View Single Post
  #15  
Old April 8th 18, 08:06 PM posted to alt.comp.os.windows-10
Paul[_32_]
external usenet poster
 
Posts: 11,873
Default Really slow wifi

Chris wrote:
On 07/04/2018 13:00, Frank Slootweg wrote:
Chris wrote:
Hi all,

Am a longtime computer user, but new to win10 as I've just built a cheap
gaming machine. My last extensive exposure to windows was win98, so
please be gentle. My experience since then has been linux (mostly)
and Macs.

This is a home machine running Win10 Pro version 1709. Rest of the spec
is an intel i3 8100 (coffee lake), 8GB RAM, nVidia 3GB GTX1060, a
Samsung 820 64GB SSD and a 1TB HDD. I have an Asus PCE-N15 Wireless
PCI-E card for the wifi.

The problem is that the networking is crushingly slow on this machine
(i.e. measured in low kilobits per sec). Every other device in the house
incl. smart TVs, Mac laptops, linux desktops, phones, etc. have really
good connectivity (measured in megabits per sec) - they are quite happy
streaming Netflix etc. However, this PC stuggles to even refresh
newsgroup headers or load googe.com in a browser.

Where do I find where the wireless device drivers are so that I try and
tweak them? Where else can I look to try and boost the networking on
this machine? Or alternatively, what could be interferring with the
internet speed?

Given the location of the router relative to the PC ethernet is not an
option, so please don't go down that road.

Other devices (incl. the linux PC which this windows machine is
replacing) have no problems using wireless in the same room as the PC.
The PC seems throttled somehow and I'm not sure where...

Any help/pointers happily received.


Long shot:

5Ghz or 2.4GHz Wi-Fi (in use)?


Yes. The router is dual band. Not sure about the card. No mention of
5/2.4 Ghz, but it does mention 20/40MHz - not sure what that means?

If 2.4GHz, does the system have Bluetooth? If yes, turn Bluetooth off.


No bluetooth.


The 20/40MHz is the channel width. It affects data rate. The standards
require the narrower setting, as a function of "green field" - if
older Wifi standards are detected "in the air", newer standards are expected
not to screw up the older standards. And then the channel width
should be 20MHz at the driver level.

https://en.wikipedia.org/wiki/802.11n

If you're on a farm and away from other Wifi setups, then you
may be able to actually configure Wifi gear for the best
possible performance using the standard you've selected.

No matter what Wifi standard you use, they hardly ever switch up
to the highest performing mode seen in a Wikipedia article. Even
when the manufacturer makes a concerted effort and not a half-hearted
effort :-) Devices like this for example, may be dual band
and MIMO, and close to the maximum number of antennas for
the max spatial in the standard. Having this many antennas,
a "cheese grater", is like an "Offering to the Gods of Wifi"
that your harvest will be good.

http://www.networkbuzz.org/wp-conten...3123d28953.jpg

*******

I like the graphic in this article, because it shows you the
nature of Wifi propagation into other rooms. Constructive
and destructive interference, multipath etc.

http://www.engadget.com/2014/09/01/w...nal-reception/

Many times, too much signal is just as bad as not enough. With
too much signal, it encourages the multipath interference,
and that's what throws the signal off. The smallnetbuilder
site, I think they might have used an RF absorptive mat
in a lab, and got better reception. I don't know of a way though,
to easily build a map like in that picture - there are portable
sniffers, but you'd need a device that knew precisely where
it was in the floor plan when taking readings, to build
a map like that.

USB3 cables emit a broad signal with a peak energy at 2.5GHz.
Right smack in the middle of the lower Wifi band. As far as the
Wifi is concerned, the modulation is random and looks like
a noise floor. leaking USB3 cables can knock out a Bluetooth
mouse or keyboard (until you move the two apart).

Bluetooth (frequency hopping) is not supposed to interfere
with Wifi. (I would have to dig up an article for an
explanation of why it's not expected to interfere.) Bluetooth
changes channels 1200-1800 times a second or so. So as an
interference source, it doesn't stay absolutely steady
with respect to the more fixed channels on Wifi. The Wifi
defines a smaller set of channels, and then uses an even
smaller subset of the defined channel marks, in order that
the skirts of the channels not overlap. And then in North America,
there might be three channel allocations you can select
on a router. And you select the one that "doesn't seem to
have a lot of traffic on it".

In the sketchy pictures here, you can see 1,6,11,14 in the
diagram, with the intention to show that 1 and 6 don't
interfere, because the skirt is down enough at the edges.
North America and Europe have slightly different allocations,
and perhaps 14 is illegal in NA, and then the three
channels might be 1,6,11. You should have the option
to adjust your Wifi (somewhere) for one of the three
non-overlapping choices.

https://en.wikipedia.org/wiki/Wifi_channel

I don't use Wifi here, so cannot vouch for any tuning
procedures or give practical advice. The only Bluetooth
I had here, was the lowest class kind, and it barely
worked at five feet :-) And the Windows 10 software
for it, sucked (couldn't get it to set up a Piconet).
I can just imagine how a Wifi experiment would go here.

I already had an experience here, where someone down the
street from me, was using some electronics that wiped
out OTA TV reception on a couple channels. There's
really no way to know just what kind of a jammed
mess you're in, in any given residence. Maybe a leaking
microwave oven could do it, so your download stops,
every time you microwave a burrito.

Paul
Ads