Powerline Adapter - real world usage

funkey_monkey

Reading up on powerline adapters - there doesn't seem much in actual performance between the top ones and the middle of road ones.

This is a review for the TL8010P (which I think is the one Clareman has):

https://www.techadvisor.co.uk/review...eview-3610443/

Quote:
We do two tests on Powerline. One is not very real-world but is tested in a house. We put both Powerlines on the same wall power socket and transfer a large file from one computer to another.

On this test the TP-Link AV1200 equalled our previous top speed of 410Mbps, by transferring a 1GB file between PC and laptop in 20 seconds.

But a more realistic test is with the second Powerline adapter a couple of floors away from the first one that’s connected to the router.

In this tougher test the TP-Link AV1200 managed just over 100Mbps, which is only slightly slower than the other Gigabit Powerlines we have tested: the uglier but cheaper Solwise SmartLink 1200AV2 (107Mbps) and larger but more costly Devolo 1200 (104Mbps).

This is the new one from TP-Link:

https://www.techadvisor.co.uk/review...eview-3643494/

Quote:
In our first Room Test we check speeds when both adapters are in the same room – which is not how you use Powerline! We use this to test its top speed and The AV2000 did very well, scoring 432Mbps by transferring a 1GB file in 19 seconds.

When we positioned the second adapter in a room two floors down and about 30 metres apart speeds naturally dropped. This time the AV2000 scored 117Mbps way below the 2,000Mbps on the box but still the fastest we’ve seen from a Powerline.

Quote:
In general the 500Mbps Powerlines are faster than the 200Mbps, and the 1,200Mbps adapters are faster than the 500Mbps – so use the speeds as a gauge of speed between models.
Not sure how they can say this as the AV1200 is not ~40% slower than the AV2000 - unless I'm missing something?

So to me it seems that there is no point in opting for the more expensive AV2000 as in the real word the difference is not startling between both items.

Comparison:
https://www.tp-link.com/uk/compare/?...%2C5788%2C5790

TL-PA8010P = £54 (Amazon)
TL-PA8033P = £68 (Amazon)
TL-PA9020P = £103 (Amazon)


Is there any point even considering the PA9020P? Specs look good, but the real world test says otherwise.

I'll be using these for a calving camera and maybe some additional security cameras (note the link for sheds to the NVR will be okay - the NVR to internet may have blips, but any footage should be recorded onto NVR).
Calving camera may have dropouts, but I'll just have to deal with that until I figure out how to get the cabling into attic.

Is there no shielded network cabling on the market that would allow it in same conduit as power cable?

smuggler.ie

I suggest you put that £100 toward proper CAT link installation cost.

As everything, powerline adapters would be spec'ed at ideal environment hence "up to". Among other factors effecting transmission speed will be: distance(actual power cable) from point A to point B, connection quality on every power outlet(socket) and noise generated by other electrical equipment used on power circuit(phone charger, TV, water/shower pump, etc.).

Reviews written by consumers might give you general indication, however, reflect that specific environment device is in use.

On other hand, your camera NIC most likely will be 100Mbps anyway and will require just fraction of that to operate, will depend on camera and settings. https://www.cctvcalculator.net/en/calculations/bandwidth-calculator/
Amazon have(had) good return policy if willing to do real life testing.

Johnboy1951

Is there no shielded network cabling on the market that would allow it in same conduit as power cable?

I would expect shielded cable to be fine if the shield is correctly earthed. Of course if there is machinery switching at the remote location then a lot of noise could be generated which can impact the data cable.
How long is the run?
You could possibly run outdoor shielded CAT5e/6 cable?

EDIT:
For the cost of the cable I suggest it might be worthwhile to try running it with the mains cable.

.

funkey_monkey

Johnboy1951 wrote: »

EDIT:
For the cost of the cable I suggest it might be worthwhile to try running it with the mains cable.

.

It is against code to run them in the same trunking. They can induce a voltage and blow whatever is connected. I can't do this.

funkey_monkey

smuggler.ie wrote: »

I suggest you put that £100 toward proper CAT link installation cost.

I would prefer them to do that but they don't want to chase the walls. Unless you can think of another way to get wired connection into attic.

smuggler.ie wrote: »

On other hand, your camera NIC most likely will be 100Mbps anyway and will require just fraction of that to operate, will depend on camera and settings. https://www.cctvcalculator.net/en/calculations/bandwidth-calculator/
Amazon have(had) good return policy if willing to do real life testing.

I have already checked out the demand and you're right that 100Mbps will cover most of it.
However, the demands might be lower as all installed security cameras and calving camera will eventually feed into an nvr in the attic. The calving camera will be a live feed only. So nvr will eventually be only connection on the powerline.

smuggler.ie

funkey_monkey wrote: »

I would prefer them to do that but they don't want to chase the walls. Unless you can think of another way to get wired connection into attic.

Without knowing all details hard to suggest, what is current network and building layout?
IIRC you have CAT link to the shed dug-in(or had intention) so i guess this will feed your camera(s).
Also you have switch mounted on the wall in some room(is it dedicated COM's room?). Could this be option to avoid breaking walls? Mount in the corner to cover up run to the ceiling.

If rooms adjacent to external walls have Ethernet feed, you could drill out and run external CAT outside up to the soffit ant then to the attic

Johnboy1951

funkey_monkey wrote: »

It is against code to run them in the same trunking. They can induce a voltage and blow whatever is connected. I can't do this.

My difficulty is in understanding how this can happen to a properly screened and earthed Cat cable.

funkey_monkey

Johnboy1951 wrote: »

My difficulty is in understanding how this can happen to a properly screened and earthed Cat cable.

I'm not sure either but I'm not going to go against the regs in a relations house.
Would be so simple to do as the router is right beside the socket faceplate. Annoying to say the least!

funkey_monkey

smuggler.ie wrote: »

Without knowing all details hard to suggest, what is current network and building layout?
IIRC you have CAT link to the shed dug-in(or had intention) so i guess this will feed your camera(s).
Also you have switch mounted on the wall in some room(is it dedicated COM's room?). Could this be option to avoid breaking walls? Mount in the corner to cover up run to the ceiling.

If rooms adjacent to external walls have Ethernet feed, you could drill out and run external CAT outside up to the soffit ant then to the attic

CAT link to shed to be dug in and routed up exterior wall into attic.

The comms cupboard is in another house!
This property is a bungalow with entrance in centre, living are on one side and bedrooms on other. The router is located just inside front hallway beside porch door. The floor is laminate - so I can't route CAT under carpet to a suitable location to rise into attic. The hallway also has a deep coving around the ceiling which makes conduit difficult. Plus, they have recently decorated the hallway (wallpaper) so they are hesitant about adding trunking at this stage.

I think I'll try the powerline first and if it doesn't give suitable performance/stability, then I'll make them decide about installing some trunking.

human 19

I'm not sure of your layout here but don't powerline adapters have to be on the same electrical network? is this the case here? Maybe buy a couple of cheapos just to confirm they can connect?

I used to use powerline adapters in the house but switched to netgear Wi-Fi extenders. would that be an option? I use these to stream Wi-Fi throughout the house. If I'm in the garden I plug one in near the window

andublin

funkey_monkey wrote: »

CAT link to shed to be dug in and routed up exterior wall into attic.

The comms cupboard is in another house!
This property is a bungalow with entrance in centre, living are on one side and bedrooms on other. The router is located just inside front hallway beside porch door. The floor is laminate - so I can't route CAT under carpet to a suitable location to rise into attic. The hallway also has a deep coving around the ceiling which makes conduit difficult. Plus, they have recently decorated the hallway (wallpaper) so they are hesitant about adding trunking at this stage.

I think I'll try the powerline first and if it doesn't give suitable performance/stability, then I'll make them decide about installing some trunking.

Just a thought:
I have three 4K cameras, powered from POE switch, going over two tplink av600 power line adapters, to an nvr. The adapters report about 130Mb/sec being sent, if that helps your calculations. Compression being used by the cameras is h.265; I believe h.264 would need a lot more bandwidth.

I didn’t expect the power line adapters to be good enough. They were only going to be a trial measure as I had them lying around, and was surprised they worked well enough. I still plan on replacing them with cable though not urgently; nothing compares for reliability, speed, throughput, and ease of troubleshooting.

By the way, the Devolo Magic 2 lan power line adapters claim 2400Mb/s and are based on a different technology standard, but are very expensive. No idea if better or worse for your use case.

Keep us informed, whatever your final result!

funkey_monkey

I've bought the av1300 tp link items. Should be here middle of week. I'll try them out and let you know the performance.
The sockets are on same ring so shouldn't have to go through consumer unit.

ED E

Johnboy1951 wrote: »

My difficulty is in understanding how this can happen to a properly screened and earthed Cat cable.

When a sparks earths a chassis or device theys mount a cable on a screwdown. Users dont tend to remove these. Users unplug cables.

Johnboy1951

Quote=funkey_monkey
It is against code to run them in the same trunking. They can induce a voltage and blow whatever is connected. I can't do this.

Johnboy1951 wrote: »

My difficulty is in understanding how this can happen to a properly screened and earthed Cat cable.

ED E wrote: »

When a sparks earths a chassis or device theys mount a cable on a screwdown. Users dont tend to remove these. Users unplug cables.

I am unsure of the point you are making or how it relates.

Johnboy1951

funkey_monkey wrote:

It is against code to run them in the same trunking. They can induce a voltage and blow whatever is connected. I can't do this.

Johnboy1951 wrote: »

My difficulty is in understanding how this can happen to a properly screened and earthed Cat cable.

ED E wrote: »

When a sparks earths a chassis or device theys mount a cable on a screwdown. Users dont tend to remove these. Users unplug cables.

I am unsure of the point you are making or how it relates.

funkey_monkey

andublin wrote: »

Keep us informed, whatever your final result!

Okay - so the adapters have arrived. Much quicker than I expected so I've tried them out in my house.

Internet speed (wired into router)
Download: 49.69Mbps
Upload: 7.99Mbps
Ping: 20ms

Office (Powerline - same ring)
Download: 27.7Mbps
Upload: 7.99Mbps
Ping: 22ms

Office (WiFi)
Download: 48.63Mbps
Upload: 7.48Mbps
Ping: 22ms

Bedroom (Powerline - Different ring)
Download: 24.09Mbps
Upload: 8.04Mbps
Ping: 23ms



I used www.broadbandspeedchecker.co.uk to test the speeds. These are internet connection speeds and not the speed between each end of the home plug. Not sure how to test that as I've only 1 laptop.
I will check again when at the house where they are to be installed, but 50% efficiency should be sufficient - remembering that all cameras feed into an NVR and it will be the NVR that is connected to the network. It won't be X cameras independently requiring network access.

Camera I currently have is H2.565+ compression and any additional ones will match or better this.


Note: The fastest speeds in our area is approx 65Mbps.

smuggler.ie

This was on your network, final destination might differ, hopefully not into worse side.
For local stream should be fine, for stream online watch upload speed

funkey_monkey

smuggler.ie wrote: »

This was on your network, final destination might differ, hopefully not into worse side.
For local stream should be fine, for stream online watch upload speed

Final destination may well be worse as it is an older property. This one has a recent re-wire so is essentially all new wiring.
What upload speed would be the boundary value between usable and unusable?

Much quicker than I expected so I've tried them out in my house.

Delivery that is...

smuggler.ie

funkey_monkey wrote: »

Delivery that is...

Got this

funkey_monkey wrote: »

What upload speed would be the boundary value between usable and unusable?

Well, i guess you need steady ~5Mbps(bit less) per stream(full HD) for smooth playback or quality will drop(stutter)

Should not be issue with local if you manage to maintain ~25 , for few cam's

funkey_monkey

Finally got the powerlines installed in situ and the results are disappointing.

Connecting to host 192.xxx.x.xxx, port 5201
[  4] local 192.xxx.x.xx port 53557 connected to 192.xxx.x.xxx port 5201
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-1.01   sec  11.0 MBytes  91.5 Mbits/sec                  
[  4]   1.01-2.00   sec  10.6 MBytes  89.6 Mbits/sec                  
[  4]   2.00-3.00   sec  10.9 MBytes  91.1 Mbits/sec                  
[  4]   3.00-4.01   sec  10.1 MBytes  84.3 Mbits/sec                  
[  4]   4.01-5.01   sec  10.8 MBytes  90.7 Mbits/sec                  
[  4]   5.01-6.02   sec  12.0 MBytes  99.6 Mbits/sec                  
[  4]   6.02-7.00   sec  11.6 MBytes  99.2 Mbits/sec                  
[  4]   7.00-8.01   sec  8.62 MBytes  71.5 Mbits/sec                  
[  4]   8.01-9.01   sec  8.25 MBytes  69.4 Mbits/sec                  
[  4]   9.01-10.00  sec  8.38 MBytes  70.8 Mbits/sec                  
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-10.00  sec   102 MBytes  85.8 Mbits/sec                  sender
[  4]   0.00-10.00  sec   102 MBytes  85.7 Mbits/sec                  receiver

iperf Done.

Connecting to host 192.xxx.x.xxx, port 5201
Reverse mode, remote host 192.xxx.x.xxx is sending
[  4] local 192.xxx.x.xx port 53559 connected to 192.xxx.x.xxx port 5201
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-1.01   sec  3.06 MBytes  25.4 Mbits/sec                  
[  4]   1.01-2.00   sec  5.61 MBytes  47.5 Mbits/sec                  
[  4]   2.00-3.00   sec  5.80 MBytes  48.7 Mbits/sec                  
[  4]   3.00-4.00   sec  5.86 MBytes  49.3 Mbits/sec                  
[  4]   4.00-5.01   sec  6.54 MBytes  54.2 Mbits/sec                  
[  4]   5.01-6.00   sec  6.52 MBytes  55.3 Mbits/sec                  
[  4]   6.00-7.01   sec  9.87 MBytes  81.6 Mbits/sec                  
[  4]   7.01-8.00   sec  11.0 MBytes  93.3 Mbits/sec                  
[  4]   8.00-9.02   sec  9.24 MBytes  76.6 Mbits/sec                  
[  4]   9.02-10.00  sec  6.48 MBytes  55.0 Mbits/sec                  
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-10.00  sec  70.1 MBytes  58.8 Mbits/sec                  sender
[  4]   0.00-10.00  sec  70.1 MBytes  58.8 Mbits/sec                  receiver

iperf Done.

An internet speed test (broadbandspeedchecker.co.uk) returned a speed of 5.5Mb/s.

I think this would be sufficient for the camera, but although the speed between the powerlines would be sufficient, the internet speed test from the powerline is worrying.
Any advice.


Actually wondering if I performed the test correctly.

smuggler.ie

What is actual internet speed to the property, direct connection to router ?

funkey_monkey

I'm not there now, but I think they told me they were getting ~50Mb/s connection.

Edit - as stated above, the connection from the powerline in attic out to internet was very slow @ 5.5Mb/s. The above speeds were between the two powerlines. So, not sure how the internet connection dropped to 5.5Mb/s.

I may persevere with these as the alternative is to install a fibre optic cable up the conduit alongside the mains power.

smuggler.ie

Better you verify this. It could be 5 Mbps or it could be time of a day test taken, depending on internet connection type.
Local looks OK, you maxing 100Mbps - is this limitation on devices NIC's/cable or adapters?

funkey_monkey

Client laptop: Intel I219-LM
Server Laptop: Realtek PCIe GbE Family Controller

The cabling used was bundled with the AC1300 powerlines - I didn't specifically check, but assumed the were Gb rated.
Both laptops had their ethernet speed set to 'Auto Negotiation'.
I've changed them to '1.0 Gbps Full Duplex' - maybe that will make a difference?

Edit - Seems like Auto Neg is the correct setting, so I'll revert.

Is there any reason why the reverse test results (second set) were worse than the normal run?

Tea drinker

funkey_monkey wrote: »

Client laptop: Intel I219-LM
Server Laptop: Realtek PCIe GbE Family Controller

The cabling used was bundled with the AC1300 powerlines - I didn't specifically check, but assumed the were Gb rated.
Both laptops had their ethernet speed set to 'Auto Negotiation'. I've changed them to '1.0 Gbps Full Duplex' - maybe that will make a difference?

It could do, but you should verify the powerline adaptors are working at 1000/full also.
How many powerlines do you have connected now, do they share the spectrum somehow? What happens if you once again unplug both at the same time and plug in and retest?

funkey_monkey

I should also check the ethernet connection status when it is all set up - as that will tell me what speed they agreed upon.
Only 2 powerlines from the same family (boxed pair).

Never tried that unplugging them.

smuggler.ie

What command switches you use? Were you limiting your test?

Post the line

edit : omit personal details

funkey_monkey

Server (attic):

iperf3 -s

Client (beside router):

iperf3 -c 192.xxx.x.xx

and then

iperf3 -c 192.xxx.x.xx -R

for reverse test.

smuggler.ie

As no bandwidth limit switch used (-b XXX) and if all devices/cables GB rated it leaves drop on power-line itself i guess


Try :
iperf3 -c 192.xxx.x.xx -t 100 -u 10000


-t 100 ->> 100 seconds
-u ->> UDP (default runs on TCP)
10000 ->> port number (random)


This should show if you have packet loss on the line

funkey_monkey

Will try that today as it is my last day with the second laptop.
Anything else to try before I go?

smuggler.ie

funkey_monkey wrote: »

Will try that today as it is my last day with the second laptop.
Anything else to try before I go?

My plan would be

Have few "known good" CAT leads
Test local speed without adapters
Check link speed to each adapter on NIC
Test local speed with adapters on short distance(same room)
Test Internet speed on router(Find out ISP, connection type)


As for speed difference on iPerf, try swap iperf3 -s to another device, repeat test - could be NIC specs per device related.

and check this out:
https://www.tp-link.com/uk/support/faq/2928/


Fiber link would be nice bit of investment