TP-Link PA9020P kit giving REDUCED internet speeds after I upgraded from Vivid 200 to Vivid 350!

[phillipwj]

Model : TL-PA9020P Kit
Hardware Version : V1
Firmware Version : 32.4_902111_170628_901
ISP : Virgin Media Vivid 350Mbos

Hi all,

I've been using a TP-Link TL-PA9020P ethernet powerline adapter kit very successfully with Vivid 200 for many months, always achieving very near to or in excess of 200+Mbps internet speed to my Mac Pro, which was great.

However, we have recently upgraded to Vivid 350 and my ethernet Powerline internet speed has actually DECREASED, now never achieving more than 160-170Mbps!

The reported line speed with TP Utility is and has always been c1350Mbps. I realise this is the line speed not the throughput speed, but TP Link have themselves corroborate this with me, that this fast line speed should certainly be able to cope with 350Mbps throughput.

Also using iperf, networking-wise we are seeing c.318Mpbs through the powerline. For internet speeds, when we had Vivid 200 I would regularly get up to 230Mbps through the power line, now we have moved to Vivid 350 it's steadfastly reduced. to c170Mbps

So I contacted TP Link support and reported this rather odd throughput speed decrease via the powerline despite the almost 50% speed increase to the router (confirmed with a test Direct connection and also with a test connection to the spare ethernet socket on Powerline Adapter 1, i.e. the adapter is receiving the full speed also, before is goes through the powerline).

After several days of to'ing and fro'ing with TP-Link support the issue was escalated to their R&D programmers for advice and this is what came back from them:

*************
Sorry for the reply late, I have discussed your issue with our R&D. The communication frequency of Coaxial cable is 5Mhz~42Mhz and the communication frequency of Powerline adapter is 1.8Mhz~68Mhz. So there will be some influence between them, the higher speed of Coaxial cable, the more affected on powerline. So we think your issue may due to this reason.
*************

So they are basically stating that the higher speed Vivid 350 is likely causing more interference than the Vivid 200, to the extent that it has actually reduced the throughout to a LOWER rate than Vivid 200! Wonderful!

Can anybody help to clarify and/or corroborate this claim? i.e. the communication frequency of 200 vs 350 service and affect on powerline adapters?

Also, I did extensive testing across the network using iperf 2 to try and diagnose the issue, so I will post below in the hope that someone can interpret them usefully:

Up until a couple of days ago our network topology was as follows:

ISP(Virgin)---Virgin Hub3---TL-PA9020P(No1.)~~~TL-PA9020P(No2.) —PC2 (Mac Pro Desktop)

In an effort to reduce interference I separated the Hub3 further from the Powerline by adding in the gigabit switch, so it is now:

ISP(Virgin)---Virgin Hub3---Netgear Switch(Gigabit)---TL-PA9020P(No1.)~~~TL-PA9020P(No2.) —Mac Pro Desktop (PC2)

In the following iperf tests:

PC1 is my Laptop which I can move around

PC2 is my Mac Pro Desktop setup which is anything but portable and cannot be moved around and this is my standard setup, i.e. internet coming through: Virgin Hub3---Netgear Switch(Gigabit)---TL-PA9020P(No1.)~~~TL-PA9020P(No2.) to my Mac Pro.

Iperf tests I carried out below to check NETWORK traffic speeds as follows, at no point did I switch off the Hub3 or disconnect it from the gigabit switch. I simply left it in place and did the tests around it. Obviously ensuring the IP addresses on each computer remained the same while testing.

A: PC1 (Laptop)---Netgear Switch(Gigabit)---TL-PA9020P(No2.)—PC2 (Mac Pro Desktop) >>>> 938Mbps / 941Mbps (i.e. not going across powerline network, just testing the switches on the powerline adapter and the netgear switch)

B: PC1 (Laptop)—TL-PA9020P(No1.)~~~TL-PA9020P(No2.)---PC2 (Mac Pro Desktop) : >>>>>> 318Mbps.

C: PC1 (Laptop)---Netgear Switch(Gigabit)---TL-PA9020P(No1.)~~~TL-PA9020P(No2.)—PC2 (Mac Pro Desktop) >>>>> 344 / 362 / 372 Mbps

D: PC1 (Laptop)---(LAN)Virgin Hub3(LAN)---TL-PA9020P(No1.)~~~TL-PA9020P(No2.)—PC2 (Mac Pro Desktop) >>>>>> 295 / 311/ 302 Mbps

E: PC1 (Laptop)---(LAN)Virgin Hub3(LAN)---Netgear Switch(Gigabit)---TL-PA9020P(No1.)~~~TL-PA9020P(No2.)—PC2 (Mac Pro Desktop) >>>>>> 298 / 311 / 321 Mbps

We also know that the coaxial cables are delivering INTERNET traffic to the Hub3 at full speed 350-385Mbs since I connected the laptop direct to hub with Cat6 cable and as expected this gave me immediate full speed of c350-380Mbps.

We also know that INTERNET traffic is reaching PowerLine Adapter No1 at full speed of 350-385Mbos as I just tested that again this morning as follows:

ISP(Virgin)---Virgin Hub3---Netgear Switch(Gigabit)---TL-PA9020P(No1.) Ethernet Port 1 ~~~TL-PA9020P(No1.) Ethernet Port 2 —Laptop (i.e. not going across the powerline network, just into one of the powerline adapters ethernet ports and out the other one).

We know that NON-INTERNET NETWORK traffic can get through the 4 port switch on the Hub3 at speeds between 295-321Mbps as we’ve tested the speed of that switch.

We know that NON-INTERNET NETWORK traffic can get through the gigabit switch at essentially full gigabit speeds 941Mbps

We know that NON-INTERNET NETWORK traffic can get through the PowerLine network at speeds ranging between 344-372Mbps.

We know that INTERNET traffic can only get through the PowerLine network at c170Mbps, i.e. half the speed of the other network traffic.

So, considering all the above, does anyone have any ideas as to what is causing the 50% reduction in internet download speed even when we know we are getting full speed to the router?

Many thanks for any pointers.

 

[ColinTaylor]

Very strange. Especially as you're using exactly the same modem as before, only the provisioned speed is different.

The "5Mhz~42Mhz" is in reference to the DOCSIS upstream frequency range of 5MHz to 42MHz. In fact that's the American frequencies. Virgin Media use EuroDOCSIS which operates on 5MHz to 65MHz.

So now that you have been provisioned with more upstream bandwidth you're probably using more upstream channels and those channels may now be on different frequencies. So TP-Link are suggesting that the cable modem is causing RF interference with the powerline system.*

* Log into the VM Hub and make a note of what upstream channels it's using. Then try powering it off, waiting 5 minutes and then turning it back on. Check the channel usage again. Sometimes it will use different channels. If it has see if that has effected your powerline throughput.

So if TP-Link's theory about RF interference is correct then all you can really do is make sure that the coaxial cable into the VM Hub doesn't go anywhere near the powerline adaptors or your mains wiring. So don't lie the cables over each other, don't put the coax cable near the mains wall sockets, and don't run it along a wall that has a ring main behind it.

Also, at the risk of stating the obvious, always plug the powerline adapters directly into the wall sockets and not into an extension strip.

 

[thiggins]

Did you change modems with the service speed upgrade?

How far away did you move the powerline adapter from the Virgin Hub? Why did you add the switch?

 

[phillipwj]

Very strange. Especially as you're using exactly the same modem as before, only the provisioned speed is different.

The "5Mhz~42Mhz" is in reference to the DOCSIS upstream frequency range of 5MHz to 42MHz. In fact that's the American frequencies. Virgin Media use EuroDOCSIS which operates on 5MHz to 65MHz.

So now that you have been provisioned with more upstream bandwidth you're probably using more upstream channels and those channels may now be on different frequencies. So TP-Link are suggesting that the cable modem is causing RF interference with the powerline system.*

* Log into the VM Hub and make a note of what upstream channels it's using. Then try powering it off, waiting 5 minutes and then turning it back on. Check the channel usage again. Sometimes it will use different channels. If it has see if that has effected your powerline throughput.

So if TP-Link's theory about RF interference is correct then all you can really do is make sure that the coaxial cable into the VM Hub doesn't go anywhere near the powerline adaptors or your mains wiring. So don't lie the cables over each other, don't put the coax cable near the mains wall sockets, and don't run it along a wall that has a ring main behind it.

Also, at the risk of stating the obvious, always plug the powerline adapters directly into the wall sockets and not into an extension strip.

Thanks for your reply. Yes thats right we're using the same modem/hub as before, Virgin Hub 3. Absolutely nothing changed apart from the provisioned speeds increased from 200 Down 12 Up to 350 Down 20 Up. The upload speeds are so slow anyway with Virgin that nothing has been effected there, it's just the download speeds via the Powerline that have seemingly nonsensically decreased. Most frustrating.

What has been quite frustrating is that TPLInk support have been quite vague as to what exactly is interfering with what and where. They seem to suggest it's the coaxial (cable or hub, or the signal itself, I'm not sure which) which increasingly interferes with the powerline system (they don't say whether with adapters themselves or the actual electric ring main in the house), as the coaxial power increase in order to accommodate the higher speed Vivid 350 vs the Vivid 200 service.

You mentioned the Upstream frequencies and channels etc, apologies for my naivety, but does Upstream equate to the Upload data part of the modem? If so, as I mentioned, we don't have any problems with uploads, it's just the downloads, not being able to get at least the 200-230Mbps we used to get through the powerline, now we have the speed upgrade to 350Mbps.

I've copied the channels/frequencies info from my Hub 3 below, would be interesting to see what you or anyone can make of them in light of this hypothesised Coaxial/Powerline EM Interference. They were quite specific this is a case of EM interference, not RF.


Cable Modem Status
Item Status Comments
Acquired Downstream Channel (Hz)
298750000
Locked
Ranged Upstream Channel (Hz)
39400000
Locked

Upstream bonded channels
Channel Frequency (Hz) Power (dBmV) Symbol Rate (ksps) Modulation Channel ID
1 39400000 30.8 5120 64 qam 50
2 25800000 29.5 5120 16 qam 52
3 32600000 30.3 5120 64 qam 51
4 46200000 29.8 5120 64 qam 49


Upstream bonded channels
Channel Channel Type T1 Timeouts T2 Timeouts T3 Timeouts T4 Timeouts
1 ATDMA 0 0 0 0
2 ATDMA 0 0 0 0
3 ATDMA 0 0 0 0
4 ATDMA 0 0 1 0


Downstream bonded channels
Channel Frequency (Hz) Power (dBmV) SNR (dB) Modulation Channel ID
1 298750000 15.9 38 256 qam 21
2 322750000 15.8 38 256 qam 24
3 314750000 15.9 38 256 qam 23
4 306750000 15.8 38 256 qam 22
5 290750000 16 38 256 qam 20
6 282750000 16 38 256 qam 19
7 274750000 15.9 38 256 qam 18
8 266750000 16.1 38 256 qam 17
9 258750000 16.5 38 256 qam 16
10 250750000 17 38 256 qam 15
11 242750000 17 38 256 qam 14
12 234750000 17 38 256 qam 13
13 226750000 17.3 38 256 qam 12
14 218750000 17.3 38 256 qam 11
15 210750000 17 38 256 qam 10
16 202750000 17 38 256 qam 9
17 194750000 17 38 256 qam 8
18 186750000 16.9 38 256 qam 7
19 178750000 16.8 38 256 qam 6
20 170750000 16.6 38 256 qam 5
21 162750000 16.6 38 256 qam 4
22 154750000 16.4 38 256 qam 3
23 146750000 15.6 38 256 qam 2
24 138750000 15.5 38 256 qam 1


Downstream bonded channels
Channel Locked Status RxMER (dB) Pre RS Errors Post RS Errors
1 Locked 38.9 274 0
2 Locked 38.6 295 0
3 Locked 38.6 200 0
4 Locked 38.9 204 0
5 Locked 38.9 174 0
6 Locked 38.6 158 0
7 Locked 38.9 113 0
8 Locked 38.9 158 0
9 Locked 38.6 580 0
10 Locked 38.9 71 0
11 Locked 38.6 89 0
12 Locked 38.9 72 0
13 Locked 38.6 46 0
14 Locked 38.9 60 0
15 Locked 38.9 60 0
16 Locked 38.9 43 0
17 Locked 38.9 41 0
18 Locked 38.9 45 0
19 Locked 38.9 154 0
20 Locked 38.9 33 0
21 Locked 38.9 40 0
22 Locked 38.6 41 0
23 Locked 38.6 95 0
24 Locked 38.6 45 0

 

[ColinTaylor]

Yes, upstream relates to uploads. That doesn't necessarily mean that you have to be uploading data for the upstream channels to be in use. There's always going to be electrical signals going in both directions whether you're doing anything or not.

That said, I'm not entirely buying this "interference" theory. But as you seem to have thoroughly tested everything else it's worth investigating. EM vs. RF, it's the same thing.

I'm on exactly the same package as you, also with a Hub3. Here are my upstream channels and frequencies.

So yes, they are within the range TP-Link says their adaptors use (1.8MHz - 68MHz) but it's not exactly covering the entire spectrum is it.

Have you tried physically moving the equipment (Hub3 and Ethernet cables) about, just as a test? These things tend to be squashed together behind a TV or something. Try spreading everything out as much as possible and doing your test again. EM/RF interference is particularly sensitive to positioning. If it is an interference problem I'd expect that moving the Hub3 (and its coax cable) say 3 feet further away from the powerline adaptor would result in a significant change in behaviour.

 

[phillipwj]

That said, I'm not entirely buying this "interference" theory. But as you seem to have thoroughly tested everything else it's worth investigating. EM vs. RF, it's the same thing.

I'm on exactly the same package as you, also with a Hub3. Here are my upstream channels and frequencies.
View attachment 13954
So yes, they are within the range TP-Link says their adaptors use (1.8MHz - 68MHz) but it's not exactly covering the entire spectrum is it.

Have you tried physically moving the equipment (Hub3 and Ethernet cables) about, just as a test? These things tend to be squashed together behind a TV or something. Try spreading everything out as much as possible and doing your test again. EM/RF interference is particularly sensitive to positioning. If it is an interference problem I'd expect that moving the Hub3 (and its coax cable) say 3 feet further away from the powerline adaptor would result in a significant change in behaviour.

Believe me I'm not really sold on the coaxial interference idea either. As Virgin said, the shielding on the latest virgin coaxial cables is very strong indeed. Any interference would be messing up everything. Looking through all the various test results it's also interesting to note that the Virgin Hub 3's 4 port switch doesn't seem to be performing well either. It should be Gigabit speeds and yet compare it's ~300Mbps speeds vs the Netgear Gigabit switch ~950Mbps + speeds and also bear in mind NONE of that traffic was even going through the coaxial or powerline. That was all tested with iperf to check each piece of hardware looking for a weak link.

TP Link of course blame Virgin and yet Virgin are delivering 350-380Mbps to the hub as promised. Seems the Virgin Hub3's 4 port switch doesn't help matters, but anyway.

Next thing on my list is to move the powerline adapter #1 to another power outlet on other side of the room away from the Hub and use a long ethernet cable to join them.

TPLink have also offered to exchange my Tl-PA9020P Kit V1 for a V2 hardware version instead which may make a nominal difference. They said they also have a V3 version but said the UK warehouse don't have any in stock. I did hours of researching but couldn't find any info re specific changes between v1, v2 and v3 hardware and their specs. If anyone has any info that would be amazing!

Thanks for all your help so far! Will keep you posted re my ongoing efforts.

 

[ColinTaylor]

Looking through all the various test results it's also interesting to note that the Virgin Hub 3's 4 port switch doesn't seem to be performing well either. It should be Gigabit speeds and yet compare it's ~300Mbps speeds vs the Netgear Gigabit switch ~950Mbps + speeds and also bear in mind NONE of that traffic was even going through the coaxial or powerline.

I can't see that test in your post. Tests D and E are going through the powerline and the other test you mentioned was testing the download speed from the internet. (I probably can't see it for looking )

 

[phillipwj]

Did you change modems with the service speed upgrade?

How far away did you move the powerline adapter from the Virgin Hub? Why did you add the switch?

No it's still the Hub 3.0 which we have had for ages. Originally they were about 30cm apart. Then I tried moving the Hub about a metre away. Today I tried moving the powerline to a socket on the other side of the room and connected to the Hub with a long ethernet cable. All ultimately gave exactly the same speed downloads 170Mbps. Bizarre.

TPLink are sending me a replacement TL-PA9020P kit which is V2 hardware whereas my current one is V1. They said theres a small chance it might be able to deal with the interference better but to be honest I'm highly dubious. Seems crazy that the proposed interference would only affect the internet download speed across powerline but not any of the other speeds across the powerline network.

 

[System Error Message]

i should advice though, when you are using powerline, it is recommended to plug in switching PSU based devices into them rather than in parallel, devices that cause interference when plugged into the powerline device, if it has a filter would filter it out.

the best i got with mine was 200Mb/s across the fuse.

 

[phillipwj]

i should advice though, when you are using powerline, it is recommended to plug in switching PSU based devices into them rather than in parallel, devices that cause interference when plugged into the powerline device, if it has a filter would filter it out.

the best i got with mine was 200Mb/s across the fuse.

Hi there, could you just clarify what you mean by this?

I've added a photo below so you can see the current cable layout. There are a fair few PSU's. One extension cable handles the TV, Sound System, and two Lights and is plugged into the other socket in the wall. Then the second extension handles the Philips Hue Hub, the Virgin Hub, the Tivo V6 and the Netgear Switch and that is plugged into the Powerline Pass through socket.

You can also see the coaxial cable, which comes in from the left, goes down along the skirting board then back up to a splitter, one half to the Virgin Hub to supply broadband, the other to the Tivo to supply the TV signal.

I did try locating the Powerline adapter on the other side of the room but it made zero difference to the internet speed across the powerline, remained steady at 170Mbps.

 

[phillipwj]

I can't see that test in your post. Tests D and E are going through the powerline and the other test you mentioned was testing the download speed from the internet. (I probably can't see it for looking )

Sorry, my mistake, I did NOT actually test just the Virgin Hub3's 4 port hub alone, as my MacPro setup is in a different room and I don't have an ethernet cable long enough to reach it.

 

[sfx2000]

A: PC1 (Laptop)---Netgear Switch(Gigabit)---TL-PA9020P(No2.)—PC2 (Mac Pro Desktop) >>>> 938Mbps / 941Mbps (i.e. not going across powerline network, just testing the switches on the powerline adapter and the netgear switch)

B: PC1 (Laptop)—TL-PA9020P(No1.)~~~TL-PA9020P(No2.)---PC2 (Mac Pro Desktop) : >>>>>> 318Mbps.

C: PC1 (Laptop)---Netgear Switch(Gigabit)---TL-PA9020P(No1.)~~~TL-PA9020P(No2.)—PC2 (Mac Pro Desktop) >>>>> 344 / 362 / 372 Mbps

D: PC1 (Laptop)---(LAN)Virgin Hub3(LAN)---TL-PA9020P(No1.)~~~TL-PA9020P(No2.)—PC2 (Mac Pro Desktop) >>>>>> 295 / 311/ 302 Mbps

E: PC1 (Laptop)---(LAN)Virgin Hub3(LAN)---Netgear Switch(Gigabit)---TL-PA9020P(No1.)~~~TL-PA9020P(No2.)—PC2 (Mac Pro Desktop) >>>>>> 298 / 311 / 321 Mbps

We also know that the coaxial cables are delivering INTERNET traffic to the Hub3 at full speed 350-385Mbs since I connected the laptop direct to hub with Cat6 cable and as expected this gave me immediate full speed of c350-380Mbps.

We also know that INTERNET traffic is reaching PowerLine Adapter No1 at full speed of 350-385Mbos as I just tested that again this morning as follows:

ISP(Virgin)---Virgin Hub3---Netgear Switch(Gigabit)---TL-PA9020P(No1.) Ethernet Port 1 ~~~TL-PA9020P(No1.) Ethernet Port 2 —Laptop (i.e. not going across the powerline network, just into one of the powerline adapters ethernet ports and out the other one).

We know that NON-INTERNET NETWORK traffic can get through the 4 port switch on the Hub3 at speeds between 295-321Mbps as we’ve tested the speed of that switch.

We know that NON-INTERNET NETWORK traffic can get through the gigabit switch at essentially full gigabit speeds 941Mbps

We know that NON-INTERNET NETWORK traffic can get through the PowerLine network at speeds ranging between 344-372Mbps.

We know that INTERNET traffic can only get through the PowerLine network at c170Mbps, i.e. half the speed of the other network traffic.

Hi - you've sorted that it's not the TP-Link directly with your testing - the HPAV2 link is more than sufficient to support the upgrade on your WAN connection

Check the Settings on the TP-Links on both ends - plug directly into them with a ethernet connection and the TP-Link util - check the QoS settings on each one - if you change, it's only for the connected one, it doesn't push across all of them.

If you're getting half-speed on the far end, check the ethernet cable attached - you could be half-duplex for whatever reason (yes, this happens).

Barring other changes in your setup - HomePlugs are notorious for interference from AC to DC adapters - not specific to TP-Link there.

 

[System Error Message]

Hi there, could you just clarify what you mean by this?

I've added a photo below so you can see the current cable layout. There are a fair few PSU's. One extension cable handles the TV, Sound System, and two Lights and is plugged into the other socket in the wall. Then the second extension handles the Philips Hue Hub, the Virgin Hub, the Tivo V6 and the Netgear Switch and that is plugged into the Powerline Pass through socket.

You can also see the coaxial cable, which comes in from the left, goes down along the skirting board then back up to a splitter, one half to the Virgin Hub to supply broadband, the other to the Tivo to supply the TV signal.

I did try locating the Powerline adapter on the other side of the room but it made zero difference to the internet speed across the powerline, remained steady at 170Mbps.

Thats a really bad setup, at least use a surge protector. Firstly though theres one strip plugged next to the powerline, if the total wattage is below 3KW you can plug the power strip into another one, because not all the devices that have interference is behind the powerline. And virgin media is absolutely crap, especially their hub. Do a LAN test before complaining because its not unusual to have problems with virgin media

 

[phillipwj]

Thats a really bad setup, at least use a surge protector. Firstly though theres one strip plugged next to the powerline, if the total wattage is below 3KW you can plug the power strip into another one, because not all the devices that have interference is behind the powerline. And virgin media is absolutely crap, especially their hub. Do a LAN test before complaining because its not unusual to have problems with virgin media

Hi again 'System Error Message'. Thanks for your input.

So to clarify, you're saying I should unplug from the second wall socket the extension cable which powers the TV, Sound System and two lights and instead plug that into the end of the extension which currently powers the Philips Hue Hub, the Virgin Hub, the Tivo V6 and the Netgear Switch and have the whole lot go through the powerline pass through socket, using a surge protection extension, thereby placing everything BEHIND the powerline rather than parallel to it?

Yes, I can definitely try that. I also have a single, bigger, longer, surge protected extension that could carry the whole lot and move it further away from the powerline adapter.

Is there anything else you can expand on re my 'really bad setup'?

We can both agree that Virgin Media, while fast in certain situations, is definitely crap in terms of it's Hubs. When you say 'do a LAN test before complaining', what exactly do you mean? A LAN test on which components? I've done so many I'm now almost losing the will to live. Do you mean as per above, testing the LAN speed through the Hub's 4 port switch? The one test I hadn't managed to carry out?

 

[phillipwj]

Hi - you've sorted that it's not the TP-Link directly with your testing - the HPAV2 link is more than sufficient to support the upgrade on your WAN connection

Check the Settings on the TP-Links on both ends - plug directly into them with a ethernet connection and the TP-Link util - check the QoS settings on each one - if you change, it's only for the connected one, it doesn't push across all of them.

If you're getting half-speed on the far end, check the ethernet cable attached - you could be half-duplex for whatever reason (yes, this happens).

Barring other changes in your setup - HomePlugs are notorious for interference from AC to DC adapters - not specific to TP-Link there.

Hi SFX2000, This is exactly whats bothering me, I also believe the TPLInk should be more than capable of handling the internet download speeds, it is after all handling everything else fine, just not the internet downloads.

I'll certainly try setting the QoS settings using the method you outline when I get the replacement TP-LInks back from them (they've kindly offered to replace my PA9020P V1 hardware with PA9020P V2 hardware). Up until now I used the TPLink Utility to set both at the same time even though only connected to one end. The Utility reported that the specified QoS settings had been successful, but I'll see whether connection to each locally does indeed make a difference.

I've done a lot of different cable testing over time. True, the odd cable was operating sub-par, but I've now got the high quality shielded Cat6 purple cables acting as the backbone between my key components so I'm fairly certain that area is now covered.

Re Duplex, I was sure I read somewhere that Powerlines are always Half Duplex anyway and hence cannot ever live up to the full speeds of ethernet. Another argument for ultimately chucking them out the window and installing ethernet cabling - I wish. But at least I know that way I'm guaranteed the full 350-380Mbps internet speed I KNOW the Hub is delivering as promised, as per my tests.

 

[System Error Message]

Hi again 'System Error Message'. Thanks for your input.

So to clarify, you're saying I should unplug from the second wall socket the extension cable which powers the TV, Sound System and two lights and instead plug that into the end of the extension which currently powers the Philips Hue Hub, the Virgin Hub, the Tivo V6 and the Netgear Switch and have the whole lot go through the powerline pass through socket, using a surge protection extension, thereby placing everything BEHIND the powerline rather than parallel to it?

Yes, I can definitely try that. I also have a single, bigger, longer, surge protected extension that could carry the whole lot and move it further away from the powerline adapter.

Is there anything else you can expand on re my 'really bad setup'?

We can both agree that Virgin Media, while fast in certain situations, is definitely crap in terms of it's Hubs. When you say 'do a LAN test before complaining', what exactly do you mean? A LAN test on which components? I've done so many I'm now almost losing the will to live. Do you mean as per above, testing the LAN speed through the Hub's 4 port switch? The one test I hadn't managed to carry out?

well the main thing in your bad setup is just the lack of a surge protector and VM, other than that its fine. But yes the TV and lights do actually generate interferences especially if its one of those efficient LED lights due to the way their power circuits operate. As with the VM hub, you should just get another router to use behind it and use it as a modem only, that would significantly improve your internet experience.

When i say LAN test i mean try to transfer a file from one PC to another, You can also use iperf between 2 PCs. your VM hub will not give you your full promised speeds, even BT has improved their routers significantly and i have one with me. With the VM hub, it takes a minute or 2 just to connect to its wifi.

In my case i got powerline because i was only renting a room, so am not allowed to put long cables around and BT's best was far below its capacity but i picked BT over VM simply because of better uploads. Ethernet would be far better and long cables are cheaper (my tp link AV2000 wasnt as cheap). And yes powerline is half duplex, but its interesting that people say that the power lines (on the street) has so much unused space.

Honestly, your upload is the actual measure of your internet speed, not your download, and an ISP being stingy on uploads makes internet more expensive, because a country that imports data rather than exports will have to pay to the ISPs in other countries. higher consumer uploads, along with local datacenters will improve data exports and make your country's network more desirable. This is why asymmetrical speeds, one with slower uploads are detrimental and why i particularly hate cable ISPs, at least with BT they're limited by tech (VDSL).

 

[phillipwj]

well the main thing in your bad setup is just the lack of a surge protector and VM, other than that its fine. But yes the TV and lights do actually generate interferences especially if its one of those efficient LED lights due to the way their power circuits operate. As with the VM hub, you should just get another router to use behind it and use it as a modem only, that would significantly improve your internet experience.

When i say LAN test i mean try to transfer a file from one PC to another, You can also use iperf between 2 PCs. your VM hub will not give you your full promised speeds, even BT has improved their routers significantly and i have one with me. With the VM hub, it takes a minute or 2 just to connect to its wifi.

In my case i got powerline because i was only renting a room, so am not allowed to put long cables around and BT's best was far below its capacity but i picked BT over VM simply because of better uploads. Ethernet would be far better and long cables are cheaper (my tp link AV2000 wasnt as cheap). And yes powerline is half duplex, but its interesting that people say that the power lines (on the street) has so much unused space.

Honestly, your upload is the actual measure of your internet speed, not your download, and an ISP being stingy on uploads makes internet more expensive, because a country that imports data rather than exports will have to pay to the ISPs in other countries. higher consumer uploads, along with local datacenters will improve data exports and make your country's network more desirable. This is why asymmetrical speeds, one with slower uploads are detrimental and why i particularly hate cable ISPs, at least with BT they're limited by tech (VDSL).

I've actually got a BT Smart Hub (6) Type A (i.e. no WAN port, just 4x LAN) and blimey I've spent days and days trying to get that work efficiently and reliably as an additional Wireless Access Point, leaving the VMHub acting as router still and also leaving on it's wifi, but attaching the BT Smart Hub via ethernet, disabling it's DHCP but enabling it's wifi. It would work for a while, correctly picking up an IP address from the VM Hub DHCP server initially and allocating it to a device, but then it would never successfully manage to do it twice. So after about 30 mins nothing would be working through it. Very odd indeed. I also tried switching the VM Hub to Modem mode and going that route but I don't think that works correctly with the BT Smart Hub type A as there is no WAN port and you cannot set up the BT Smart Hub properly to use the VMHub3 as modem/gateway and Smart Hub as DHCP server / wifi. The whole thing was an experience in frustration.

 

[System Error Message]

I've actually got a BT Smart Hub (6) Type A (i.e. no WAN port, just 4x LAN) and blimey I've spent days and days trying to get that work efficiently and reliably as an additional Wireless Access Point, leaving the VMHub acting as router still and also leaving on it's wifi, but attaching the BT Smart Hub via ethernet, disabling it's DHCP but enabling it's wifi. It would work for a while, correctly picking up an IP address from the VM Hub DHCP server initially and allocating it to a device, but then it would never successfully manage to do it twice. So after about 30 mins nothing would be working through it. Very odd indeed. I also tried switching the VM Hub to Modem mode and going that route but I don't think that works correctly with the BT Smart Hub type A as there is no WAN port and you cannot set up the BT Smart Hub properly to use the VMHub3 as modem/gateway and Smart Hub as DHCP server / wifi. The whole thing was an experience in frustration.

the one i have is the 5A which has an atheros chipset rather than broadcom, it works very reliably and has a WAN port, along with gigabit ethernet ports, unlike netgear's 100Mb/s VDSL modem which will not handle the 300Mb/s BT is rolling out despite supporting that VDSL protocol, hence it indicated to me early that netgear is just becoming overpriced crap. At least dlink is cheap. Unlike with the VM hub where all of them are the same and consistently terrible.

What i meant just now was using a better router like tplink or asus.

 

[sfx2000]

Re Duplex, I was sure I read somewhere that Powerlines are always Half Duplex anyway and hence cannot ever live up to the full speeds of ethernet. Another argument for ultimately chucking them out the window and installing ethernet cabling - I wish. But at least I know that way I'm guaranteed the full 350-380Mbps internet speed I KNOW the Hub is delivering as promised, as per my tests.

HomePlug's use a TDD link between the nodes, but it's a scheduled MAC, so that's not a big deal - the ethernet ports are full duplex.

 

[Samir]

The test I would try is an iperf between just the powerlines with and without the virgin equipment plugged in. If you can get your old speeds without it plugged in and it only drops after the virgin equipment is plugged in, the newer speeds may require more ampage and power draw changing the scenario on the power circuit which the powerlines use.