Powerline performance
Performance, measured using a single-direction test with the open source iperf utility, yielded a ten-run average TCP throughput of 172Mb/s, a little lower than other 500Mb/s - the aggregate over-the-mains bandwidth, don't forget, not what you'll see at the Ethernet port; connected computers just present the RJ-45's details - adaptors I've tried, but then the test conditions weren't exactly the same.
Just to give you a guide, I've included test results from past 85Mb/s and 200Mb/s Devolo powerline adaptors, so you can see the improvement doubling the aggregate bandwidth makes.
iperf Test Results
Single-direction data transfer rate in Megabits per Second (Mb/s)
Longer bars are better
Running iperf in UDP mode got me up to 243Mb/s before packet loss nudged up to one per cent.
Both sets of adaptors provide the now-standard 128-bit AES encryption to secure the transmission between adaptors, controlled with a passphrase. Each adaptor can generate its own: just press the wee button by the Ethernet port and make sure you then press it on all the adaptors you want the new passphrase to share.
Devolo's bundled Cockpit app, available for Windows, Mac OS X and Linux, allows you to set the passphrase manually, adding the codeword to any or all of the adaptors the software can see on your network. You'll need the security code printed on each adaptor for that.
Devolo's Adobe Air-based Cockpit app provides only rudimentary device management facilities
It's pretty software, but vacant. It doesn't do much else, not even handling firmware updates. Click the Updates button and you're simply taken to Devolo's website. Firmware updates are available, but Devolo only provides a Windows-based uploader.
But if the software fails to impress, at least the hardware will. If you're already using 200Mb/s adaptors it's hard to justify an upgrade unless you're flinging rather a lot of HD streams around. After the switch from my regular 200Mb/s units to the 500Mb/s Devolos, the missus and the boy noticed no change or improvement to their regular, respectively, BBC iPlayer and Xbox Live sessions. Modern online gaming systems have enough latency compensation to deal with any the powerline adaptors add.
Then again, that 200Mb/s - 63Mb/s in the real world - is all you have. The more things you have taking a slice of that bandwidth, the quicker you'll hit the limit. Upgrading to 500Mb/s adaptors and getting 172Mb/s gives you a lot more headroom.
Tastefully styled
If you're struggling with Wi-Fi dead spots, powerline offers a way into them. So does Cat 5 cable, but not all of us want to got to the expense of getting it plumbed in, or fancy having it tacked around the skirting boards. That may not even be possible if you rent your property, or are planning to move in the not-too-distant future.
Verdict
A decent, well-performing set of adaptors, then. But they don't set themselves apart from their rivals. While physically thinner than, say, the Netgear XAVB5101 adaptor I looked at in April, the Netgear's other dimensions are smaller. And by the time you've added in the thickness of the three-pin plug mount, the dLan doesn't hug the wall much more than the Netgear does. There's barely any difference in the price, either.
Indeed, if you don't care much about an adaptor's size, you can buy lesser known brands of powerline kit for two-thirds of the price from the likes of Solwise. It has pass-through power options too. ®
More Networking Gear Reviews |
||||
 D-Link DHP-1565 |
 Netgear Nano 500 |
 HD JuiceBox
HDMI over Powerline |
 Ten... dual-band
wireless routers |
 Ten... Mi-Fi HSPA 3G
wireless mini-routers |
Devolo dLAN 500Mb/s powerline network adaptor review
COMMENTS
Tests?
Comparative tests with 2.4GHz and 5.8GHz a/b/g/n?
Comparison with wired Cat5e via 1Gbps Switch?
Effect of all the light sockets with cheap CFL lamps?
Spectrum analyser measuring EMC?
Effect of 5 x PC / Laptop cheap PSUs
Effect of cheap lightweight phone charger and flat phone on same socket
Effect of Filtered extension strip/non-filtered to connect it to a skirting board
Effect of plugging both ends and gear at both ends into 2 x UPS
Or is this just Shiny Gadget Advertising?
Re: Cat5 and Gigabit
you want at least Cat6 for a stable gigabit network. Just like I have here at home, in fact.
Overkill. Gigabit was specified for cat-5. Experience with cat-5 in the field suggested insufficient margin to tolerate all the kinks and scrapes that the cable plant got subjected to in the field, and cat-5e was born. I doubt you can still purchase non-e plain cat-5. Cat-5e is extremely widely used for Gbit ethernet (such as everywhere in my workplace) and has no stability issues at all provided you respect the distance limits (max 94m of fixed wiring with max 6m total of patch cable at the ends, no joins in the middle. If the cable run in the middle is considerably shorter, as it usually is, you can take some liberties with longer patch cables at the ends).
Re: 500Mbps?
500Mbps might be overkill if you're only using the adapter to link to the outside world, but once you start throwing files between devices on your local network you'll find yourself wanting as much speed as you can get your hands on...
Re: Tests?
Strewth, Mage, if you're so anal you need that level of testing before buying a 60 quid doohickey, you can buy 'em yourself and post your own results. Me, I don't need all that nonsense here.
Even 200meg powerline is better than 5GHz Wi-Fi. No one I know has Cat 5 and a gigabit switch at home, even the guy I know who does the IT at a major Japanese bank. Ditto UPS. The review talks about using power gangs. Powerline folks have been saying for years you shouldn't use surge protectors.
FWIW, I use this stuff and I get decent - well, good enough - speeds with all sorts of stuff plugged in around it. No fluoro lights, though; don't have any. It all lives behind the TV, and I'm not reaching down there every time I need to charge my phone. Who would?
