I threw together a simple setup to test the filters' effectiveness. I plugged a pair of NETGEAR AV200 adapters into an unfiltered outlet strip, with one adapter's Ethernet port connected to an Acer Aspire 1810T netbook and the other connected to my LAN switch.
I first tried a cellphone charger as a noise source, but it didn't put a large enough dent in throughput. So I took a halogen floor lamp with dimmer, plugged it into the same outlet strip and played with the dimming while watching an IxChariot throughput run until I got a good dip in performance.
Figure 4 shows the lamp and dimmer combo plugged in at around the 15 second mark and uplugged just before 30 seconds. Throughput is reduced around 10 Mbps with the noise source present.
Figure 4: HomePlug AV throughput with dimmer noise
I then connected the lamp and dimmer through each of the three filters, while running the same IxChariot throughput script and plugging and unplugging the filter at the same times during the run (15 seconds and 30 seconds).
Figure 5 shows how the Corinex filter did (top trace).
Figure 5: Corinex filter effect
Figure 6 shows how the Plaster IX2 filter did (top trace).
Figure 6: Plaster IX2 filter effect
And Figure 7 shows the effect of the NETGEAR AV+ 200's built-in filter.
Figure 7: NETGEAR AV+ 200 filter effect
In each case, the filter was quite effective in keeping the lamp / dimmer noise from reducing throughput.
I'm sometimes asked how far away (or close) noise sources can be from powerline networking gear to have them not affect (or affect) performance. The answer, of course, is "it depends". Depending on how many sources and how strong they are, you could see a problem with noise sources plugged into outlets on the other side of a room. Effects from noise sources in another room? Not so likely.
My experience with high-frequency noise sources (dimmers, chargers, power supplies) is that it's a very close-range (same outlet) effect. Filtering, however, should be applied as close to the powerline adapter as possible. Which makes adapters with built-in filters like the NETGEAR AV+ 200 very effective.
Updated 12/17/2010: Added line cord results
After this review was published, a reader email prompted me to try plugging my noise sources into a simple 6 foot extension cord instead of a powerline filter. To my surprise, the extension cord worked as well as the filters!
This might not work on strong noise sources. But it's worth a shot before spending $20.
So if you're not getting the throughput that you think you should be getting from your powerline gear, a $20 powerline filter (or even an ordinary extension cord) could be a very worthwhile investment.