1
\$\begingroup\$

I am trying to differentiate between two kind of marbles. I am using the marbles as a second resistor in a voltage divider and measuring the voltage drop on it. Theoretically there must be a difference in the voltage drop since the materials are different but the resistances of the marbles are too small to be measured (multi-meter gives 0 voltage for the each marble ) . My goal is to amplify the voltage difference across each marble , so I can connect them to a microcontroller that can differentiate the material of the marbles according to the value of the voltage. Thank you .

UPDATE: The two marbles (sphere balls) are of materials brass and steel both having diameter 20mm.Resitance of brass:1.003* 10^-6 , resistance of steel : 1.1 * 10^-5 The voltage drop on the marbles will be connected to a microcontroller that should differentiate between the material of the balls and control a gate (opens it or not depending on which material).

Measuring setup: A ball will be moving on a track and two contacts will be provided by thin aluminum sheets covering the inner track

\$\endgroup\$
4
  • 1
    \$\begingroup\$ are we talking marbles as in round, glass objects or metal ball bearings? \$\endgroup\$ Apr 23, 2016 at 21:16
  • 6
    \$\begingroup\$ Measuring resistance differences between similar glass objects is unlikely to be successful. Glass is such a good insulator that contaminants on the surface will affect the measured resistance far more than the actual glass. I think that you need to come up with a different technique. \$\endgroup\$ Apr 23, 2016 at 21:17
  • \$\begingroup\$ We are talking metal balls; brass amd steel \$\endgroup\$ Apr 24, 2016 at 7:50
  • 1
    \$\begingroup\$ Your update gives information that should have been in the original question although you have given resistivity of the materials instead of calculating expected resistance of the balls which would demonstrate how difficult the problem is with your approach. You are unlikely to succeed with only contact on a track. I suggest you forget that idea and differentiate by ferro-magnetic properties instead. Either use an inductive proximity sensor to sense the steel balls or use an electromagnet to steer them. \$\endgroup\$
    – Transistor
    Apr 24, 2016 at 8:50

3 Answers 3

3
\$\begingroup\$

As you have discovered, the resistance of a 20 mm sphere of metal is very low. This combined with contact resistance will make it impossible to use a resistance measurement technique to differentiate between steel and brass. A much more reliable approach would be to detect ferrous balls and switch the diverter when these are detected. Let everything else go undiverted.

Rockwell Automation's Principles of Operation for Inductive Proximity Sensors gives a good explanation of practical use and operation of these devices.

enter image description here

Figure 1. A basic inductive proximity switch setup.

enter image description here

Figure 2. Response as a function of target distance.

enter image description here

Figure 3. Correction factors for steel and brass.

Figure 3 is the interesting one. It shows that the sensor can discriminate between steel and brass simply by increasing the distance. In your case you would:

  • Mount the sensor over the track looking down at the balls passing below.
  • Move the sensor in until it detected steel balls reliably.
  • Confirm that it doesn't react to brass balls.

schematic

simulate this circuit – Schematic created using CircuitLab

Figure 4. Full circuit.

Figure 4 shows the addition of a timer with an (adjustable) off delay to hold the solenoid on for long enough that the ball has been diverted before switching back to default position.

  • Sensors are available to operate from 10 to 24 V DC.
  • PNP types switch a load that is connected to ground and are sometimes referred to as "sourcing" type since the current comes from the switch.
  • NPN types switch a load that is connected to V+ and are sometimes referred to as "sinking" type since the current comes from somewhere else but is "sunk" to common through the switch.
  • Pick PNP or NPN to suit your timer.
\$\endgroup\$
0
1
\$\begingroup\$

schematic

simulate this circuit – Schematic created using CircuitLab

Figure 1. Your test circuit. Figure 2. Equivalent schematic.

You are forgetting that your measurement is going to give you contact resistance which will be much higher than your ball resistance.

schematic

simulate this circuit

Figure 3 and 4. Kelvin connection.

You need to make a Kelvin connection to the part being measured. Since little current is flowing in the meter circuit and it has high resistance, the meter contact resistance becomes less significant.

How to measure:

  • Drive as high a current as you can through the ball to generate a significant voltage.
  • Measure the current with an ammeter.
  • Measure the voltage across the ball using Kelvin contacts.
  • Calculate the resistance from R = V / I.
\$\endgroup\$
6
  • \$\begingroup\$ I have never heard of the Kelvin connection and I don't understand how it will allow the multimeter to get a non zero value? \$\endgroup\$ Apr 24, 2016 at 8:03
  • 1
    \$\begingroup\$ Did you look it up? The high current will give the non-zero value. The Kelvin connection will ensure that you are not measuring voltage drops caused by contact resistance. Your original question is poor because you haven't described your setup and test equipment. Put this information in your original question and not in the comments. Include information on the materials and diameters. Calculate what resistance you expect. (You could assume a cylinder rather than a sphere to make the calculations easier.) \$\endgroup\$
    – Transistor
    Apr 24, 2016 at 8:09
  • \$\begingroup\$ I am sorry for not being descriptive; I have updated the question. I have searched on kelvin connection but I wasn't able to understand it well, it would be great if you provide me with a good source for assembling this circuit. \$\endgroup\$ Apr 24, 2016 at 8:39
  • \$\begingroup\$ See websrv.mece.ualberta.ca/electrowiki/index.php/Kelvin_Connected for a description of a Kelvin connection. \$\endgroup\$ Apr 24, 2016 at 15:44
  • \$\begingroup\$ @PeterSmith: This approach is doomed. The ball bearings are running down a track coated somehow with aluminium sheets. See my proximity switch answer. \$\endgroup\$
    – Transistor
    Apr 24, 2016 at 15:47
1
\$\begingroup\$

I'm going to go to what I see as the end-game on this...

My goal is to amplify the voltage difference across each marble , so I can connect them to a microcontroller that can differentiate the material of the marbles according to the value of the voltage.

Use AC but not regular AC - I'm talking about several hundred kHz typically 300 kHz as used in food and pharmaceutical metal detectors (yes I've designed a couple). These machines can discriminate size and material content quite successfully. I'm talking about iron to brass to stainless steel differentiation: -

enter image description here

If you want something much simpler then state your needs more explicitly.

\$\endgroup\$
5
  • \$\begingroup\$ It is brass and steel, but how can I assemble such circuit since this is my first electronics course \$\endgroup\$ Apr 24, 2016 at 7:53
  • \$\begingroup\$ Also is there a way to convert from the mains supply to this frequency ? \$\endgroup\$ Apr 24, 2016 at 8:00
  • \$\begingroup\$ It's called an oscillator and its output would drive a coil which the marbles pass through. You'd also use two antiphase detect coils to detect material content. Amplitude and phase of signal is what you are using. Go look up inductive balance metal detector. \$\endgroup\$
    – Andy aka
    Apr 24, 2016 at 8:36
  • \$\begingroup\$ So this is how I imagined the circuit imgur.com/w9tcuni , however where anti-phase coils go or they are included in the main coil. \$\endgroup\$ Apr 24, 2016 at 9:07
  • \$\begingroup\$ See my answer to this question: electronics.stackexchange.com/questions/74952/… \$\endgroup\$
    – Andy aka
    Apr 24, 2016 at 9:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.