Recently, I had some issues with download/upload speed on my WAN. Due to this problems which I solved with my ISP, I have noticed that I have a low upload speed via WiFi on some clients.
My nominal WAN speed should be 100/100 Mbps.
Testing against Netmetr via reForis shows mostly 96/95 Mbps.
Testing via Netmetr.cz Android app with my Galaxy Note 9 shows 94/94 Mbps.
On PC 1 (notebook) connected by WiFi, testing via Firefox against www.netmetr.cz web test shows 96/1-6 Mbps (!!!).
On PC 2 (desktop) connected by WiFi, testing via Firefox against www.netmetr.cz web test shows 96/1-6 Mbps (!!!).
On PC 3 (desktop) connected by LAN, testing via Firefox against www.netmetr.cz web test shows 96/96 Mbps.
The problem didn’t exist about three weeks ago when I changed a position of TO due to poor WiFi signal at my PC 2. I measured the link speed with www.rychlost.cz and I had 94/90 Mbps.
I’m using 5 GHz WiFi only, the ac mode, 80 MHz channel width and I think is is somehow a software problem either of TO configuration or my client configurations.
PC 1 (WiFi adapter: Marvell AVASTAR Wireless-AC Network Controller)
See my updated question. The numbers are quite consistent for download and inconsistent for upload. The hardware of my PCs is irrelevant, it didn’t change in last three years.
the problem is therefore only on wifi for some combination of client and measuring server. How is the WMM parameter set in general setup wifi, is ON? Has he tried a narrower bandwidth than 80MHz? I generally think that this is not a problem with the router, because the error is caused by changing the other side of the measurement … the measurement server.
My sleep is one thing but a fluent conference calls is the other thing which I care about. I spend a half of my working day on conference calls I need undisturbed interaction with others.
I have not yet registered a mention of operational problems upload video. I’ve been concentrating on your speed tests, where the cause may be unclear.
Upload problem with video conferences always occurs and with all talking ? 1- If you have insufficient upload, the counterparty should have problems watching your video? 2 - Or is the video cut to you? What geographic location your partners are in?
How exactly do you have a router located in a new place - height from the ground, distance from the wall ? The antennas are in position… \ | / … Is the position optimal?
What is the location of problematic PC1 and PC2 relative to the router … number of walls and ceilings and distance ?
What 5GHz channel do you use … non DFS ?
I also suggest trying band 2.4 as it has a larger range and penetrates the walls better.
I have not yet registered a mention of operational problems upload video. I’ve been concentrating on your speed tests, where the cause may be unclear.
Speed tests brought my attention to it.
Upload problem with video conferences always occurs and with all talking ? 1- If you have insufficient upload, the counterparty should have problems watching your video? 2 - Or is the video cut to you? What geographic location your partners are in?
Mostly Prague (me Brno), it happens when I am speaking. Both video and audio is jerky. Audio not so much when camera is off but I don’t want to turn it off.
How exactly do you have a router located in a new place - height from the ground, distance from the wall ? The antennas are in position… \ | / … Is the position optimal?
I use | | | antenna setup. I changed the position from behind TV where antennas where close to wall to next to TV where antennas are further from the wall and have basically open space around them. The upload was fine in the current position. That is the reason why I know that something is wrong. I run the speed tests after I changed the router’s position to the current one.
What is the location of problematic PC1 and PC2 relative to the router … number of walls and ceilings and distance ?
The PC2 & mobile had problems with loosing the connection about 8 m away from the router in my study. There is one wall between PC2 and the router. There is a kitchen equipment on the one side of the wall, a wardrobe on the other side of the wall.
I also suggest trying band 2.4 as it has a larger range and penetrates the walls better.
The 2,4 GHz band is just slower on download and has a very same problem with the upload. The numbers in upload are basically same, just download speed is about the half of the 5 GHz band.
I would start by changing the position of the router (only for the test), because it is the easiest. The router should be higher on the ground 1.5-2 and next 30 cm to the wall. Placing next to the TV can be a problem, a change of position will confirm or prove it. Antennas should be \ | / because reflections in the room are omnidirectional - it probably won’t be a major influence.
!!! If you claim to see a video jerky (calling partner) then it’s not a problem of your upload, it’s a partner upload. If the caller (counterparty) has some metallic (telephone line - asymmetric connection for example phone operator xDLS)- then there may be a problem there.
I think … if you think you have a limited upload you - you would have a fine image, but your partner would see you jerkily.
Upload problem with video conferences always occurs and with all talking (in multiple locations - workplace) ? Let’s try a video call with me (skype, messenger, etc.) … send private message.
First of all, the problem is really only on my site, in upload through WiFi. Even when I download/upload files to my NAS, upload is three times slower than download.
Four days ago, I changed the position from | | | to \ | /. At the beginning the numbers & Teams behavior looked good. Next day, the upload was even worse (measured on www.netmetr.cz). The upload was almost zero and websites responsibility felt that way.
Next day, my 5GHz WiFi collapsed during conference call. Since the collapse, the speed is where it was before the antennas change. On the other hand, the coverage is better and my upstream to Teams started to be more fluent.
Test on Speedtest server is great. Other measuring servers that give you bad results are not relevant. The best results is valid (Speedtest) !!! It can be seen that the speed does not limited your local WLAN !!!
You may have to put together a speed test against a server in an unknown location and with unknown parameters with an operating speed against other entities. What Speedtest show you your technical maximum.
I can check my own wifi speed with FTP (attention, not SFTP) from a Windows PC to a mobile phone or Synology NAS. I use Google Play app “Ftp Server” from The Olive Tree. Any other internet connection may not reach the speed of my WLAN.
Each speed in traffic is affected by three (four) factors: the properties of the start and end point (for example, slow write to SD card storage ). … WAN route (plus your wifi).
You probably can’t influence anything outside your network and you can’t compare it to a connection to Speedtest. Therefore, do not focus on your wifi, which is obviously perfectly fine as you did to prove yourself.
(skype, messenger, etc.) … send private message.