Slow WiFi Speeds for Smart TV -- Router Setting or Bad TV ?

[copperhead]

I recently purchased a Sony Smart TV (KD75X800H) --Specs: 802.11ac, 2x2 MU-MIMO. My router is an Asus RT-AC86U -- Specs: 802.11ac, 4x4 MU-MIMO, 1024 QAM, 80 Mhz Channel Width.

Within the router's Wireless Logs i can see a Link Speed of 780 / 866 Mhz (Rx / Tx), RSSI = -60 dBm. The router is about 25 feet away, no obstructions. Asus Wireless Logs state the following FLAGS are active: P=Powersave Mode, S=Short GI, T=STBC, M=MU Beamforming, A=Associated, U=Authenticate.

I then SSH into the router and perform iperf3 -c from the router to the TV. Results: 250 Mbps (transferring from router to TV). When i do a Reverse, Result: 185 Mbps (transferring from TV to router).

I would have expected maybe a 30% to 40% drop from the negotiated Link Speed, so transfer speeds would be 450+ Mbps, but it appears im getting nearly 70% drop in speeds producing a transfer rate of 250 Mbps.

Using WiFi Radar within the Router, i see very little to no interference in the chosen band / channel. See below. My router's AP is "MuffinMan_SRC".

For the record i moved the router to 5 feet away from the TV and was still getting the exact same transfer speeds.

My questions:

1. Could the "slow" speed be due to the Smart TV operating at 40 Mhz channel width? Besides the link speeds stated by the router i am unable to conclusively determine what channel width the smart TV is actually accepting.

2. Can the few competing APs, as shown below, have that much of an affect on my transfer speeds?

 

[OzarkEdge]

I recently purchased a Sony Smart TV (KD75X800H) --Specs: 802.11ac, 2x2 MU-MIMO. My router is an Asus RT-AC86U -- Specs: 802.11ac, 4x4 MU-MIMO, 1024 QAM, 80 Mhz Channel Width.

Within the router's Wireless Logs i can see a Link Speed of 780 / 866 Mhz (Rx / Tx), RSSI = -60 dBm. The router is about 25 feet away, no obstructions. Asus Wireless Logs state the following FLAGS are active: P=Powersave Mode, S=Short GI, T=STBC, M=MU Beamforming, A=Associated, U=Authenticate.

I then SSH into the router and perform iperf3 -c from the router to the TV. Results: 250 Mbps (transferring from router to TV). When i do a Reverse, Result: 185 Mbps (transferring from TV to router).

I would have expected maybe a 30% to 40% drop from the negotiated Link Speed, so transfer speeds would be 450+ Mbps, but it appears im getting nearly 70% drop in speeds producing a transfer rate of 250 Mbps.

Using WiFi Radar within the Router, i see very little to no interference in the chosen band / channel. See below. My router's AP is "MuffinMan_SRC".

For the record i moved the router to 5 feet away from the TV and was still getting the exact same transfer speeds.

My questions:

1. Could the "slow" speed be due to the Smart TV operating at 40 Mhz channel width? Besides the link speeds stated by the router i am unable to conclusively determine what channel width the smart TV is actually accepting.

2. Can the few competing APs, as shown below, have that much of an affect on my transfer speeds?

In the bottom pic, what are the MACs/clients connecting at link rate 433? If one of those is the SONY TV, then where is its second antenna/stream?

You could disable Airtime Fairness and Universal Beamforming per band on the router under Wireless\Professional.

Otherwise, I don't see a problem other than your speed concern... keep in mind that WiFi is half-duplex and with some overhead, subject to interference (but I don't see any).

I wonder if the power save mode means the TV is not active and throttled down?

In my experience, SONY's quality control is pretty darn good, if that helps.

OE

 

[L&LD]

I can't see how you're running iperf from the TV?

Can you compare what it's capable of over a wired Ethernet connection with quality Cat5e or better cable?

 

[copperhead]

In the bottom pic, what are the MACs/clients connecting at link rate 433? If one of those is the SONY TV, then where is its second antenna/stream?

You could disable Airtime Fairness and Universal Beamforming per band on the router under Wireless\Professional.

Otherwise, I don't see a problem other than your speed concern... keep in mind that WiFi is half-duplex and with some overhead, subject to interference (but I don't see any).

I wonder if the power save mode means the TV is not active and throttled down?

In my experience, SONY's quality control is pretty darn good, if that helps.

OE

Apologies about the last pic above. The MACs are other devices NOT the Sony TV. Two are mobile phones and the other is a tablet. I dont know why the TV isnt showing up. Prob it went back into power save mode right when i snapped the photo. Sheesh. Anyway i will try disabling those options suggested.

 

[copperhead]

I can't see how you're running iperf from the TV?

Can you compare what it's capable of over a wired Ethernet connection with quality Cat5e or better cable?

To run iperf im using the app, Analiti. It runs on the TV; i select iperf Server. I've tried running as a client but running as a server allows me to SSH into the router to then run the command:

iperf3 -c 192.168.1.200 -w 300K
iperf3 -c 192.168.1.200 -w 300K -R

where the stated IP is the TV's IP address.

Good idea to test the speeds of ethernet directly. I'll give it a try

 

[L&LD]

Maybe the TV's CPU is too weak for anything faster, over Wi-Fi, then?

Even if the Ethernet shows GbE speeds, the CPU is suspect here (Ethernet ports are rarely CPU bound).

 

[copperhead]

UPDATE

As reference i executed iperf from Router to PC (the path uses CAT6a ethernet completely from Router to Switch to PC) for the following result:

Ethernet connected Router to Ethernet connected PC RESULT = 937 Mbps

Next, i attached Cat5e ethernet cable from Router to TV; I do not have long enough CAT6a cabling. WiFi on the TV has been disabled. Running iperf on the router to the TV over ethernet produced the follow result:

Ethernet connected Router to Ethernet connected TV RESULT = 90 Mbps

I exchanged the CAT 5e cable with a different CAT5e cable to get the same result, 90 Mbps.

Now i'm totally confused!

For OzarkEdge, i disabled Airtime Fairness and Universal Beamforming for both 2.4 and 5 Ghz. Here are updates from WiFi Radar. MuffinMan_SRC is my AP. I disabled every other 5 Ghz device within my home and only left the single Sony TV connected via Wifi

 

[copperhead]

And for the record here is iperf3 again but the TV is back to being connected wirelessly (ethernet removed). Airtime Fairness and Universal Beamforming are disabled. iperf results:

Router communicating Wireless to Sony TV RESULT = 290 Mbps

Wireless Log shows:

Note: The changes in IP for the Sony TV (MAC: 64:FF:0A:7F:7A:90) are due to me switching between ethernet (192.168.1.145) and wifi (192.168.1.120).

 

[L&LD]

The TV is only rated for 100Mbps Ethernet.

Consider yourself lucky that the AC Wi-Fi is over 3x faster.

Not only is the switch (Ethernet Port) limited, but it seems like the AC wireless is limited too by the CPU. I don't think there is much more to do here.

 

[OzarkEdge]

The TV is only rated for 100Mbps Ethernet.

Consider yourself lucky that the AC Wi-Fi is over 3x faster.

Not only is the switch (Ethernet Port) limited, but it seems like the AC wireless is limited too by the CPU. I don't think there is much more to do here.

I agree. Seems as good as it will get. 300 Mbps capped by the TV and not by the WiFi connection should be good enough.

One other possibility is the location of the antennas relative to the TV circuits. You could try moving the router to the other sides of the TV to see if you get higher throughput.

When the TV is wired direct, you can check the router webUI Status to see the LAN port link rate... we are expecting you will see 100 Mbps, not 1 Gbps. If so, then SONY thinks you will be fine with 100 Mbps full duplex.

OE

 

[copperhead]

Terrific responses!

I completely agree that the CPU capabilities must be at fault and the TV's ethernet port is 100 Mbps. It all makes sense. Being in the year 2020 i made too many assumptions about newly released technology.

Much appreciated, OzarkEdge and L&LD!

 

[doczenith1]

Try the "-w 4M" option in iPerf. I found that my wifi speeds improved with that option.

This is a loooonnnng shot but I wonder if your TV is somehow experiencing what is being seen on the new Chromecast with Google TV device. As with your TV my CCwGTV is reporting a 866 link speed but when running speed tests (with Analiti) I only get around 300 Mbps both ways. I seem to remember running iPerf and the results were lower than expected. I have since connected the CCwGTV via gigabit ethernet and speed test speeds are no better than wifi. Someone on Reddit mentioned that this could be a result of some internal hardware being USB 2.0 which would limit the transfer rates.

 

[bones5050]

Speed aside for a minute, are you experiencing any issues with playing shows on Netflix, or any other streaming service? If not, I would worry about it.

 

[cooloutac]

thats a super fast tv. wow. i got a couple LG's and a vizio and they only do 50 (vizio) or 100 MBs (LG)

 

[drabisan]

Unless you're seeing a problem (like Netflix with Atmos chocking), I would say you're overthinking the problem here!
My Sony TV is wired to the router, they are 50cm apart and link is doing "only" 100mbps. That "only" has no problems in serving Netflix 4k with Atmos.
Even on a busy wireless environment, your connection rates should be more than enough. All decently written applications will buffer enough to overcome any short glitches. Just take a look on how traffic from router to TV is having some 1-2s spikes every ~10 seconds while rest being virtually 0.

 

[copperhead]

Definitely streaming from the internet is just fine. I just wanted to make sure whatever i wanted to watch or do within my LAN was also possible at top speed, as i'm already used to between PCs over ethernet.

I never realized so many Smart TVs only achieved 50 to 100 Mbps. I errored imagining all 2020 devices would only be a few rungs down from top speed as opposed to reality of being just good enough to satisfy a particular want. Not sure what i was thinking. But lesson learned.