Electronics > Projects, Designs, and Technical Stuff
Ground leakage measurement
(1/1)
petergriffin56:
Hi everyone,

I am looking to detect whether when a motor is running inside a machine I don't have access to. This motor is driven by a PWM signal, from the design of the device and how powerful the signal is I expect a significant amount of capacitive coupling to the chassis. Something similar has already been done by researchers at Inverse Path (link :https://dev.inversepath.com/download/tempest/tempest_2009.pdf), they measured these signals by providing an alternate path back to earth, look at the presentation if you want a good explanation.

I'm interested in shunting the ground connection between the DUT and the power line (at the outlet). I'm just not sure what resistor value I should use and what kind of protection I need for my measurement equipment (it is an SDR). Since the input is 50ohm, I was thinking of matching that with the same value in series and using a capacitor in series to high pass + remove DC. Also, the SDR has good AGC and dynamic range so I don't think I need to provide additional amplification.


What do you guys think ?
moffy:
Can't you just measure the load current into the black box to tell when the motor is on or not?
Gregg:
Various companies make solid and split core current transformers with adjustable thresholds just for the purpose you have outlined. 
Veris Industries Hawkeye line is one that I have used in the past; they make them in many configurations.
https://www.veris.com/Category/Current-spcMonitoring.aspx
EEEnthusiast:
If you just want a non invasive way and if the motor is loud enough, a simple microphone pick-up will do the trick for you.
petergriffin56:
I probably should have been clearer about my intentions. The end goal is to pick up EMI from this motor. Considering the low frequency range, I'm expecting the ground wire to radiate the coupled signal pretty effectively. I'm really interested in the amount of noise that couples to this ground wire as I can't really calculate the parasitic capacitance inside, and measuring it would be simpler and more accurate than trying to estimate it.
Like I stated, the end goal is non intrusive in nature, I've thought about current probes and vibration / sound side channels but they don't fit my bill (for other reasons). I just want to know for now the kinds of signals that leak to the ground (which harmonics couple more than others). Also, the amount of signal leakage is dependent on the parasitic capacitance (constant, as intrinsic to the design of the machine) but also the inductance / impedance of the ground wire. As the wire gets longer, so does its inductance. Ideally a sort of LISN (line impedance stabilisation) device would be used to simulate how the signal responds to different lengths of wire, all this by means of a tunable inductor.

I've added a schematic to show what I have in mind. What I would really like is your opinion on whether this is the best design for my need and if my equipment is protected from any damage.
Navigation
Message Index
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod