Why do I get poor speed through my powerline adapters?
Powerline adapters can be a great way to expand the coverage area of your home network. They can add ethernet ports where one is not available, avoid messy cable configurations, and can even create or extend a wireless connection. But there are some draw backs. First, they are susceptible to electrical noise and interference. Powerline adapters should not be plugged into surge protectors as these device filter some of this noise out and can affect performance. And the network speed can greatly be affected by distance and load on the powerline itself. One very common cause of speed issues with powerline adapter is location of the adapters.
Take this image for example. A user has placed their “base” unit connected to their router which is located on circuit 1. And the remote unit which is on the last circuit in their panel.
Assuming that the circuits are in linear configuration, the data must transmit through each circuit to get to the main unit and out to the internet like so.
Each time the data hops a circuit there will a reduction in overall network speed and performance.
So how do resolve this issue and improve speed between the adapters? Unfortunately, the only option is relocation. Reducing the hops will improve the network performance of the devices. Ideally, we would recommend relocating your modem/router and base unit to a more centralized location Thus allowing remote adapters and the far ends of the panel to achieve the best possible speeds.
- Copy Link
- Subscribe
- Bookmark
- Report Inappropriate Content
And that is okay. We welcome and apperciate everyone's feedback on our products. Our stance does remain though that internet speeds have a mulitude of factors that affect them. LAN and everything that goes with this does play a factor. We are not trying to mislead customer and there is nothing wrong with the design or how the Powerline products are marketed.
- Copy Link
- Report Inappropriate Content
Good afternoon. I have forwarded you last 2 post #20 and #21 to the test engineer who is most familar with powerline. I am pretty sure i can answer your questions from these post but I want to make sure that your get a more detailed response then I would likely be able to provide. I will respond once I have this infromation.
- Copy Link
- Report Inappropriate Content
Here is the response. Let me know if I need to provide anything back to the engineer.
- How much of the reported link rate between adapters can be realized?
How much overhead is there in Powerline transmission compared to TCP/IP? (I believe there is extensive error checking and redundancy. Does that take up 5% of the bandwidth, or 50%?) We are so used to working with "switched ethernet" that we forget how networking with shared media was constrained. (Remember when we had ethernet "hubs" before switches? or, when ethernet was run on an enormous coax cable?)
The overhead of Powerline transmission is usually more than 50%. The main overhead comes from modulation, demodulation and MAC framing.
- Is transmission between adapters affected by concurrent transmission on the wire?
i.e. if between A and B I have 200mbs, and between C and D I have 200mbs, can I put a total of 400mbs (less overhead) on Powerline at the same time?
Or, is it more likely that as soon as any adapter begins sending, the available capacity is immediately reduced?
Powerline communication is time division multiplexed. If between A and B I have 200mbs, and between C and D I have 200mbs, then you can put a total of 200mbs on Powerline at the same time.
- What causes an adapter to "lose sync" (Powerline LED turns RED) and essentially quit working?
Shouldn't it temporarily slow down, but resume when whatever caused the problem goes away?
Powerline LED turns RED indicates a link rate below 50Mbps or disconnection. This is usually because of the distance is too far or strong interference, resulting in the signal instability.
Whatever is being reported by tpPLC is not a measure of actual capacity. In every case, The amount of data transmitted is in every case 1/5th to 1/6th of the number provided by tpPLC. It would be really fascinating to learn how the engineers describe this number.
tpPLC reports the link rate read from the chip. This is the physical layer rate other than the application layer real performance.
For example, here's the calculation behind the link rate 2000 Mbps (PA9020 - AV2000):
- The chip can use a bandwidth of up to 84.3 MHz (from ~1.8 MHz up to ~86.1 MHz).
- In this bandwidth, a total of 3455 OFDM carriers are employed with a minimum symbol time of 40.96 us
- The maximum bit allocation per carrier is 12
- The maximum PHY rate is obtained as follows: 3455 carriers x 12 bits/carrier ÷ 40.96us = 1012.2 Mbps
- MIMO technology USES two way(L-N, L-G) to transmit data: 1012.2 Mbps x 2 = 2024.4Mbps
The overhead of Powerline transmission in physical layer and data link layer is usually more than 50%. The main overhead comes from modulation, demodulation and MAC framing.
My PROBLEM is with #4. In the few minutes that I was watching iPerf3, the throughput dropped to 4mbs. I think we also know that electrical noise affects the receiving end more than the sending end. So, #4 is able to send to #1, but occasionally has real trouble receiving from #1.
Yes, electrical noise always has a greater impact on the reception performance of the node it is close to.
#1 --------------------------------------------------- noise source --- #4
For #4:
In the sending direction, the noise is attenuated as well as the signal.
While in the receiving direction, the signal is attenuated, but the noise is not.
- Copy Link
- Report Inappropriate Content
Fabulous. Very clear explanation. And, my practical experience matches "overhead over 50%". NOW....
What specific products can be deployed to block electrical noise from 1.8MHz to 86MHz? If I have a device close to PA#4 and cannot move it, what can I do to keep electrical noise from the device from reaching the electrical circuit?
- Can I plug an EMI/RFI filter into the electrical outlet and then plug the device into it?
But, are there any 120v EMI/RFI filters that block this frequency range?
What are they and where do I buy them?
- Can I put ferrite cores on the power cord of the device?
Should they be at the end nearest the device, or at the end nearest the electrical plug?
Are there any ferrite cores that will make any difference?
Is there a spec I should look for?
Where do I get them?
- A lot of my 15amp circuits now have LED bulbs (that's about all a person can buy these days).
I also have circuits with cell phone chargers, UPS, stereo equipment, etc.
How can I block electrical noise generated by those LED bulbs and other devices from reaching the electrical circuits that carry Powerline?
(I have only two circuits carrying Powerline. Of course, they have "stuff" on them, but not LED's like the lighting circuits.
Could I clamp ferrite cores on the "hot lead" of those circuits inside the electrical panel?
If so, exactly which ferrite cores?
Where do I purchase them?
Thanks
- Copy Link
- Report Inappropriate Content
Still interested in recommendations (see above). Thought it might be useful to report on my experiment.
Since PA#4 is my "problem node", I did the following:
- Installed an EchoGear ECHO-ASW61 6 outlet surge protector/EMI-RFI filter on a nearby outlet.
- This leaves the PA-9020 alone on an electrical outlet, with nothing plugged into the pass through port.
- Plugged the refrigerator and plug strip powering my IP PoE cameras into the EchoGear using extension cords.
- Snapped a couple of ferrite cores around the power cords for the refrigerator and plug strip.
As a result, the typical reported transmission rate has increased in both directions.
However, a potential problem remains. The transmission rate reported every 10 seconds by tpPLC varies a lot. While "mostly" in the range of 90-110, it will be as high as 150 and as low as 25. Since actual throughput is only 1/5 to 1/6 of the reported rate, "25" is not enough to support IP camera streams. Since making these changes, the PA9020 has not "died" and needed to be reset.
- Copy Link
- Report Inappropriate Content
- Copy Link
- Report Inappropriate Content
Have you installed the tpPLC Utility? I find it really useful in two ways: (1) It gives me an instant view from my desktop of the status of my four TP-PA9020's. When I see the transmit rate change from a number to a question mark, I know to look at the Powerline LED status (still green? OK. Red? Time to reset.) (2) it provides a rough test of whether some change I have made "did anything". When I rearranged my garage wiring to isolate the PA and plugged my other devices into an EMI/RFI filter, speed and stability definitely improved.
I have issues with the tpPLC utility, but overall I recommend it over nothing.
What I am looking for is sources of EMI filters and ferrite cores that handle the frequency range used by Powerline AV2. (1.8MHz to 86MHz) The most common (apparently) ferrite cores start at 25MHz, which is way too high. Likewise, some seem to "top out" at 300MHz or 50MHz. (too low). Ferrites sold on Amazon (where I bought a box to try out) never list any technical information. No idea what I purchased.
- Copy Link
- Report Inappropriate Content
Well, I guess it is about time to "throw iin the towell".
The issue that started all this is my adapter #4 in the garage that once in a while just dies and I have to reset and resync it. tpPLC reports that transmittion TO the adapter is consistently less than half of the transmission FROM the adapter. This led to the idea that there must be some electrical equipment close to adapter #4 that is making electrical noise and the noise is much greater relative to the signal from 200+ft away. I installed surge protector/EMI filters/ferrite cores. No improvement. I installed surge/EMI devices on every outlet in the entire pathway from adapter #1 to #4. No improvement. I got a long extension cord and moved everything in the garage from that circuit to an entirely different circuit. No improvement. Transmission TO adapter #4 can be as high as 120mbs and as low as 15mbs.
Just now, transmission FROM was 195mbs and TO was 45bms. 10 seconds later TO was 90.
I want to thank Carl for staying with this, asking the engineers questions and relaying information. This is more support direct from the company than I get on other user forums. If they come up with any more tips, please pass them on. I am searching now for a network monitor that will reliably alert me when the Powerline network drops.
- Copy Link
- Report Inappropriate Content
I am not ready to give up quite yet. There is one more person i would like to have review the thread to see if they can provide some insight. Let me get this over to them and I will let you know what I get from them.
- Copy Link
- Report Inappropriate Content
Thanks for following up. I continue to be mystified about this particular Powerline adapter. Just now when I checked tpPLC, the speed "to" the adapter was 129mbs and the speed "from" was 235mbs. Since I have witnessed as low as 17mbs to the adapter, I set up a "ping" watch on it. The adapter has a PoE switch connected to it that services a number of IP security cameras. I am using EMCO Ping Monitor (because it is "free" and easy to set up.) In the first 24 hours it logged "Down" at least one time when doing a ping once a minute. ICMP (ping) is almost no data at all, but it is a UDP packet. Either it didn't get there, or the response didn't get back.
I understand that "something in my electrical wiring environment" causes the wild fluctuations in Powerline performance.
I am tempted to swap in one of the "new" (Version 3.6) adapters, but I have no confidence that the versions are entirely compatible.
- Copy Link
- Report Inappropriate Content
Information
Helpful: 0
Views: 31430
Replies: 32
Voters 0
No one has voted for it yet.