Connect with us

Testing the inductance and resistance with oscilloscope

Discussion in 'Electronics Homework Help' started by mike_980, Mar 24, 2013.

Scroll to continue with content
  1. mike_980

    mike_980

    16
    0
    Mar 24, 2013
    Hi guys,
    no doubt this will be a very easy question for everyone here but electronics is far from my strongest subject!

    For my final year uni project I am working on an eddy current engine dynamometer. For those that don't know it is basically a rotor which spins with coils around it, when current is run through the coils eddy currents are produced in the rotor and applies a braking torque through ohmic heating.

    For my calculations of how much current is required etc, I need some of the electrical characteristics of the coils, these being the resistance and the inductance.

    My supervisor suggested using an oscilloscope to test these where the basic method would be:

    1. Ensure both switches are in the open position, then create a series circuit connecting the battery, switch A, precision 1ohm resistor and dynamometer coil

    2. Connect Switch B and 1Kohm resistor across the dynamometer coil

    3. Connect the oscilloscope across the precision resistor

    4. Turn on the oscilloscope and adjust the settings to read the correct scale

    5. Adjust the settings on the Oscilloscope to store the waveform

    6. Start the oscilloscope recording

    7. Close switch A and record the data into the oscilloscope, noting the filename on the lab test sheet

    8. Discharge the energy in the coils (close switch B)

    9. Open switch A

    10. Fit an exponential curve to the data to obtain the RL time constant and the resistance of the circuit

    11. Subtract the effect of the 1Ohm precision resistor to obtain achieve an estimate of the properties of the windings.


    As I understand it, I am calculating the RL time constant using this method from the time is takes the coils to discharge and I use a laplace transform table to get to the values I need? I have never touched on this theory before though so any help at all would be greatly appreciated!
     
  2. Harald Kapp

    Harald Kapp Moderator Moderator

    11,667
    2,700
    Nov 17, 2011
    It would help very much if you include a schematic showing us the setup and which switches are A and B.

    Your conclusion is right.
    - measure
    - fit the curve (find the parameters of an RL-circuit that match the measured data best)
    - subtract the effect of the measurement resistor (1 Ohm) to get the date for the coil.

    You don't have to use laplace transform or anything similar. Wolfram explains how to fit data to an expected exponential curve. Even Excel can do this.

    Btw: It was me who moved the thread to the homeworks section. I think it belongs here.
     
    Last edited: Mar 24, 2013
  3. mike_980

    mike_980

    16
    0
    Mar 24, 2013
    No problem, I wasn't entirely sure where to place it and I am not looking for straight answers to plug into my work just help in understanding the theory.

    The above method is basically what my tutor sent me and I know nothing other than what is there really, my knowledge is in the mechanical side of engineering so it feels strange coming back to basics for something again!

    Good to know I wont have to use laplace transforms, the reading I have done just gave me a headache.

    So presumably, the oscilloscope will give me an output file (or datapoints I can plot) which I can fit a curve to using your very helpful link (thankyou), which I can then basically pull my values from. Then how do my resistance and conductance values come from this curve?
     
  4. Harald Kapp

    Harald Kapp Moderator Moderator

    11,667
    2,700
    Nov 17, 2011
    Basically you have an RL circuit.
    If you apply a step input (aka close a switch), the voltages and currents in this circuit have an exponential step response. The oscilloscope records cvoltae over time (and if you measure the voltage across the 1 Ohm sense resistor that voltage is equivalent to currentg over tiem because of Ohm law).
    The wiki link gives the requires equations. Now find parameters L and R such that the equations give the same curve a the one you fitted from the measured data. If both computed curve and fitted curve match, you have found R and L of the equivalent circuit. All you need to do is subtract the sense resistance since it was artificially introduced to measure the current, but it will not be part of the final circuit.
     
  5. mike_980

    mike_980

    16
    0
    Mar 24, 2013
    Yeah that makes sense, thanks for your help! I am doing the test tomorrow so if I have any other troubles I will post them back in here! thanks again
     
  6. mike_980

    mike_980

    16
    0
    Mar 24, 2013
    OK I spoke with one of the technicians at uni today and he said to lay out my circuit like this, the oscilloscope would measure the voltage across the resistor (1ohm) and I think the voltage across the output of the coil with respect to earth then subtract one from the other to get the curve I need is this correct?

    The black box is the power source with 3 AA batteries (4.5V) and the when I measured the resistance in the coils with my multimeter this morning I read 9Ohms

    Sorry about the picture, I have attached it now instead of using the URL hopefully that has worked.
     

    Attached Files:

    Last edited: Mar 26, 2013
  7. Harald Kapp

    Harald Kapp Moderator Moderator

    11,667
    2,700
    Nov 17, 2011
    No picture. Is it on another seerv, unaccessible to me? Post it here ->go advanced -> insert image
     
  8. mike_980

    mike_980

    16
    0
    Mar 24, 2013
    Sorry, I have attached it now instead of using the URL hopefully you can now view the image
     
  9. Harald Kapp

    Harald Kapp Moderator Moderator

    11,667
    2,700
    Nov 17, 2011
    Sorry, I don't think so.
    The voltage across the resistor is only a proxy for the current.

    Let me see if I understand correctly. These are your instructions:
    In step 2 the resistor and switch are in parallel to the coil, but switch, 10k resistor and coil are in series with the 1Ohm resistor. Steps 8 and 9 do not make sense to me. In step 8 you are advised to close switch B to discharge the coil (btw: a coil is not "charged", it stores energy inductively). However, at this time switch A is still closed meaning that the voltage source can still drive a current through the inductor, The 10k resistor in series with switch B will distract a negligible amount of current. Only when you open switch A in step 9 will the current through the coil be diverted throigh the 10k resistor and switch B.
    Or is it me who gets all this backwards?

    And where does the diode come from that you show in your setup? It's not mentioned in the task description.

    Would you kindly read this writeup? It contains all the theory and on page 7-5 there is an example. Once you know the time constant of the circuit, (tau=R/L) you can use your knowledge of the know Rs to determine L.
    The setup is a bit different from what your task description says, but it's well explained.
     
  10. mike_980

    mike_980

    16
    0
    Mar 24, 2013
    Thanks for the help with this so far! Unfortunately due to university messing me about with equipment and not been able to get hold of things and with me been ill for a few weeks I have not progressed very far and taking this up again now some of my other work is out of the way.

    my next question:
    With me being an automotive student and a huge car enthusiast and working on cars often, I guess it would be useful for me to have my own oscilloscope, but I don't want or need a big one but I see there are pocket ones out there. I have seen the Velleman HPS140i which I can get for about £100 and the ARM Nano dso201 which I can get for about £50, would both of these be good enough for this project? And would both be good enough for testing things like crank sensors on my car? I read the pocket ones arn't great because the sample rate is quite slow and they only have one channel but I guess I would only ever need one channel for my uses?

    Thanks in advance
     
    Last edited: Apr 24, 2013
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-