Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Motor em interference and tapping into sensor

  1. Sep 12, 2005 #1
    I have a photointerrupter sensor that is near an electric motor in a scooter and i'm tyring to tap into it. When I tapped into the sensor with a wire, the scooter stops running. Initially I thought it was my speedo circuit but then i found out that if I tap into the sensor with just a plain wire the scooter would also stop running. The wire is not connected to anything! If the wire is long enough the scooter would stop running. This suggests that it is because of the EM waves of the motor cause voltage spikes in the wire tapping into the sensor and distored to the signal thus cause the computer to stop the motor.

    Can somebody tell me how to overcome this?
  2. jcsd
  3. Sep 13, 2005 #2


    User Avatar

    Staff: Mentor

    Which scooter is it, and what is its photointerrupter circuit used for? Why do you want to tap into it? Here are a couple of ideas:

    -- Mount your own photointerrupter next to the stock one, if they can share the same interrupting mechanism. Like, if the first one is placed over a rotating disk with holes, just stick the 2nd one next to it and do your own sensor circuit.

    -- If you end up having to tap off of the stock sensor, do it in a less invasive way. Use an isolating impedance (a few kOhms and some inductance, for example), and use your own opamp circuit to amplify and condition the signal. Be sure to put the isolating impedance right at the tap-off point.
  4. Sep 20, 2005 #3
    can you elaborate on "ioslating impendance?" What does that circuit look like?

    THe scooter seems to be affected even when a wire connected to nothing is tapped into the sensor.:confused:

    The stock photoi. is for RPM calculation. The scooter uses it to brake after certain speed and for movement detection.
    Last edited: Sep 20, 2005
  5. Sep 21, 2005 #4


    User Avatar

    Staff: Mentor

    Sure. When probing sensitive points in a circuit, it is often necessary to use an isolating impedance and amplifier to keep from disturbing the "device under test". Think of it like an oscilloscope probe -- the input impedance of most probes looks like a 10MegOhm resistor to GND in parallel with a few pF to GND. When you connect a 'scope probe to a node in a circuit, it usually will not alter the signal at that point, and you get a good representation of the signal on your 'scope display. Check out this tutorial about 'scope probes at Tek's website:


    The components in a typical 10x 'scope probe are shown at the bottom of the web page. The fixed parallel "compensation" capacitor around the 9MegOhm input resistor in the probe tip works with the adjustable compensation capacitor in the probe body at the 'scope to flatten the frequency response of the overall probe.

    In your application, when you hook a 'scope probe to the sensor output, does that create any problems? Probably not. That's because the 'scope probe cable is coax with a grounded shield, and because the isolating 9MegOhm resistor in the probe tip is isolating the 'scope probe cable from the sensor node, and noise is not getting injected into the sensor circuit. The compensation capacitor around the 9MegOhm input resistor in the probe tip would tend to inject noise into the node if the probe cable were not coax. If you were not using coax for the sensing wire, then you would not put a cap around your isolating resistor, and you would have to minimize the input capacitance of your sensing circuit by locating the first amp right at the sensing point, with the isolating resistor between the node being probed and your first amp's input.

    When you connect a wire to the sensor, the noise from the ignition is probably coupling into the wire capacitively, and since there is no isolating resistor right where the wire connects to the sensor, all that noise gets injected into the sensitive node. If the frequency of the noise is high enough, then the wire can also act as a receiving antenna, and inject its terminal voltage into the sensitive node.

    When you isolate your measuring circuit from the node with some impedance, your measuring circuit will generally only see a fraction of the original signal (because your goal is to be as non-invasive as possible in your measurement). That means that you will generally need to amplify the monitored signal, to get it back up to an amplitude that you can use in the rest of your measuring circuit. Opamps are usually good choices for this amplification stage.

    Since the ignition noise is pretty high frequency in nature, you may be able to use an inductance as an isolating element. Use an oscilloscope and try different isolating schemes to see what works.

    BTW, since this opto sensor on your scooter sounds like part of a safety circuit, I'll go back to my original suggestion that you put your own photointerrupter next to the scooter's opto, and sense the holes or slots or whatever with your own sensor. That will be a totally non-invasive way of getting the info that you want, so that there is no chance that your circuit will interfere with a safety feature of the bike.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook