarily looking for is a way to detect movement of the Train, and I thought resistance might be the easiest way.

Any thoughts would be greatly appreciated, even if it's just to let me know this is impossible.

crutschow

Joined Mar 14, 2008
32,041
What's the gauge of the wire, and is it copper?

Captain Bob

Joined Nov 9, 2022
7
It can be any material that makes this work, but I'm really hoping for aluminum. Gauge can be anything that makes this work.

Tonyr1084

Joined Sep 24, 2015
7,356
I would imagine that detecting the slightest change in resistance would be quite difficult. But if your "Train" was a capacitor of some value (I have no idea what), changing the length of the leads to the cap would change a frequency which would be easier to detect.

Years ago I saw a home made radar detector that the actual detector itself was a capacitor with specific lead lengths. Shorter lengths would read a higher frequency, longer leads - a lower frequency.

crutschow

Joined Mar 14, 2008
32,041
It can be any material that makes this work, but I'm really hoping for aluminum.
Al has a relatively low resistance (slightly above copper).

It's easier to detect a small resistance change if the wire is as small a gauge as feasible, and the wire is high resistance such as steel, or (best case) nichrome.

Ian0

Joined Aug 7, 2020
7,442
2mm in 18" is 0.4%.
The temperature coefficient of metals is generally 0.4%/°C. So you wouldn't be able to tell if the train had moved 2mm or the tracks had got 1°C warmer.
So nichrome (or constantan, manganin etc.) would be a good idea. They also have a lot higher resistivity than copper or aluminium so that the current needed to produce a measurable change in voltage would be less.
No good, though, if the tracks supply power to the train.
Aluminium immediately oxidises on exposure to air to produce a thin film of alumina which is an insulator. I suspect that contact resistance would be a problem, unless you could arrange it as a potentiometer like the 1 metre bridges we used to use in physics classes.

BobTPH

Joined Jun 5, 2013
6,516
If the goal is to determine the position of the “train” there are better ways to do it.

crutschow

Joined Mar 14, 2008
32,041
2mm in 18" is 0.4%.
Actually two wires for a total of 36" for a 0.8% change.

Ian0

Joined Aug 7, 2020
7,442
Actually two wires for a total of 36" for a 0.8% change.
But if the train goes 2mm down the track, the length of track in circuit increases by 4mm.

Captain Bob

Joined Nov 9, 2022
7
If the goal is to determine the position of the “train” there are better ways to do it.
The goal is more to determine "is the train moving right now".

Ian0

Joined Aug 7, 2020
7,442
The goal is more to determine "is the train moving right now".
Assuming it has a DC motor, then look for the noise superimposed on the supply by the commutator.

WBahn

Joined Mar 31, 2012
28,474
You should consider a number of potential alternatives, such as an ultrasound module.

Ya’akov

Joined Jan 27, 2019
7,442
Welcome to AAC.

This is one of those cases that is going to go around in speculative circles unless you provide more information about what you are actually trying to do. You've explained, what you have labeled a naïve solution, but not what the solution is meant to solve.

Please explain the context so you can get real help.

MrChips

Joined Oct 2, 2009
28,485
How to detect if an object has moved 2mm?

There are a number of ways to do this via resistance, capacitance, inductance, optical, lasers, linear encoders, phase shift.

We need to know specifics such as the actual object, track, total range of movement, etc.

BobTPH

Joined Jun 5, 2013
6,516
An encoder somewhere in the drive train would be my first thought. (if that is even applicable)

Tonyr1084

Joined Sep 24, 2015
7,356
An encoder somewhere in the drive train would be my first thought. (if that is even applicable)
Two parallel wires, no more than eighteen inches long, connected to an ohmmeter. Circuit is completed by a conductive something (let's call it the Train) that rides on the two wires like a train on tracks. Is there any way the ohmmeter could detect the change in resistance if the Train moves as little as 2mm?
The TS has used a train as an example. I doubt there's a drive mechanism, at least not from where my imaginings have gone. I agree with Ya’akov we need to know more about what you're trying to sense as opposed to imagining a train on a couple of 18 inch long wires. Otherwise you're going to get suggestions such as motion detectors, DME (Distance Measuring Equipment) and a whole host of other non-related solutions to your project. If your project is "Top Secret" and you're not willing to help us help you then you're never going to find a solution. Suggestions as potentially useless as this.

Ian0

Joined Aug 7, 2020
7,442
Turn off the power supply briefly (using PWM for instance) then measure the voltage that the motor is generating.

MrChips

Joined Oct 2, 2009
28,485
As we don't know exactly what is moving, here are some resistance sensors:

1) slider potentiometer
2) contact sliding on carbon track
3) contact sliding on resistance wire

For non-contact solution you can use photo detectors and a wedge-shaped slot.
For a digital quadrature encoder, you can salvage an encoder strip from a printer or bed scanner.

BobTPH

Joined Jun 5, 2013
6,516
The TS has used a train as an example. I doubt there's a drive mechanism, at least not from where my imaginings have gone. I agree with Ya’akov we need to know more about what you're trying to sense as opposed to imagining a train on a couple of 18 inch long wires. Otherwise you're going to get suggestions such as motion detectors, DME (Distance Measuring Equipment) and a whole host of other non-related solutions to your project. If your project is "Top Secret" and you're not willing to help us help you then you're never going to find a solution. Suggestions as potentially useless as this.
Yes, that is why I said if it is applicable.

joeyd999

Joined Jun 6, 2011
4,816
About a hundred years ago, I designed a "ball balancer" with an analog PID loop. I used a pivoting stretch of N gauge model rail road track attached to a servo, and a ball bearing (or is it bearing ball???) rode on the track, conducting one rail to the other at the point of contact.

A 100mA current source drove one rail, and a high precision differential amp with appropriate gain picked up the small voltage changes on the second rail.

The biggest problem was noise during motion of the ball (make/break kind of stuff), but was resolved with some filtering (and polishing of the ball and rails).

Worked well. My intention was to build a three track balancing ball clock -- the seconds rail was to rock the lower ball from end to end each 30 seconds, and the minutes and hours would keep the ball locked at their respective positions and roll on the minute/hour to the next position.

Never finished the final clock. But the one track worked well.