Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Nothing easy without cost: network powerline adapters (Read 11749 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Nothing easy without cost: network powerline adapters

I am seeking advice, thanks. 
I am using some powerline adapters to get "better" speed to my basement and kitchen.  They do provide faster than wifi speed, at a cost, a ton of electrical noise, I'm not sure which, EMI or RFI.  It is most noticeable when using headphones.

Here is the setup:
NAS->FiOS Router->ZyXEL Powerline PLA5205 Adapter->Computer (Gigabyte Gaming 5 motherboard with DACUP USB)->Audioquest Dragonfly->Audioquest Golden Gate Cable, 3 meter->NAD HP50 headphones

I can hear the noise, pops, crackles and clicking, in the headphones  when no music is playing or during a quiet section of a track.  When I unplug the network cable the noise completely goes away.  Playing from a local drive is the same, noise when the network cable is plugged into the jack, quiet when unplugged.  Furthermore, if I transfer a file to my NAS from the local drive or to my local drive from the NAS the line  noise "intensifies."  It is a bit louder and more of a consistent buzz.  The motherboard has the so-called DACUP USB ports which supposedly supply "cleaner" power to the USB ports over conventional USB ports.
 
What have I tried?  Not much, but here is some data.

NOISE:
1. I tried a ferrite clamp on the network cable - no effect.
2. I tried different outlets for the powerline adapter (probably same circuit) - no effect.
3. I tried USB3 ports (non DACUP ports) - no effect.
4. 1/8" female/male 6 foot extension cable - no effect.
5. I plugged the powerline adapter into my surge protector.  (It worked but at 1/2 the speed when plugged in the wall outlet.) - no effect

NO NOISE (but not a workable setup):
6. I temporarily moved my FiOS router to the basement to bypass the powerline adapter.  (The NAS was disconnected so I could not test file transfer noise.)  The line is silent, all conditions.
7.  Stock headphone cable  (too short to use without extension) and "lower quality" 1/8" inch cable.  The line is silent, all conditions.
8. Cheap external powered computer speakers - no EMI/RFI.  The line is silent, all conditions.


What can I do to eliminate the noise but not filter out the throughput of the powerline adapter?
A power conditioner?  (Hopefully inexpensive, i.e., <$100)
A network switch between the powerline and my computer? (Hopefully inexpensive too)
String RJ45 along the floor about 60 feet?  Yuk.  I cannot string cable though walls, so if I have to use a direct line, it will be very long and tacked down along the floorboard, doorways, and stairs - not cool.

Thanks for your help!!

Nothing easy without cost: network powerline adapters

Reply #1 –
Weird, man!

Have you tried the regular soundcard output?    I'm thinking the noise is getting-in through the USB power.    Usually an external device is more noise-immune than an internal device, but it might be worth trying. 

You could also try a powered USB hub.  (A lot of people recommend against using a hub for audio, but if you don't use it as a "hub" and only plug-in one device and it solves the problem that would be good.)

Nothing easy without cost: network powerline adapters

Reply #2 –
Quote from: tdxloki on
Here is the setup:
NAS->FiOS Router->ZyXEL Powerline PLA5205 Adapter->Computer (Gigabyte Gaming 5 motherboard with DACUP USB)->Audioquest Dragonfly->Audioquest Golden Gate Cable, 3 meter->NAD HP50 headphones


I don't think the networking setup is very relevant here.  Instead, explain how you have the wires in your basement connected.  Try putting everything on the same circuit (power and data) and see if that helps.

Quote from: tdxloki on
1. I tried a ferrite clamp on the network cable - no effect.
2. I tried different outlets for the powerline adapter (probably same circuit) - no effect.
3. I tried USB3 ports (non DACUP ports) - no effect.
4. 1/8" female/male 6 foot extension cable - no effect.
5. I plugged the powerline adapter into my surge protector.  (It worked but at 1/2 the speed when plugged in the wall outlet.) - no effect


None of that is going to matter except possibly changing which circuits devices are plugged into.  Try that.

Quote from: tdxloki on
What can I do to eliminate the noise but not filter out the throughput of the powerline adapter?


Assuming moving everything to the same outlet doesn't help, you could try buying a different power line adapter.  However, at that point I would just use the money for a better wifi repeater and avoid this problem entirely.

Nothing easy without cost: network powerline adapters

Reply #3 –
Quote from: DVDdoug on
Have you tried the regular soundcard output?    I'm thinking the noise is getting-in through the USB power.    Usually an external device is more noise-immune than an internal device, but it might be worth trying.


I tried the internal soundcard - silent.  So, you are right, its getting to the USB power.
 

Quote from: DVDdoug on
You could also try a powered USB hub.  (A lot of people recommend against using a hub for audio, but if you don't use it as a "hub" and only plug-in one device and it solves the problem that would be good.)


This seems reasonable.  I can give it a try but I currently do not have a powered hub.  Like you say, it does add an extra connection between the source and output but hey, if it works, ok.

As another test, I connected my basement powerline to a laptop and listened for line noise.  Exactly the same noise if not worse, especially on file transfer.

The powerline is definitely causing the problem.

Nothing easy without cost: network powerline adapters

Reply #4 –
Wait a minute, simply eliminating the headphone extension cable fixes the problem? If so, that would seem to suggest EMI being picked up via the wire, but why it should matter at the speaker level is a bit of a mystery to me. Try a different non-audiophile extension cable say by Monoprice or Radio Shack?

Nothing easy without cost: network powerline adapters

Reply #5 –
Quote from: 4season on
Wait a minute, simply eliminating the headphone extension cable fixes the problem? If so, that would seem to suggest EMI being picked up via the wire, but why it should matter at the speaker level is a bit of a mystery to me. Try a different non-audiophile extension cable say by Monoprice or Radio Shack?


Right, it is strange that I pick up the noise in the better cable.  The stock cable plus a cheap extension also reveals the EMI, but, like I said above, the stock cable alone or a throwaway cheap cable are silent.  The stock cable is a flat cable, seems nice, but its too short to reach from the computer, hence the 3 meter Audioquest cable.  The throwaway cable makes a poor fit in the earcup, bad production tolerance, but its tight enough to get stereo output.

I ordered another cable from Blue Jeans cable.  I should have it in a few days and I'll report back.

I thought by mentioning the cable nuance that I would detract from the main question, the EMI from the powerline adapters.  Regardless, I would like to use an approximately 10 foot cable, so I need to make that setup work.

Thanks!!

Nothing easy without cost: network powerline adapters

Reply #6 –
This might be basic, but have you actually tested the speed? I tried to use a power line for networking once, and it was terrible. There was 75%+ packet loss and I could barely pull single-digit megabits. I guess it depends on how "clean" the power is.

Nothing easy without cost: network powerline adapters

Reply #7 –
Is the network cable shielded or unshielded?

Nothing easy without cost: network powerline adapters

Reply #8 –
Quote from: slks on
This might be basic, but have you actually tested the speed? I tried to use a power line for networking once, and it was terrible. There was 75%+ packet loss and I could barely pull single-digit megabits. I guess it depends on how "clean" the power is.


When I copy a file to or from my Synology DS212 via powerline I get about 8 MB/sec.  Is that junk?

Well I am not experienced with quantifying packet loss the speed is about double relative to wifi.  It doesn't come close to advertised speeds or direct cable connection (obviously) but I don't get advertised speed, let alone speeds observed in review articles, for any of my tech.

We live in a small house.  Not too much power draw, aside from normal appliances, fridge, dryer, dishwasher, couple tvs, etc.

Nothing easy without cost: network powerline adapters

Reply #9 –
Quote from: pelmazo on
Is the network cable shielded or unshielded?


Definitely unshielded.  The ferrite clap has no effect either.  I think the ferrite serves a similar purpose to shielding, right?

Nothing easy without cost: network powerline adapters

Reply #10 –
Quote from: tdxloki on
Quote from: pelmazo on
Is the network cable shielded or unshielded?


Definitely unshielded.  The ferrite clap has no effect either.  I think the ferrite serves a similar purpose to shielding, right?


No the ferrite acts as low pass filter, attenuating the signal on the cable via inductive coupling.  Ferrites have a fairly narrow bandwidth though, so if you picked the right one it should reduce or even break your network connection, but most likely its not tuned to the bandwidth of interest so it probably does little/nothing.

Nothing easy without cost: network powerline adapters

Reply #11 –
Quote from: saratoga on
Quote from: tdxloki on
Quote from: pelmazo on
Is the network cable shielded or unshielded?


Definitely unshielded.  The ferrite clap has no effect either.  I think the ferrite serves a similar purpose to shielding, right?


No the ferrite acts as low pass filter, attenuating the signal on the cable via inductive coupling.  Ferrites have a fairly narrow bandwidth though, so if you picked the right one it should reduce or even break your network connection, but most likely its not tuned to the bandwidth of interest so it probably does little/nothing.


Nice!  Thanks!  I suspected I have a knowledge gap.

Should I try a shielded network cable?

I need to dig deeper but I think the powerline adapters use a high frequency.  Frequencies usually filtered by cheap surge protectors.  Would a shielded network cable help?

Nothing easy without cost: network powerline adapters

Reply #12 –
Quote from: tdxloki on
Should I try a shielded network cable?

I need to dig deeper but I think the powerline adapters use a high frequency.  Frequencies usually filtered by cheap surge protectors.  Would a shielded network cable help?


I have to correct saratoga's view. A ferrite clamp attenuates the common mode signals on the cable. The differential signals ought to be unaffected. Ethernet uses differential mode signalling, so if you find that a ferrite clamp chokes the data transmission, there's something rather seriously wrong. A ferrite clamp is a useful thing for EMI suppression because the type of signal that radiates off a cable is common-mode in nature.

It is quite different from a shield, and ferrite clamps can be effective against EMI problems even in the presence of a shield, because common mode signals can quite happily ride on the shield of a cable.

A shield can reduce the emissions from a network cable, but the effect usually matters only in close proximity of the cable. The twisting of the wire pairs in the network cable is already quite effective by itself, particularly once you are away from the cable by a handful of inches. Hence shielded network cables are beneficial mostly where a whole bunch of cables are running side by side over some distance, such as in cable ducts. For a patch cable that is both rather short, and also not in close neighborhood with other such cables, shielding achieves very little in terms of EMI.

The shielding can, however, close a ground loop, and thereby contribute to a completely different interference problem. Without the shield, the mandatory pulse transformers in each ethernet device isolate the circuits and prevent ground loops. Therefore, one can frequently improve the situation by deliberately using unshielded patch cables between networked audio devices and the rest of the network.

If you are already using unshielded network cable, my educated guess would be that the network cable itself is not the cause of your problem.

Nothing easy without cost: network powerline adapters

Reply #13 –
This is only a semi serious suggestion: you could try a long USB extender cable, and then a shorter headphone cable.

Can you try running the computer off a separate mains socket/supply from the powerline adapter?

Is there anything else in the basement/kitchen connected to the mains power that could be acting as a transmitter of the RFI from the powerline? What if you disconnect all your other appliances?

Cheers,
David.

Nothing easy without cost: network powerline adapters

Reply #14 –
Quote from: tdxloki on
When I copy a file to or from my Synology DS212 via powerline I get about 8 MB/sec.  Is that junk?

Given that 8 MB/sec is 64 Mbit/sec and your adapters claim to be 600 Mbit/sec I would say that is junk.  You are getting less than 1/9th the rated speed.  Wireless never gets its claimed speeds either but it is not common for it to be that bad.  If you are in a "noisy" 2.4 GHz wireless area then switching to 5 GHz might help if your distances aren't too great. 

I'm pretty well versed with networking and powerline adapters are universally junk.  You might get closed to reviewed speed on the same circuit with nothing else on them, but that is very unlikely in the real world. 

On thing to check, are your ethernet cables proper cat5 or better?  Do they have all 8 wires?  Some cheap cables only have 4 which limits them to 100 mbit.

Nothing easy without cost: network powerline adapters

Reply #15 –
Quote from: washu on
Quote from: tdxloki on
When I copy a file to or from my Synology DS212 via powerline I get about 8 MB/sec.  Is that junk?

Given that 8 MB/sec is 64 Mbit/sec and your adapters claim to be 600 Mbit/sec I would say that is junk.  You are getting less than 1/9th the rated speed.  Wireless never gets its claimed speeds either but it is not common for it to be that bad.  If you are in a "noisy" 2.4 GHz wireless area then switching to 5 GHz might help if your distances aren't too great.


I do believe I am running on 5 GHz, or at least I am on channel 6 or 11 for 2.4 GHz.  The PC in question is in the basement, one floor below the router which is in a cabinet.  I'm not sure if that would be too much to block the signal.  The PC currently does not have a wifi card or wifi usb.  I might revisit switching to AC wifi if you think that is a good route.  All of my hardware is n, but the basement PC could be AC...
 

Quote from: washu on
I'm pretty well versed with networking and powerline adapters are universally junk.  You might get closed to reviewed speed on the same circuit with nothing else on them, but that is very unlikely in the real world.


I suspected this, and I keep my expectations in check with claimed specs, but with my hardware, the powerline adapters gave me double the transfer speed of wifi.  So, relatively speaking, it was better, if not the proclaimed bandwidth.  Still, very disappointing, even more so now that I have encountered this crazy EMI. 

Quote from: washu on
On thing to check, are your ethernet cables proper cat5 or better?  Do they have all 8 wires?  Some cheap cables only have 4 which limits them to 100 mbit.

Yes, I belive they are, 8 wires, 4 white, 1 brown, 1 green, 1 blue, and 1 orange.

Nothing easy without cost: network powerline adapters

Reply #16 –
Did you try setting up wifi instead? If I recall powerline adapter speeds are theoretical (= speed of the chip inside them), highly unreliable (type of cables used, other stuff on your line, etc.) which means during real life scenarios they slow to a crawl and introduce a myriad of other problems. I'd rather spend your $100 on a router instead of trying to salvage your current setup.

Nothing easy without cost: network powerline adapters

Reply #17 –
Quote from: Daeron on
Did you try setting up wifi instead? If I recall powerline adapter speeds are theoretical (= speed of the chip inside them), highly unreliable (type of cables used, other stuff on your line, etc.) which means during real life scenarios they slow to a crawl and introduce a myriad of other problems. I'd rather spend your $100 on a router instead of trying to salvage your current setup.


I just installed a new PC in the basement where I will connect my headphones.  I can rerun this test with my laptop in the basement, thanks for the idea.  My previous setup was a laptop in the same room as the wifi router (Asus RT N66R - not junk).  In that setup the powerline produced double the transfer speed of the wifi when writing and reading from my NAS.  Still, nothing comparable to advertised or reviewed speeds (agreed, cables, refrigerator, tv, etc.)

Wifi n is just so underwhelming to me, compared to theoretical and cabled, that I hesitate to go for a massive upgrade to ac.

Nothing easy without cost: network powerline adapters

Reply #18 –
Quote from: tdxloki on
I do believe I am running on 5 GHz, or at least I am on channel 6 or 11 for 2.4 GHz.  The PC in question is in the basement, one floor below the router which is in a cabinet.  I'm not sure if that would be too much to block the signal.  The PC currently does not have a wifi card or wifi usb.  I might revisit switching to AC wifi if you think that is a good route.  All of my hardware is n, but the basement PC could be AC...

Your router is fine, but you need compatible wireless NICs to get the full possible speed from it.  Your router's theoretical max is 450 Mbit/sec, but since it is N it is split among 3 150 Mbit channels.  Your wireless NICs have to support 3 channels as well.  A basic "Wireless N" adapter is very unlikely to support 3 channels or even 2, so you are limited to a single channel at 150 Mbit. 

Also unless you live in the middle of nowhere you are very unlikely to get all three 2.4 GHz channels (1, 6, 11) clear at the same time required for 450 Mbit/sec.  You pretty much have to use 5 GHz, so make sure your wireless NICs support that as well.

Depending on how the router and your PC are positioned relative to each other could make a big difference.  If the PC is really directly below it should be fine.  If it is offset horizontally your signal could be going through a lot more material in the floor.  If you have a duct or something else metal in the path that can reduce the signal strength significantly.

In a decent wireless setup with no interference I've seen up to 2/3 of the claimed bandwidth, or 300 Mbit in your case.  That would give you ~37.5 MByte/sec file transfers, minus a bit of overhead.  Since you have a floor in between you are likely to get less, but you should still be able to better than 64 Mbit real world even with just N wireless.

Nothing easy without cost: network powerline adapters

Reply #19 –
Quote from: tdxloki on
Wifi n is just so underwhelming to me, compared to theoretical and cabled, that I hesitate to go for a massive upgrade to ac.

If you don't want to spend money on a wireless upgrade, why not just pay a cable installer to run a network line for you?  Then you get real 1000 Mbit, none of this theoretical crap. 

Nothing easy without cost: network powerline adapters

Reply #20 –
Quote from: pelmazo on
Quote from: tdxloki on
Should I try a shielded network cable?

I need to dig deeper but I think the powerline adapters use a high frequency.  Frequencies usually filtered by cheap surge protectors.  Would a shielded network cable help?


I have to correct saratoga's view. A ferrite clamp attenuates the common mode signals on the cable. The differential signals ought to be unaffected. Ethernet uses differential mode signalling, so if you find that a ferrite clamp chokes the data transmission, there's something rather seriously wrong. A ferrite clamp is a useful thing for EMI suppression because the type of signal that radiates off a cable is common-mode in nature.


Sorry, I misunderstood and thought he was putting it on the power line side of the cable, not the ethernet side.

Quote from: tdxloki on
I just installed a new PC in the basement where I will connect my headphones.  I can rerun this test with my laptop in the basement, thanks for the idea.  My previous setup was a laptop in the same room as the wifi router (Asus RT N66R - not junk).  In that setup the powerline produced double the transfer speed of the wifi when writing and reading from my NAS.  Still, nothing comparable to advertised or reviewed speeds (agreed, cables, refrigerator, tv, etc.)


The speed of a powerline system will depend on how much wire is between the devices.  If they're on the same local circuit, you will get reasonable speed.  If they're on separate circuits (that is, with a breaker and hundreds of feet of wire between them) then performance is going to be very poor.  Powerline is good for networking devices in the same room or sharing the same circuit, or for low bandwidth communication across a whole house that is too large for wireless. 

Quote from: tdxloki on
Wifi n is just so underwhelming to me, compared to theoretical and cabled, that I hesitate to go for a massive upgrade to ac.


Yeah but how good is your laptop's antenna and chipset?  A comparison of devices between 2 rooms on different floors of a house:



Is yours the 8 or 70 Mbit device depending on the quality of the receiver (same access point and 2.4G).

http://www.anandtech.com/show/7127/the-joy...-80211ac-wifi/2

Maybe think about a 2x2 802.11N device if you don't want to upgrade to an 802.11ac AP.  They're not very expensive at all.

Nothing easy without cost: network powerline adapters

Reply #21 –
Also, directional antennas can really help if you put them on both ends. I used two of these to get a reliable 2.4 GHz connection between some apartments that was otherwise a bit dodgy with regular omnidirectional antennas (multiple walls, refrigerators, furnaces, and ovens in the way.)

Nothing easy without cost: network powerline adapters

Reply #22 –
Quote from: 2Bdecided on
This is only a semi serious suggestion: you could try a long USB extender cable, and then a shorter headphone cable.

Can you try running the computer off a separate mains socket/supply from the powerline adapter?

Is there anything else in the basement/kitchen connected to the mains power that could be acting as a transmitter of the RFI from the powerline? What if you disconnect all your other appliances?

Cheers,
David.


The USB extension is a nice idea.  I looked into that option but given the strong EMI I wonder if the result would be the same.  What I would like to get is a nice quality USB A male to A female cable.  It could extend the setup of the Audioquest Dragonfly.  I do not see any of these type of cables sold by audiophile-type cable makers.  They are all USB A to B male cables.

I can try running the powerline adapter off a different mains socket with a long extension cord, but, if this fixes the problem, I won't be able to keep it setup that way.

Thanks for the suggestions!

Nothing easy without cost: network powerline adapters

Reply #23 –
Quote from: Rotareneg on
Also, directional antennas can really help if you put them on both ends. I used two of these to get a reliable 2.4 GHz connection between some apartments that was otherwise a bit dodgy with regular omnidirectional antennas (multiple walls, refrigerators, furnaces, and ovens in the way.)



Good idea, but one that can be taken quite a bit further.







Also: High powered wifi adaptors and access pointes are readily available and can also help.

Nothing easy without cost: network powerline adapters

Reply #24 –
Quote from: washu on
Quote from: tdxloki on
I do believe I am running on 5 GHz, or at least I am on channel 6 or 11 for 2.4 GHz.  The PC in question is in the basement, one floor below the router which is in a cabinet.  I'm not sure if that would be too much to block the signal.  The PC currently does not have a wifi card or wifi usb.  I might revisit switching to AC wifi if you think that is a good route.  All of my hardware is n, but the basement PC could be AC...

Your router is fine, but you need compatible wireless NICs to get the full possible speed from it.  Your router's theoretical max is 450 Mbit/sec, but since it is N it is split among 3 150 Mbit channels.  Your wireless NICs have to support 3 channels as well.  A basic "Wireless N" adapter is very unlikely to support 3 channels or even 2, so you are limited to a single channel at 150 Mbit.


I'm not sure where I would find if my NICs support 3 channels.  The NICs in my laptops are Intel WiFi Link 5100 AGN.  Granted, the laptops/NICs are about 5 years old, but they were supposedly good at the time.  The PC in question is brand new but currently only has a Gigbit port.

Quote from: washu on
Also unless you live in the middle of nowhere you are very unlikely to get all three 2.4 GHz channels (1, 6, 11) clear at the same time required for 450 Mbit/sec.  You pretty much have to use 5 GHz, so make sure your wireless NICs support that as well.


I'll have to double check with my FiOS router if I am on 5 GHz, but when I was using the Asus RT N66R I was on 5 GHz.  The 2.4 GHz was also broadcasting but I laptops were connected at 5 GHz.

Quote from: washu on
Depending on how the router and your PC are positioned relative to each other could make a big difference.  If the PC is really directly below it should be fine.  If it is offset horizontally your signal could be going through a lot more material in the floor.  If you have a duct or something else metal in the path that can reduce the signal strength significantly.


It is a floor below and on the opposite wall.  I think there is a decent amount of material for the signal to traverse, but nothing metal.

Quote from: washu on
In a decent wireless setup with no interference I've seen up to 2/3 of the claimed bandwidth, or 300 Mbit in your case.  That would give you ~37.5 MByte/sec file transfers, minus a bit of overhead.  Since you have a floor in between you are likely to get less, but you should still be able to better than 64 Mbit real world even with just N wireless.


I will test this out later tonight, the speed of my laptop with powerline and with wifi (FiOS router) in the basement.  Thanks for your help!!