RF question: Filter output voltage bigger than input?

Factual Questions
1

I posted this on another forum too:
I am doing a frequency sweep of a typical ADSL filter using a scope and a generator as shown here: EEVblog #396 - Bode Plotting on Your Osciloscope - YouTube

The filter has a Line input on one side and Modem and Phone outputs on the other side. I am feeding the output from the generator to the input of the filter and

I have set up the generator 1 Hz to 1MHz, 1 second linear sweep and 1 Volt peak to peak.

When testing the Modem output, things go as expected:

But when testing the Phone output it seems I am getting out almost twice the input voltage :o

What is going on here?

2

Resonance?

3

You should put a 600 ohm resistor across the phone line output from the filter. It will probably only work correctly that way.

The filters are built to work into a particular load. If that load isn’t present, then you will get strange results - improper frequency response being the expected thing.

Filters are sort of like springs. If you feed them a signal at their resonant frequency, then the amplitude can exceed the input amplitude. Being passive, though, they cannot exceed the input power. With the intended design load, the output amplitude should go back down.

4

Thanks!

This is with a 580 ohm resistor in parallel:

Now thats a lot closer to expected.