Top critical review
548 people found this helpful
Grossly Misleading - Actual performance is up to 32Mbps, not 500Mbps as advertised
on April 27, 2015
I wanted to test the actual performance of these powerline adapters and I did so with a professional Ethernet transmission tester used by large telco companies to qualify links. The test I ran is called RFC2544 and it determines the maximum possible data rate at various Ethernet frame (packet) sizes.
First of all, the device is advertised as as providing data rates at "up to 500 Mbps". Now I am always cautious whenever I see "up to" in advertising claims. Literally, this means that in a best case scenario it is capable of transmitting data at 500Mbps. However when you connect an Ethernet device to one of these adapters the link speed is only 100Mbps. So how is it possible to transmit "up to 500Mbps" when the physical layer Ethernet port (PHY) is a 100Mbps device? So right away my personal opinion is that the claims are totally fraudulent. If the PHY connected at 1,000Mbs there would at least be the possibility of achieving a 500Mbps transfer rate between these two devices. If there were an asterisk next to the claim and the fine print read something like "Total combined data rate of 500Mbps possible with x number of powerline adapters on the same network." That would tell me that 500Mbps is not between two adapters, but is the total network capacity. However I found no such information on the packaging.
Nonetheless, I setup my transmission tester with a maximum data rate of 100Mbps since that all these powerline adapters are technically able of providing between any two units.
How the RFC2544 test works is that it will start with a specified frame size and transmit data between the two components of the test system at a low data rate, increasing the data rate until errors are detected. The data/frame rate that can be achieved without errors is reported as the maximum data rate for that frame size. Then the next largest frame size is tested and the process repeats until all frame sizes are tested.
The test system reports information in several categories:
THROUGHPUT (Note that on my test results below the Throughput test shows PASS because I did not set a minimum required data rate for the test, the threshold is set to 0Mbps)
- Maximum data rate for each frame size (Mbps). This is the key test that reveals the actual performance delivered by these powerline adapters. The TP-Link claims 500Mbps, which is impossible since the Ethernet port on the adapter is limited to 100Mbps.
- Maximum frame rate (frames/second)
- Utilization rate which is simply the data rate divided by the link rate, in this case it will be the same as the data rate since the link rate is 100Mbps between the tester and the powerline adapter
- Frame loss at the maximum data rate which is usually zero since the data rate is determined as the the max data rate that was achieved without errors. However for the sake of speeding up the test, the device can be set to test to the maximum specified data rate in larger steps. My tester was configured to test each frame size up to 100Mbps in 15Mbps increments.
- At the maximum data rate determined by the Throughput test the time it takes a frame transmitted from one device to reach the other device and return back to the originating device (round trip time). In a telecom network this can be affected by internet routers between two locations but in this case the two devices are separated by just a few feet as they are on the same electrical circuit in my home office.
As with Throughput, Latency is tested at each frame size.
- For each frame size, at the max data rate the measured, jitter is the difference in time between subsequent frames. Ideally jitter is zero which means all the frames in the data stream are evenly spaced making for consistent data transmission.
- Measures the number and percentage of frames lost at the desired test data rates. The Throughput tests tell you how fast the network can run without dropping frames and this test tells you what will happen if the network is forced to run at certain data rates.
Prepare to be disappointed. The first set of photos are screen grabs from the transmission testing of the powerline adapters. In short I achieved a best data rate of 32Mbps at a frame size of 2014kb. Most Ethernet applications use a frame size of 1518kb which should be more efficient, however these adapters were able to provide only 26Mbps at that frame size. So, in most uses, these provide 20Mbps transmission, not the claimed 500Mbps.
The first set of pictures with the red header bars are the results from the TP-Link powerline adapters. The green line on the Throughput plot indicates what data rate (Information Rate) should be achieved. The red line is what was actually achieved by these powerline adapters.
The second set of plots with the green header bars are from the same RFC2544 test run through wired Ethernet switches. As you can see the actual data rates were identical to the theoretical maximum data rate, delivering 100Mbps at all frame sizes. Since the Ethernet switch is a 1Gbps model I could have reconfigured the tester to run to that speed but for consistency I left the test configuration the same as when testing the powerline adapters.
There is no substitute for good old wired Ethernet if you want the best performance possible. Obviously we can't always put network cabling where we would like it and these powerline adapters offer an easy alternative. In spite of their poor performance they should be fine for general web surfing, email, social media, etc... But I would I would not rely on these for streaming video or other bandwidth hungry applications.
I decided not to rate 1 star because they do provide some functionality. However the blatant misrepresentation of the actual performance these devices provide makes me rethink buying TP-Link products in the future. I understand that TP-Link realize there is little chance any consumer will be able to test the performance of these powerline adapters so there's little risk in exaggerating the marketing claims. Make whatever assumptions you like with the factual data I have provided in this review.