Extending wifi

Takedap

fred246 said:

It seems daft constantly persuading everyone that powerline adapters are no good when they obviously work very well for a large proportion of people. Buying a couple to try isn't an enormous expense and you can sell them if they don't work for you. Get all the fixed stuff on powerline and keep the wifi on the mesh for mobile devices. Works for me anyway.

I am currently using these & they DON'T work very well for me.   I am the OP.   That is why I asked the question.

fred246

I am talking about the ethernet police who tell everyone they have to put wires everywhere and nothing else works. You will end up with people thinking that power line adapters aren't worth trying.

neilmcl

Undervalued said:

Mickey666 said:

You probably don't have an internet connection as fast as the OP.

The TP-Link power line adaptors I mentioned (in my post half way down the first page) were significantly faster than my WiFi ever achieved even when the computer was in the same room as the router! Providing they were on the same ring main.

The OP, in his reply, said that his were on the same ring main but gave disappointing results, which I found surprising.

At the time I had 100 MB cable BB (which speed tested at about 105 with a wired connection to the supplied router. The best I ever got on 2.4 GHz WiFi with that router was around 50 MBs and often far less. The TP link units gave a reliable 70+ anywhere on that ring main.

Agreed, there's definitely something amiss if speeds drop from 100Mbps to just 10Mbps when using a powerline adapter.

CoastingHatbox

fred246 said:

I am talking about the ethernet police who tell everyone they have to put wires everywhere and nothing else works. You will end up with people thinking that power line adapters aren't worth trying.

They are a bit of a mixed bag. And, a bit of a security risk. Models that force encryption out of the box are a security risk (...have you paired yours?)

Takedap

neilmcl said:

Undervalued said:

Mickey666 said:

You probably don't have an internet connection as fast as the OP.

The TP-Link power line adaptors I mentioned (in my post half way down the first page) were significantly faster than my WiFi ever achieved even when the computer was in the same room as the router! Providing they were on the same ring main.

The OP, in his reply, said that his were on the same ring main but gave disappointing results, which I found surprising.

At the time I had 100 MB cable BB (which speed tested at about 105 with a wired connection to the supplied router. The best I ever got on 2.4 GHz WiFi with that router was around 50 MBs and often far less. The TP link units gave a reliable 70+ anywhere on that ring main.

Agreed, there's definitely something amiss if speeds drop from 100Mbps to just 10Mbps when using a powerline adapter.

I'm surprised too.  But that's what is happening.   Since first posting, I've done a bit of experimenting with the position of the router & the orientation of the antennae & have now got around 20 Mbs on WiFi at the furthest point.  I can probably live with this as the PC at the far end is used for more office type work rather than streaming.  Thanks for all the suggestions though.

Mickey666

fred246 said:

I am talking about the ethernet police who tell everyone they have to put wires everywhere and nothing else works. You will end up with people thinking that power line adapters aren't worth trying.

For the OP's situation, which is what we are supposed to be helping with in this thread, powerline adapters ARE NOT worth trying because they WILL result in a drop in speed . . . as the OP has already confirmed!

You seem to be taking my comments too personally, when they are not even aimed at you   Of course, power line adapters work, that is not in dispute, the issue is how they PERFORM and they WILL NOT perform as well as cheap Ethernet cable.  Whether or not that matters to someone is up to them.  It clearly doesn't matter to you but it clearly does matter to the OP, which is why I mentioned the Ethernet cable and access point solution.

Mickey666

Takedap said:

neilmcl said:

Undervalued said:

Mickey666 said:

You probably don't have an internet connection as fast as the OP.

The TP-Link power line adaptors I mentioned (in my post half way down the first page) were significantly faster than my WiFi ever achieved even when the computer was in the same room as the router! Providing they were on the same ring main.

The OP, in his reply, said that his were on the same ring main but gave disappointing results, which I found surprising.

At the time I had 100 MB cable BB (which speed tested at about 105 with a wired connection to the supplied router. The best I ever got on 2.4 GHz WiFi with that router was around 50 MBs and often far less. The TP link units gave a reliable 70+ anywhere on that ring main.

Agreed, there's definitely something amiss if speeds drop from 100Mbps to just 10Mbps when using a powerline adapter.

I'm surprised too.  But that's what is happening.   Since first posting, I've done a bit of experimenting with the position of the router & the orientation of the antennae & have now got around 20 Mbs on WiFi at the furthest point.  I can probably live with this as the PC at the far end is used for more office type work rather than streaming.  Thanks for all the suggestions though.

It's likely to be a combination of the powerline AND wifi.  Powerline adapters come in two 'flavours' - an ethernet-to-ethernet link or an ethernet-to-wifi link.  Sometimes both.  Both types of connection sill slow down if/when there is interference and distance increases susceptibility and the effect is reduced speed due to data packets having to be retransmitted more often before an uncorrupted data packet gets through.
Thus, imagine an ethernet-to-ethernet connection running over powerline units.  Let's assume it a good connection and only loses 10% speed.  This is likely to be unnoticed by all but the most demanding applications.  Fine.  Now imagine an ethernet-to-wifi connection running over powerline units and let's assume the wifi part of the link also loses 10% speed.  In this case the wifi is losing 10% of the 90% powerline link speed, ie the overall link speed will be 90% or 90% = 81%.  Again, this may or may not be noticeable, depending on the user.  But if the two link losses were 50% then the overall speed would drop to just 25% of possible maximum (ie 50% of 50%).
Computer networking is a powerful technology and at the consumer level has been amazingly well engineered to become plug-and-play with little or no knowledge required about what is actually happening 'behind the scenes' - which is generally a good thing.  But it is one thing to just 'plug-and-play' and quite another to optimise network throughput performance.

[Deleted User]

Mickey666 said:

Thus, imagine an ethernet-to-ethernet connection running over powerline units.  Let's assume it a good connection and only loses 10% speed.  This is likely to be unnoticed by all but the most demanding applications.  Fine.  Now imagine an ethernet-to-wifi connection running over powerline units and let's assume the wifi part of the link also loses 10% speed.  In this case the wifi is losing 10% of the 90% powerline link speed, ie the overall link speed will be 90% or 90% = 81%.  Again, this may or may not be noticeable, depending on the user.  But if the two link losses were 50% then the overall speed would drop to just 25% of possible maximum (ie 50% of 50%).

It doesn't really work like this, each link in the network has a maximum throughput and the link in the chain with the lowest throughput is the limiting factor in the maximum bandwidth.

If the powerlines are rated at 100mbps but only perform at 50mbps (50% loss) and the wifi is rated 100mpbs but only performs at 50mbps (50% loss) then the maximum throughput is still 50mbps - overall loss still 50% not 25% as you say in your example.

And if you are only trying to transmit 25mpbs then the total loss is 0% because you are below the maximum throughput of any link in the chain.

FaceHead

Ethernet cabling will provide the best performance, reliability and longevity. However, the OP has already suggested that chasing walls and ripping up floorboards is unacceptable, which isn't an unreasonable position. This is when people start to look at powerline, WiFi boosters and the like.

Can ethernet cabling be run outside the house? It really depends on what in the layout are external walls. Most cables are weather protected and designed for this.

From a faceplate near the router, cable can go straight outside, avoiding chasing that wall. Outside it may be able go around the house (potential below ground level, or discreetly clipped to the wall) and back through the wall, straight into the second faceplate. For minimum disruption the two faceplates can be on pattress boxes, but ideally they are mounted flush by sinking backboxes into the wall. 

Mickey666

[Deleted User] said:

Mickey666 said:

Thus, imagine an ethernet-to-ethernet connection running over powerline units.  Let's assume it a good connection and only loses 10% speed.  This is likely to be unnoticed by all but the most demanding applications.  Fine.  Now imagine an ethernet-to-wifi connection running over powerline units and let's assume the wifi part of the link also loses 10% speed.  In this case the wifi is losing 10% of the 90% powerline link speed, ie the overall link speed will be 90% or 90% = 81%.  Again, this may or may not be noticeable, depending on the user.  But if the two link losses were 50% then the overall speed would drop to just 25% of possible maximum (ie 50% of 50%).

It doesn't really work like this, each link in the network has a maximum throughput and the link in the chain with the lowest throughput is the limiting factor in the maximum bandwidth.

If the powerlines are rated at 100mbps but only perform at 50mbps (50% loss) and the wifi is rated 100mpbs but only performs at 50mbps (50% loss) then the maximum throughput is still 50mbps - overall loss still 50% not 25% as you say in your example.

And if you are only trying to transmit 25mpbs then the total loss is 0% because you are below the maximum throughput of any link in the chain.

I think you are confusing data speed (mbps) with data throughput.  Data errors in the second segment of the chain result in error messages having to be sent back to the source of the message, meaning it has to go back through BOTH segments before the source then re-transmits the packet through BOTH segments again.  This creates a lot of additional network traffic.  The segment SPEEDS remain the same but the data throughput falls because of all the extra data that needs to be sent before an uncorrupted packet makes it through the overall link.