Powerline Adapter only providing 100mbps when I have a Gigabit Plan, 2 Cat 5E cables, and a Nighthawk N7000

Werepig · Oct 12, 2018

Just like the title suggests, I have the stuff mentioned above, but for some reason my Powerline Adapter, which is advertised to give up to 600mbps, is only giving me 94.

Powerline Adapter: https://www.amazon.com/HooToo-Electrical-Networking-Standalone-One-Button/dp/B01M0EEJBP/ref=cm_cr_arp_d_product_top?ie=UTF8

bill001g · Oct 12, 2018

Pretty standard lies they tell. The device only has a 10/100 ethernet port so how could it ever run faster.

Even the so called 1200 models only get 150mbps or so. Look at this chart for more realistic expectation.
https://www.smallnetbuilder.com/tools/charts/powerline/view

quilciri · Oct 12, 2018

They're not lies. They are however, under perfect conditions; i.e. on the same circuit, excellent in-wall wiring and shielding, plugged directly into outlets, and transferring a single large file, etc. I've pulled 700Mbps with the Trendnet AV2 1200 when testing a pair in the same outlet.

This is the same way every network technology is rated; under ideal circumstances. A direct gigabit ethernet connection tops out around 800Mbps in the best real world circumstances.

Problem is, if your devices are close enough for the adapters to be on the same circuit, you may as well use a straight ethernet connection. Where the trendnet ended up (a floor and two rooms away), I get about 120Mbps with a single large file, less with many small files.

P.S. It's a gigabit port on those adapters, not a 10/100. If it were 10/100, the throughput would be MUCH less than 100Mbps.

bill001g · Oct 12, 2018

quilciri :

Why do you think it is gigabit ports. The manual clearly says 10/100.....I did look this up before I posted that.

https://www.hootoo.com/media/downloads/81-30002-101_HT-ND002%28NO%20WIFI%29%20-User%20Guide%20V1.3-20160613.pdf

quilciri · Oct 12, 2018

Because it is physically impossible if the OP is getting 100Mbps throughput.

That manual is wrong. If you had a 1 ft Cat5 cable plugged into two 10/100 ports on either end, you still wouldn't get 100Mbps throughput. You'd cap out at 80-90Mbps. Even if one end was a gigabit port, and you were using cat5e or 6, it still would never reach 100.

When you send data through any kind of cable, there is always loss. The 100Mbps refers to how much data can go through the sending port. There will always be some portion of the data that doesn't reach the other end. How much data is lost depends on the quality of the cabling, the shielding, how much outside interference there is, etc. Cabling has a max length for similar reasons (for instance, cat5/5e/6 has a 100 meter max length)

shmoochie · Oct 12, 2018

Just read through the customer reviews and you will see that this powerline adapter is notorious for slower speeds and a bad manual. The electrical wiring in the house plays a massive part, but I would bet money on this just being a shoddy adapter.

quilciri · Oct 12, 2018

shmoochie :

No argument there

TJ Hooker · Oct 12, 2018

@quilciri the OP didn't say they were actually getting 100 Mbps, they say they were getting 94 Mbps.

And no, you don't necessarily always get data loss through a cable. The reason you can't ever get exactly 100 Mbps with Fast Ethernet is because of packet header/preamble stuff, that isn't actually payload data.

The obvious answer is that the ethernet ports are only 100 Mbps, as @bill001g (and the specifications) said.

bill001g · Oct 12, 2018

Believe what you like I will trust the manuals. You are making you assumption based on him stating 94mbps which is all of 4mbps faster than your so called maximum of 90mbps. The actual maximum if you want to be nit picky is somewhere in the 97% range. The math is pretty straight forward.

If you take a maximum packet size of 1500 bytes and then account for the overhead of the mac headers and the preamble and others you get 1538 bytes. So 100mbit/1538bytes after conversion give 8120 frames/sec. So when you only consider the usable part of the data you get
8120 frames *1500bytes * 8 gives about 97mbit

shmoochie · Oct 12, 2018

Ew they do have a 10/100 port. How can they advertise 600 mbps data transfer when it is not physically possible?

quilciri · Oct 12, 2018

TJ Hooker :

10/100 ports on powerline adapters connected to outlets that are unlikely to be on the same circuit are not going to achieve 94 Mbps throughput. They won't even come close. I didn't think I needed to say that.

If you have a perfectly straight cable in which even the atoms of copper are aligned, and the path has absolutely zero electromagnetic interference, then you can theoretically achieve zero data loss. Those conditions don't exist in any capacity we'll ever see, so for us, you will always have some degree of data loss.

The packet headers are irrelevant. There are other things I didn't mention, such as receive window size, and the difference between throughput and goodput, because they were unnecessary to make the point, and introduce unneeded complexity.

*edit* Sorry, I shouldn't say they're irrelevant, they *are* relevant, just not the primary cause of the OP's issue.

quilciri · Oct 12, 2018

TJ Hooker :

Btw, even according to this argument, it's still impossible for them to be 10/100 ports and achieve 94Mbps....and that's not even counting all the reasons I already mentioned.

https://www.cablefree.net/wireless-technology/maximum-throughput-gigabit-ethernet/

TL;DR - "The approximate throughput for Gigabit Ethernet without jumbo frames and using TCP is around 928Mbps or 116MB/s."

That would place the theoretical maximum for 10/100 at 92.8Mbps without jumbo frames. No powerline adapter that I know of supports jumbo frames. I strongly doubt the HooToo ones do

bill001g · Oct 12, 2018

So your 80-90 number you said previously you now prove wrong and say it is 92.8. So now you are only 1 away from his claimed 94. This is your basis that the documentation from the vendor must be incorrect and the device must have gigabit ports.

He never said what tool he used to measure but almost all network measurement tools like speedtest consider the tcp headers as part of the data. This tends to be why all network equipment measure rates in mbits because they don't want to be concerned with what is payload and what is not. That is partially why you can not convert file transfer MByte rates to network mbit rates because the file transfers only care about payload.

So if we look at document you post it also show 98% maximum rates for ethernet frames.

You are really grasping at straws trying to prove that the device has gigabit ports even when the documentation show it has 10/100. You can believe what you like but I will still say the limitation can easily be the 100mbit ports. Now could it be the powerline technology also limiting him sure. I am very surprised he gets 94mbit on those unit.

quilciri · Oct 14, 2018

I'll explain more simply.

Two powerline adapters in any realistic environment(e.g rooms away, not on the same circuit, with who-knows-what interference), even with perfect controllers are not going to be **anywhere** near the theoretical maximum transfer speed. Period. End of story.

Therefore, because the OP is seeing speed so near (or over depending on how you want to measure) of what 10/100 is capable of, they aren't 10/100 ports on his adapters.

As you can see below, just going from 25 to 50 meters of seperation drops the throughput by over 10%, and that's on a ridiculously clean lab test grid. In the next chart, as soon as they add an amount of interference as could be expected in normal home wiring... the throughput drops by HALF.

https://us.hardware.info/reviews/7129/8/6-homeplug-av2-powerline-adapters-with-wifi-review-tests-2-data-streams

Please stop spreading misinformation. You clearly have little experience here.

bill001g · Oct 14, 2018

This is going nowhere you can believe what you want. I will still trust the vendor documentation. What vendor would put a gigabit port in a unit and then sell it as 10/100

I can promise I know much more than you about networking in general. Go look up what CCIE means. I had that for over 15yrs before I recently retired

quilciri · Oct 15, 2018

Lol! Sure you do.

You are flat out wrong here, and don't want to admit it. You have extremely little experience with powerline adapters, and it shows that you haven't even bothered to research. I post facts and tests to back up my points and you post nothing but "trust in the vendor"

If you really had any significant IT experience, we wouldn't be having this conversation. If you were familiar with technical documentation, you'd know how often the technical writer gets something wrong and no one notices it before the documentation goes out. This happens even with large vendors and very competent technical writers..... and HooToo is a tiny Chinese company. Give me a break.

If you couldn't glean from the third chart in my last post how it is impossible for the Hootoo to have 10/100 ports, your teachers failed you.

To anyone that might be confused by this bill001g's misinformation, go buy any powerline adapter with a no-questions asked return policy and test it out yourself. Plug them into the same outlet and test transfer rate, then plug one in a room or two away and see how much the data transfer drops off.

TJ Hooker · Oct 15, 2018

If you have a perfectly straight cable in which even the atoms of copper are aligned, and the path has absolutely zero electromagnetic interference, then you can theoretically achieve zero data loss. Those conditions don't exist in any capacity we'll ever see, so for us, you will always have some degree of data loss.

I thought it was implied due to the context of this thread, but maybe I should have clarified I am talking about data loss in a digital signal, e.g. bit errors resulting in dropped packets. It's absolutely possible and realistic to achieve zero data loss in normal scenarios that involve signal attenuation and noise/interference. Just need a sufficient SNR. I'll admit I have no idea what sort of noise and attenuation levels you'd be dealing with for powerline adapters.

Edit: Ah, just realized you were probably talking about powerline adapters specifically, I thought you were talking about the general case, including ethernet. /Edit

Powerline adapters will also have forward error correction such that they can handle at least some errors without throughput being affected. It's also possible that they're smart enough to re-transmit dropped packets on their own (rather than having to tell the network device to resend), the IEEE spec appears to be behind a paywall though so I'm not sure. With those two things in mind, the raw bitrate of the powerline adapter could be significantly higher than the bitrate coming in to the ethernet port. That's probably part of their justification for selling it as 600 Mbps when the max theoretical throughput you could get from device to device is on par with 100 Mbps ethernet. That's also why a powerline adapter could potentially be able to keep a 100 Mbps ethernet connection saturated even while dealing with dropped packets.

Edit: The Homeplug AV2 spec at least mentions that it handles error checking and automatic resend. /Edit

I'll 2nd what @bill001g said about how we can't say with 100% certainty what the max theoretical 100 Mbps ethernet throughput should be without knowing exactly how the OP is measuring their speed.

quilciri · Oct 16, 2018

TJ Hooker :

You might be confusing data loss between the PC NIC and the switch with data loss between the adapters. If a packet is dropped between the adapters, one adapter still has to tell the other that it was lost and the packet has to be re-sent. This will reduce throughput between the adapters. There's nothing in existence that lets a device know a sent packet was dropped without another device telling it so. Maybe when quantum networking is a thing

As was shown in the testing link I posted, dropped packets happen quite often in a normal use environment for powerline adapters. With 50m seperation and a reasonable amount of interference, the throughput was cut in half on all tested adapters. So if the OP was seeing 94Mbps transfer, and he had his adapters a couple rooms apart, you can infer that if there were on the same circuit and closer together, you would see ~188mbps transfer rate. Even if they were only one room apart, putting them in the same outlet would see throughput much higher than 100mbps. So like I've said, they can't be 10/100 ports.

Cipherswitch · Apr 29, 2020

Werepig said:
Just like the title suggests, I have the stuff mentioned above, but for some reason my Powerline Adapter, which is advertised to give up to 600mbps, is only giving me 94.

Powerline Adapter: https://www.amazon.com/HooToo-Elect...01M0EEJBP/ref=cm_cr_arp_d_product_top?ie=UTF8

Sorry for reviving an old thread, but google dropped me here and the info isnt exactly right on this page.

Current TP-Link powerline adapters up to their AV600 series only have a 10/100M port. Everything above that upgrades to gigabit. Its in the spec, its on the box, its on the webpage, its in the product sort options, its even written on the product.

I have tested alot of their products trying to find a solution. Most recently their AV600 series vs the AV1000.

Through their own utility, a pair of AV600's shows a transfer rate of 250Mbps between each other. Connecting a laptop to the other end confirms what we already know. The laptop shows a 100m ethernet connection and my internet speed test shows 30-40Mbps while my internet speed at my switch tests at 300Mbps.

Performing the exact same test in the exact same outlets but with the AV1000 shows a 750Mbps connection between them , a 1.0Gbps connection on the laptop and speed test comes back with 275Mbps+ every time.

Hell, TP-Link even says it in their disclaimer " Actual data speeds vary and may be limited by the product’s Ethernet port and other network conditions such as data traffic, electrical noise, and wiring issues. "

That same disclaimer is not on the products with a gigabit port. They have also confirmed this themselves several times over in their own forums.

All that being said. You want these on the same circuit and avoid power strips if at all possible. If you need an internet outlet on a budget, the non gigabit versions work fine. If your internet connection is on the higher side (or even 100Mb), I would start at the AV1000.

Powerline Adapter only providing 100mbps when I have a Gigabit Plan, 2 Cat 5E cables, and a Nighthawk N7000

Prominent

Titan

Splendid

Titan

Splendid

Commendable

Splendid

Titan

Titan

Commendable

Splendid

Splendid

Titan

Splendid

Titan

Splendid

Titan

Splendid

Share this page