Powerline Unit Performance

[#1]

After having problems using Powerline Units I have been testing a pair of Devolo dLAN 500 units with different house wiring configurations. In my case I resolved my issue by installing Ethernet cabling but as this is a topic that crops up from time to time on the forum thought that it might be worth publishing my findings. The connection was monitored first using the Devolo Cockpit software and then a speed test was conducted using the Think Broadband speedtester. I must stress that this is one set of tests and is specific to my property. Others might well get totally different results.

My conclusion is that these appear to give satisfactory results providing that the Powerline Units are connected on the same circuit or if on different circuits, the two circuits are not divided from each other via a current balance earth leakage detector.

If you still have an old style ‘Fuse Box’ with wired fuses then performance should be satisfactory, but has not been tested by me.

How do you tell if you may experience poor connectivity?

The easiest way to find out is to use an appliance such as a table light and isolate the circuit by switching off all the other circuits in the Consumer Unit.

First plug the appliance in where your Router is located as this is where the master unit will be installed. Next go to where you intend to plug the slave Powerline unit and plug your appliance in there. If it works then you are on the same circuit and these tests show that you should get acceptable performance.

What if the Powerline devices are on different circuits?

If the first test fails to show power is available at both the router and the second location it is then a case of switching another MCB on one at a time until the one controlling the power to the second Powerline location is identified. Having The following picture of a split circuit Consumer Unit typical configuration should help clarify the following table which gives the results of my tests, which is one set of results in a typical house.

MCBs that need to be switched on	Configuration	Powerline Performance
		Devolo Cockpit Reported Link Speed	Think Broadband Measured Download Speed	Think Broadband Measured Upload Speed
Direct – No Powerline Unit	Direct Ethernet cable	N/A	70.1 Mb/s	18.4 Mb/s
Only one MCB in Section 1 OR 0nly one MCB in Section 2	Same ring using Powerline units	340 Mb/s	69.8 Mb/s	18.4 Mb/s
Two MCBs in Section 1 OR two MCBs in Section 2	Different rings connected via 2 MCBs	219 Mb/s	65.7 Mb/s	18.3 Mb/s
One MCB in Section 1 and One MCB in Section 2	Different rings connected via 2 RCDs and 2 MCBs.	105 Mb/s	33 Mb/s	3.9 Mb/s

If you have a high security Consumer Unit that uses RCBOs then if more than one has to be switched on then the results are similar to those shown as one MCB in Section 1 and one MCB in Section 2

MCBs that need to be switched on	Configuration	Powerline Performance
		Devolo Cockpit Reported Link Speed	Think Broadband Measured Download Speed	Think Broadband Measured Upload Speed
Only one RCBO	Same ring using Powerline units	340 Mb/s	69.8 Mb/s	18.4 Mb/s
Two RCBOs	Different rings connected via 2 RCBOs.	105 Mb/s	33 Mb/s	3.9 Mb/s

[#2]

Because your download speed is never going to exceed 70Mbits/sec, it would be worth measuring the speed of two PC's each with 1,000 Mbit/sec ethernet using a common protocol such as FTP etc.

I bet those "Cockpit" figures are fictional - unless you have just completely rewired your house only last week.

The speed depends on wiring quality and length.

There's something wrong with the last set of figures:-

Manages to Download 33 Mb/s - but can only upload 3.9 Mb/s  - surely you'd expect a symmetric performance?

[#3]

---

Thank's for taking an interest.

@VileReynard wrote:

I bet those "Cockpit" figures are fictional - unless you have just completely rewired your house 

Nope, the wiring has been there for thirty plus years. I agree that the cockpit results are not showing reality. It makes me wonder if part of the problem is an error detection and correction overhead that's not obvious.

There's something wrong with the last set of figures:-

Manages to Download 33 Mb/s - but can only upload 3.9 Mb/s  - surely you'd expect a symmetric performance?

---

I can't answer that one. In the past this was the arrangement I was using and was getting some buffering issues when streaming HD video, which should of been well within the capability of these results, but on reflection I think that was with 200Mbps dLAN devices. Either way, this is the killer combination that convinced me that Ethernet cabling is cheaper, totally reliable and the way to go, despite the bit of effort to install it (which only has to be once).

[#4]

Ethernet is hugely cheaper too!

[#5]

Sadly that's not the case for me. Due to the double height nature of my lounge room I have no choice but to use PowerLine Adapters (these in fact) as the ceiling extends in to what is the attic (I live in a bungalow) and there is no way to drop wires down the wall where the TV, etc are as the ceiling gets in the way. And the builder in all their wisdom didn't think to install a draw string for any future occasion that might arise.

Having said that I can and have, only last night, watched 4K streaming (Series 3 of The Grand Tour) on my TV using these adapters so I've no real complaints.

[#6]

A bungalow with an attic?

Sounds like you could convert that into a large house (without an attic).

[#7]

Yes. The attic being the space between the ceiling and the roof, you know, the pointy bit with the tiles on it!

What would you call it @VileReynard?

[#8]

Well, I expect the ceiling of a bungalow or house to be flat - the attic is where you store your junk and not part of your living room.

My house has flat ceilings and the attic has some rubbish & water tanks in it.

[#9]

You are right @VileReynard and in the rest of the house the ceilings are flat but for some (design) reason in the lounge it extends into the roof space with Velux windows on each side of the roof. The house is only 10 y.o this year so it may have been 'on trend' when it was constructed.

[#10]

Just in case you're still not convinced @VileReynard here's a picture.

[#11]

I didn't disbelieve you - honestly!

Seems like you have a kind of dormer window, to let in extra light, without increasing the height of the house?

[#12]

They're not dormer @VileReynard as their flush against the roof, just run of the mill Velux roof windows. Great in the summer, but the actual space they are in is a waste of warm air in the winter. In order to get to the sockets where the TV is (left side of picture) someone would need to climb over this (as it has a flat ceiling) to get to the other side to install the wires and so far no one is brave enough.

[#13]

@Anonymous - I've got my ethernet cable fastened with clips nailed to the skirting board. I expect new houses have either no skirting boards or MDF skirting boards, making it a non-starter for you.

[#14]

I have proper wooden skirting so I guess it’s one option, but my OCD wouldn’t allow me to do it as it’s not a clear run from the switch to the lounge. I have door frames and glass partitions to get past, and I’d be driven to distraction, if not insane, if it wasn’t perfect. So I’ll live with the adapters for now as it works for me.

[#15]

The option of running the Ethernet cable outside should not be ignored. Part of my routing goes through the cavity wall and up the back of a drainpipe into the house loft. There is no way that SWMBO would allow cable to be nailed to skirting boards.