Comparing the speed of a Devolo / TP-Link homeplug versus a RJ-45 cable

Procrastination can sometimes be measured in years.

In this particular case, 15 years passed between asking the electrician to run a CAT-5 RJ-45 cable to the first floor when building our house and today when I actually crimped the cable with ethernet plugs and made use of it. And boy, the speed difference is enormous !

There were of course a few years that I could not use the cable, due to the room being used for other things than intended. But the last five years or so I have used homeplug appliances like devolo and tp-link instead of connecting the network cable. The speed was okay, but not great, and I sometimes had network interrupts or slowdowns that I suspected originated from the homeplugs.

This weekend I finally bought the necessary tools to crimp the cable and perform a few speed tests. I used dslreports.com with my Safari browser, as this is an HTML5 test. Most of the other tests still seem to require flash, which I no longer wish to use.

Here is the short and sweet result of the speed test when using the homeplugs – you can get more results when clicking on the image :

And here is the result when I switched to using ethernet cable :

As you can see, the download speed of the connection is roughly 6 times faster by cable than via the electrical home network while the upload speed is only slightly higher. I redid the test with the cable using the Chrome browser to make sure that there was no caching from previous tests, and got just about the same results.

Wow. That’s a very, very big difference in download speed… I do wonder why the upload speed did not change significantly ?

Conclusion : while the homeplugs are certainly a viable solution for many houses where ethernet cable is not usable (old house, renting,…) you are certainly better of speed-wise in using a RJ-45 cat5 ethernet cable.

(Visited 597 times, 1 visits today)

One thought on “Comparing the speed of a Devolo / TP-Link homeplug versus a RJ-45 cable”

  1. You’ve run a test against the web. Have you also tested transfer across two clients on the same (dlan, ethernet, whatever) network? I suspect that would have provided cleaner outcome.

Leave a Reply

Your email address will not be published. Required fields are marked *