Powerline throughput question

[Harry Z]

Aloha. I have a situation where I am forced to use powerline adapters to get ethernet into a room in my house. So, I purchased a pair of Netgear PL1200-100PAS powerline adapters to see if they would work. I have 300 Mbps going into the first powerline adapter, and coming out of the 2nd powerline adapter, I have 70 Mbps. This is when I have both of them connected on the same circuit in my house (verified by turning the circuit breaker off).

I realize that there will be some loss of throughput when using any powerline adapter, but this seems excessive. What would be a "real world" reasonable throughput with these adapters (assuming 300 Mbps going in to adapter 1)?

And, what can I do to improve the throughput? I've already gotten a warranty replacement from Netgear and the performance did not change. The only thing left I can think of is to change the outlets. They are 25 years old. I can do this myself, but would like to know if it would possibly be worth the effort.

Mahalo for your assistance.

Harry Z.

 

[ColinTaylor]

Seems bit on the slow side, but not by much.

http://www.smallnetbuilder.com/lanwan/lanwan-reviews/32717-netgear-pl1200-powerline-1200-reviewed

Depends how far apart the adapters are. Also, make sure that you plug them straight into the wall sockets and not into power strips.

 

[thiggins]

As Colin said, 70 Mbps is within the range of possibilities.

"Wall wart" power adapters generate a lot of high frequency noise. It seems the smaller the adapter (like smartphone/USB power adapters) the higher the noise.

Start by unplugging anything like that in the "far" room, since noise has the most effect on the receiving end.

 

[sfx2000]

@thiggins - recall on the primary site - there was a good article on Homeplugs and performance - and like you mentioned, nearby devices at any node, they can be impacted by noise sources (chargers, appliances, circuit hops, etc). It's an old article, but it was relevant back then, and I suspect, even more so with the higher bandwidth devices (both HPAV2 and G.hn) now...

Can't find the link...

 

[sfx2000]

And, what can I do to improve the throughput? I've already gotten a warranty replacement from Netgear and the performance did not change. The only thing left I can think of is to change the outlets. They are 25 years old. I can do this myself, but would like to know if it would possibly be worth the effort.

Check your wiring - an outlet tester is a good tool to have - PLC or whatever, and they're cheap...

Less than 10 bucks over at Harbor Freight....

http://www.harborfreight.com/electrical-receptacle-tester-with-gfci-diagnosis-32907.html

Homes around your age (25 years) - some are copper, some are aluminum - so something to consider (there was a class-action lawsuit out here in Cali regarding aluminum wiring, FWIW) - also check hot-neutral-ground relationships across the outlets (it's easy to get wrong, and most things don't care, but it's a shock hazard) - the outlet tester I mentioned above will show this.

Jumping circuits - depends on the breakers - and in newer homes, we have GFCI to worry about - old code used to have just bathrooms/utility/kitchens, but has expanded into bedrooms - and there, there are GFCI's that are PLC friendly, and some that are definitely not...

 

[Harry Z]

Seems bit on the slow side, but not by much.

http://www.smallnetbuilder.com/lanwan/lanwan-reviews/32717-netgear-pl1200-powerline-1200-reviewed

Depends how far apart the adapters are. Also, make sure that you plug them straight into the wall sockets and not into power strips.

So, 75% loss (300 Mbps in, 70 Mbps out) in throughput is "normal" for a powerline adapter? Perhaps I should mention that I am measuring this using speedtest.net in Chrome. Same laptop used in both speed tests. The speeds are consistent through many repeated tests.

There is approximately 16 feet between the outlets. Both adapters are plugged directly into the outlets (i.e.: no extension cords or power strips).

Mahalo,

Harry Z.

 

[Harry Z]

Check your wiring - an outlet tester is a good tool to have - PLC or whatever, and they're cheap...

I'll check on this. The tester I have is probably 30+ years old. May have to buy a new one...

Homes around your age (25 years) - some are copper, some are aluminum - so something to consider (there was a class-action lawsuit out here in Cali regarding aluminum wiring, FWIW) - also check hot-neutral-ground relationships across the outlets (it's easy to get wrong, and most things don't care, but it's a shock hazard) - the outlet tester I mentioned above will show this.

All copper wiring in my house. And I think my tester checks the hot/neutral/ground relationships.

Jumping circuits - depends on the breakers - and in newer homes, we have GFCI to worry about - old code used to have just bathrooms/utility/kitchens, but has expanded into bedrooms - and there, there are GFCI's that are PLC friendly, and some that are definitely not...

I had the circuit breaker box replaced a year or two ago, so I definitely have new breakers. Would the breaker cause a problem when both outlets are on the same circuit breaker? And, is there an up-to-date list of "friendly" and unfriendly brands of circuit breakers? I found a mention of this elsewhere on this site, but I think it was a few years old.

Mahalo for taking the time to respond.

Harry Z.

 

[Harry Z]

As Colin said, 70 Mbps is within the range of possibilities.

"Wall wart" power adapters generate a lot of high frequency noise. It seems the smaller the adapter (like smartphone/USB power adapters) the higher the noise.

Start by unplugging anything like that in the "far" room, since noise has the most effect on the receiving end.

Right now, I'm testing with both powerline adapters plugged into outlets connected to the same circuit breaker. Nothing else is plugged in on this circuit.

Mahalo for responding

Harry Z.

 

[sfx2000]

So, 75% loss (300 Mbps in, 70 Mbps out) in throughput is "normal" for a powerline adapter?

PHY vs. Network - vendors always do the PHY speed, never the upper layer stuff

Hardly anyone would buy HomePlug AV2 75Mbps would they?

(same goes for the WiFi stuff)

 

[ColinTaylor]

So, 75% loss (300 Mbps in, 70 Mbps out) in throughput is "normal" for a powerline adapter? Perhaps I should mention that I am measuring this using speedtest.net in Chrome. Same laptop used in both speed tests. The speeds are consistent through many repeated tests.

There is approximately 16 feet between the outlets. Both adapters are plugged directly into the outlets (i.e.: no extension cords or power strips).

I think I'd be a bit disappointed by those figures as well. My adapters are much older than yours and only rated at 600Mbps, yet I get over 100Mbps throughput at about the same distance. That's going through a circuit breaker. But I'm in the UK so that might make a difference.

As Tim said, look for sources of interference like mobile phone chargers, etc. Perhaps for testing you could temporarily unplug everything possible in the house and try again. That might identify the cause.

Netgear should have supplied you with a utility that allows you to monitor the link rate (PHY) that sfx2000 was referring to. What does that say, does it correlate with your speed tests? Also, take one of the adapters and a laptop and move around the house testing the speed at different outlets. Starting in the same room where the other adapter is you should see something like 300+Mbps. I found that in my living room the outlet on one side of the room was significantly slower than one on the other side.

 

[thiggins]

@thiggins - recall on the primary site - there was a good article on Homeplugs and performance

http://www.smallnetbuilder.com/basics/lanwan-basics/32768-smallnetbuilders-powerline-faq-2015
http://www.smallnetbuilder.com/basics/lanwan-basics/32769-how-to-troubleshoot-your-powerline-network

 

[thiggins]

What do you get for speed when you plug the adapters into outlets in the same room?

Did you look for and remove possible noise sources?

 

[Harry Z]

As Tim said, look for sources of interference like mobile phone chargers, etc. Perhaps for testing you could temporarily unplug everything possible in the house and try again. That might identify the cause.

Netgear should have supplied you with a utility that allows you to monitor the link rate (PHY) that sfx2000 was referring to. What does that say, does it correlate with your speed tests? Also, take one of the adapters and a laptop and move around the house testing the speed at different outlets. Starting in the same room where the other adapter is you should see something like 300+Mbps. I found that in my living room the outlet on one side of the room was significantly slower than one on the other side.

Mahalo for taking the time to reply.

So, I'm back to the question I have about interference from other devices. I am testing this with both adapters plugged into two outlets that are on the same circuit breaker. No other devices are plugged into this circuit. My question is IF there is a noisy device on some other circuit in the house, will it effect this circuit?

 

[ColinTaylor]

Mahalo for taking the time to reply.

So, I'm back to the question I have about interference from other devices. I am testing this with both adapters plugged into two outlets that are on the same circuit breaker. No other devices are plugged into this circuit. My question is IF there is a noisy device on some other circuit in the house, will it effect this circuit?

Pure speculation on my part but as powerline devices are advertised as being able to communicate across different circuits, then by the same logic so can noise.

Like I said, try moving the adapters to adjacent sockets in the same room and see what speed you get.

 

[Harry Z]

What do you get for speed when you plug the adapters into outlets in the same room?

Did you look for and remove possible noise sources?

Hi Tim. Thanx for the reply. My testing is using both powerline adapters connected on the same circuit. No other devices (chargers, computers, lamps, etc.) are plugged into this circuit. Please see my previous response about checking for other noisy devices

 

[thiggins]

Hi Tim. Thanx for the reply. My testing is using both powerline adapters connected on the same circuit. No other devices (chargers, computers, lamps, etc.) are plugged into this circuit. Please see my previous response about checking for other noisy devices

Move the devices to the same room. In fact, plug them both into the same power strip (if you can fit them). We are trying to establish a performance baseline.

 

[degrub]

Hi Tim. Thanx for the reply. My testing is using both powerline adapters connected on the same circuit. No other devices (chargers, computers, lamps, etc.) are plugged into this circuit. Please see my previous response about checking for other noisy devices

You would have a common Neutral and safety ground across all circuits. You likely have two positive phases entering the house. Depending on which side the breaker is plugged into, all circuits plugged into that phase will see the same noise overlay on the 120v positive. Also, anything emitting EMI such as motor starts will broadcast on all frequencies and affect noise level on any nearby power lines. .

 

[sfx2000]

Don't forget old-school dimmer switches - they can raise a lot of EMI on the lines...

 

[Harry Z]

Sorry for the delay in responding. I really do appreciate all the input from all of you.

I don't have a power strip that does not also have surge protection, so I connected both power line adapters to a "cube tap" (see attached picture).

Using my cable company's speedtest site, I got 355Mbps going into the first PLA (black cable). And coming out of the 2nd PLA (blue cable) I got 155 Mbps. So, closer to 50% loss. Again, is this "about right" for PLA's?

Mahalo for your assistance!

Harry Z.

 

[ColinTaylor]

If I understand you correctly the other end of the black cable goes to your cable modem? You've plugged this end of the black cable into a PC and the speed test shows 355Mbps? Moving the black cable from the PC to the powerline adapter as shown in your picture and connecting the blue cable straight into your PC, you get 155Mbps. Correct?

If so, it sounds like there is something very wrong with your mains wiring (in terms of interference), or that orange cube. You should be getting over 300Mbps.