Measuring power conditioners – a fancy name for filters – is terribly complex. This is because you actually have to measure with a dynamic test signal. After all, if you connect an amplifier to a filter, for example, it will load the filter dynamically: the load changes with the music signal. And that is quite different from a static test tone or just noise. Anyway: we have to work with the resources we have!
Now we cannot – yet – emulate a music signal or heavy dynamic signal, but we can measure the effectiveness of a filter.
NB: since july 2023 we also use a programmable AC source in this test. That device can emulate all sorts of distortion and noise.Â
First of all, we can see how the filter deals with white noise (60 MHz bandwidth). We measure that both in the low spectrum on the Prism and in the higher spectrum on the spectrum analyzer. Below is an example of a Yeti Reference filter (gen1) having to deal with 5 volts of white noise. You can see that it attenuates about 40 dB at 80 kHz and starts working from about 1000 Hz.
A second test is more practical. We have a pair of dirty switching power supplies that we measure directly into the LISN AND then with the filter in between. This way we can see how much of the junk has been removed.
Above is an example of the Pura Sauropoda. In doing so, we made a measurement with switching noise from a Class D amplifier. It is clear that the Sauropoda eliminates the peaks up to at least 5 MHz and also suppresses this considerable in the lower – more important – spectrum. This creates calmness in the reproduction.
Concluding
We hope you have now gained a little more insight into our approach. We always try to paint as complete a picture of a product as possible. Now we will always try to make the measurements more complete and better. Changes in the set-up will always be explained in the reviews and / or a background article.