I'm working with the following R + parallel LC circuit:

I anticipate the following (ideal) response as I sweep the excitation frequency:

I think the "notch" frequency will be $latex f= \frac{1}{2\pi} \cdot \sqrt{\frac{1}{LC}} \approx 140 kHz $.

The phase shift near the "notch" frequency should be approaching +/- 90 degrees.

When I actually build and test this circuit, something "magical" does happen at 156.5 kHz, but it's not what I expect. I do get maximum attenuation (about -20dB), but my phase shift is zero. Here's the oscilloscope display



and in x-y mode:



I'd like to understand this. What is going on?

(Sweeping above and below, the phase shift grows as I get further from the resonance point. Attenuation falls, as expected. I measure the coil resistance as 0.4 ohms, but when I add this to my simulation it doesn't change much other than the sharpness of the notch.)

 

Look at the phase plot you posted at the point of peak attenuation. In theory it is an instantaneous phase jump. With real world inductors without infinite Q, there is a measurable phase transition region.

ak

 

Thanks, but what I observe doesn't look like a narrow transition region. The phase shift smoothly goes from a few degrees of lag (sweeping from below) to zero (at resonance) to leading. At 250 kHz, it's roughly 36 degrees leading and still increasing. At 50 kHz, it's 36 degrees lagging and still increasing. Shouldn't it be tending to zero by then?