Wireless Adapter vs. Powerline Adapter Transfer Speeds

[GabeVR]

Hi! First post here so I apologize if I come off as a complete noob!

So for Christmas I got a pair of power line adapters and I just got around to trying them out. I wanted to see the difference in speeds between my WLAN transfer speeds using the Edimax 150Mbps 802.11b/g/n Nano USB adapter, and using my NIC (100Mbps Fast Ethernet) on my desktop with the 200Mbps TrendNet TPL-307E2K powerline adapters.

I used Lan Speed Test to perform the tests. Both tranfers were made to a laptop connected directly to a gigabit port on a Linksys WRT54GS. The file size is 100MB. The test on the right is using the powerline adapters connected to the NIC on my desktop and the test on the left is using the USB wireless adapter.

As you can see from the image, the wireless adapter smoked the powerline in writing (23Mbps vs. 7Mbps) and was also better in reading (19Mbps vs. 14Mbps). I've heard both good and bad things about powerline adapters so I didn't really know what to expect but I did think that the powerline adapters would out perform the wireless adapter.

So, my question is, has anyone done similar tests and found that the wireless connection outperforms the powerline adapter? Maybe I do not have a proper test setup? Anyways, thanks for reading, and I look forward to hearing what you guys think!

 

[CaptainSTX]

Improving Powerline Speed

You should to be able to do much better than that.

I use first genertration 85Mbps adapters. I normally get 25Mbps - 40Mbps.

First establish a base line speed by plugging both adapters into a non GFI protected outlets in the same room on the same same circuit. Preferably as near your circuit breaker panel so the AC wiring runs are as short as possible.

Be sure that BOTH adapter are plugged directly into the wall, no extension cords, surge protectors, etc. Again if possible avoid testing using outlets that are GFI protected.

Once you determine the best possible speed with your network setup you can locate where the adapters are plugged in to get as much speed as possible.

 

[thiggins]

As an aside, the WRT54GS has 10/100, not Gigabit Ethernet ports. It shouldn't make a difference in this test, however.

Make sure that both powerline and Wi-Fi connections are not active at the same time. Unplug/disable the other adapter when running tests.

The Captain's other suggestions apply. You can check the Powerline Charts to see what other 200 Mbps powerline adapters performed in our testing.

 

[GabeVR]

Thanks for the replies!

@CaptainSTX: Initially I didn't see a point testing it near the router because I don't intend on using it in that room but for comparison sake I can see that it would be helpful. I did have it plugged into a surge protector and have since moved it to a dedicated non-GFI wall outlet.

@thiggins: Sorry for the mistake on the WRT54GS port speeds but you're right, it doesn't make much difference because I would never reach the maximum throughput of 100Mbps anyway. When testing I made sure to disable each adapter completely.

After moving the powerline adapter at the router to it's own wall outlet I ran the tests again. The results were better but the wireless write (left side) still beat out the powerline write (right side). The powerline read did perform better but I still thought it should have been higher. I included an image of the TrendNet Utility just to show what speed it was recording.

Anyways, like I said before I'm new to powerline networking so I didn't really know what to expect. Thanks for the help!

 

[thiggins]

That's not really a significant difference between powerline and wireless.
These results are still room-to-room or same room?

Powerline adapters always should be plugged directly into wall outlets for best performance. Never in surge-protected outlet strips.

 

[GabeVR]

That's not really a significant difference between powerline and wireless.
These results are still room-to-room or same room?

Powerline adapters always should be plugged directly into wall outlets for best performance. Never in surge-protected outlet strips.

No, you're right, it's not a significant difference between wireless and powerline, but it is a big difference if you compare the test on the right in my original post when I had the one adapter running through a surge protector.

This is still room-to-room; my room is directly above the room with the other adapter.

Do you think I should still be getting better performance or is this pretty average for 200Mbps adapters?

 

[thiggins]

My testing shows that 200 Mbps HomePlug adapters can consistently produce 40 Mbps. Test technique can account for some of the difference between our results.

In my experience, you should look for noise sources. I have found the worst offenders to be cellphone chargers. Unplug other devices in the room on both ends of the connection and see if that makes a difference.

 

[GabeVR]

My testing shows that 200 Mbps HomePlug adapters can consistently produce 40 Mbps. Test technique can account for some of the difference between our results.

In my experience, you should look for noise sources. I have found the worst offenders to be cellphone chargers. Unplug other devices in the room on both ends of the connection and see if that makes a difference.

Wow, ok so I should be getting much better performance. On the adapter in my room I am using the "bonus" outlet on the adapter itself to plug my power bar into. The power bar is facilitating two PC's, printer, speakers, audio interface etc. Do you think this would affect performance?

 

[thiggins]

On the adapter in my room I am using the "bonus" outlet on the adapter itself to plug my power bar into. The power bar is facilitating two PC's, printer, speakers, audio interface etc. Do you think this would affect performance?

Asked and answered. You need to unplug things and experiment yourself.

 

[stevech]

yes... power line data works best if the two devices (layer 2 bridges) are plugged in such that there are no computers, TVs, home theater, surge suppressors, etc. on that same circuit (breaker) - or at least there's 20' of wire (inside walls) from each device to that list of offending things.

Some types of light dimmers cause too much noise, esp. at low light levels.

and so on.

That's why I like MoCA.

But with some non-trival amount of experimenting and juggling, IP on power line usually can work. In the US, beware too of the two-phase power and the difficulty that sometimes happens when the two bridge devices are on different phases, according to happenstance in the breaker box.

 

[thiggins]

In the US, beware too of the two-phase power and the difficulty that sometimes happens when the two bridge devices are on different phases, according to happenstance in the breaker box.

That issue was solved generations ago. See SmallNetBuilder's Powerline FAQ
item 6.

 

[stevech]

That issue was solved generations ago. See SmallNetBuilder's Powerline FAQ
item 6.

I know that Tim and other may differ, but...:

That diagram makes an unsubstantiated claim.

Being an RF engineer by profession, I've not agreed sight-unseen to the premise about coupling across phases without a specific coupling device. At best, it depends on the kind of box, kind of wire from the mains, kind of 220V wiring to 220V appliances (if any, etc.). Highly variable.

 

[thiggins]

My test locations are on different phases, Steve and my results are not "highly variable". I do see performance reduction from noise sources.

You have a pair of adapters so can run your own experiments instead of spreading FUD.