line rate or more commonly referred to as powerline linkrate is not associated with tcp throughput i.e.internet throughput. i can get 1200mbps link rate on PA9020P but only 120-130mbps internet rate measured on ookla. at router direct i get 490mbps on ookla speedtest.

@Carl 

I do not believe this analysis is correct.  Behind the circuit breakers are two solid copper bus bars, each connected to the separate "phases" of a typical residential electrical supply.  Here's a link to a typical box showing how phase 1 and phase 2 are connected to "every other" breaker position.

In your picture, the breakers on the left are "double breakers", where two breakers take up the same phyical space as one single breaker.  So the top two on the left are connected to the same phase.  The next two to the other phase.  The next two to the first phase, etc.  On the right, there is a 240 volt breaker, which has one regular size breaker connected to one phase and a similar breaker connected to the other phase.

 

So, adapter #1 is on one phase and adapter #2 is on the other phase.  The signal path is not down the bus bar at all.  It goes through the main breaker from one phase to the other and then back down the bus bar to the other circuit breaker.

 

And, the communication signal goes through the bus bar, not through each of the individual circuit breakers.

 

The quickest solution is to put both breakers on the same phase, either by swapping the two double breakers on the bottom right, or by swapping the powerline circuit on the bottom double breaker with one of the circuits on the breaker above it.

 

I cannot make out whether all those breakers on the bottom right are the same amperage (they would have to be for this to work).  Also, isn't it a bit strange to have a Powerline adapter on the 15 amp circuit?  "Where I come from", 15 amp circuits are for lighting and 20 amp circuits are for wall plugs.

 

p.s. I am wrong a lot, and people usually are quick to point out my mistakes.

@Carl

My disagreement is with the concept of "jump".  The bus bar is a solid bar of copper running from top to bottom with lugs sticking up at every other position.  A signal going from the top of the bus to the bottom of the bus does not "stop" along the way.  It arrives simultaneously (speed of electricity) at every breaker on the bus at the same time.  It goes down all of the circuits at the same time.  On all but one of them, the signal just fades off to the end.  On the circuit with another adapter,that adapter senses and decodes the signal and may place a return signal back on the wire, which also goes everywhere.

 

There is no intelligence in the copper bar or the circuit breakers.  The signal does not go down one circuit, come back up and then go down another circuit.

 

Maybe I should have paid more attention in physics class.

 

I wonder, however, if the total size of the electrical wiring has any effect on performance.  In coax networks, for example, "splitters" dramatically reduce electrical signals, by 3dB per split (50%).

@Carl 

 

Sorry, I tend to be "too literal", and the science behind Powerline remains much of a mystery.  I completely agree that distance, quality of wiring & connections, etc. will degrade a signal until it is finally unusable.  It would be really useful to have some direction on how to mitigate issues.  Obviously, I can't rip out my 60 year old electrical wiring. But aren't there device(s) can be used to reduce that "electrical noise"?  Are there things that can be plugged into the outlet, and the microwave/dryer/washer/refrigerator plugged into them?  How much does going from one "phase" to another phase reduce throughput?

@Carl 

 

Thanks for the tip.  I have been "over thinking" this, haven't I.  As long as the surge protector claims "noise reduction", that means it has more then just the surge suppressor thyristors.  Back to the web I go!