Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
Shawn A
@tablatronix
I think I will just stick with the 3, trying to do anything more complicated without proper RF info from the radio is useless
You need actual SNR
liebman
@liebman
@tablatronix bandwidth is the same for all channels - you listed three non overlaping channels, usefull where you have multiple access points….
Shawn A
@tablatronix
yeah, thats the point
id say 90% of enviroments are going to be clustered on those
Michael Miller
@Makuna
But, if 90% are clustered on those, don't you want to avoid them ;-)
Mmm, I thought all channels overlap the next channel. Is this different between countries?
Well, I checked, you are correct, learn something new everyday.
Maybe I read that wrong, it states 1, 6, 11 don't overlap each other, that doesn't mean 6 doesn't overlap 5 and 7.
Michael Miller
@Makuna
Ok, I was correct. All channels overlap. Its just that channels more than five apart don't overlap each other, so to get three channels that don't overlap at all, its 1, 6, 11.
So doing anything to "detent" onto these channels doesn't seem very logically except in cases where there is very little WiFi use in your area.
I am seeing seven that aren't even mine ;-)
liebman
@liebman
image.png
Michael Miller
@Makuna
In US, there is no 12 and 13, we are limited legally to only 11.
liebman
@liebman
Correct as described in teh article
Shawn A
@tablatronix
11 22MHZ wide channels, 5 MHZ apart
some countries have channel 13 14 and they are offset
if you are trying to pick a empty channel it almost guaranteed to be 1,6,or 11, simply due to the fact that those are occupied channels in multi ap enviroments because of overlap, So I will focus on that and add dynamic later if its even needed, randomly picking other channels will not be any benefit and will only cause issues with other devices
Shawn A
@tablatronix
for US, it will differ by country code
@liebman how did you insert an image?
liebman
@liebman
on a Mac screen capture and paste
Shawn A
@tablatronix
Screen Shot 2019-09-13 at 3.11.14 PM
lol
image.png
oh cool
jaygeet
@jaygeet
hi i have some queries to share with you
Tejas H
@CodingCreate101
Just share it. Someone would respond eventually @jaygeet
Shawn A
@tablatronix
image.png
Does anyone know the internals of analogRead and how it works ? Apparently I cannot service wifi when using analogRead,
Shawn A
@tablatronix
It this some kind of vdd33_const RF Cal issue?
Shawn A
@tablatronix
Looks like ( at least in my app ) I cannot analogRead more than every 600micros
Michael Miller
@Makuna
On most uC ADCs take time to do the work. You usually can configure them, slower = more accurate, faster = less accurate. From my experience 600us doesn't seem that long for an analog read.
Develo
@devyte
also, our analogRead is a thin wrapper around the sdk function
I believe that the adc read actually does several reads and average them (not in our code). There's s fast read in the sdk api to do lots of reads into a buffer
Shawn A
@tablatronix
Yeah I am not finding any details on the SDK implementation, but that sounds about right, I was more interested in the overlap with wifi interrrupts
All specs are form observation, people testing sample rates etc
adc fast should be direct adc, but you cannot use with with wif
wifi the docs say , so the sdk is doing something to handle the adc read averaging and wifi in its black box
Develo
@devyte
the problem, per my understanding, is that wifi traffic injects noise into the adc reading. Originally, the adc would be fast, but would jump wildly depending on packet traffic. To overcome that, they implemented an averaging function over several samples. The downside is that it takes a while to do the whole sequence.
Shawn A
@tablatronix
what is odd is that I can run the adc way faster than that, so I am not sure overhead is what causes the wifi issues
Develo
@devyte
how are you running it?
Shawn A
@tablatronix
normally, analogRead
Develo
@devyte
no, I mean how are you calling it? how did you reach the conclusion about calling it so fast?
Shawn A
@tablatronix
Just in a loop and benchmarking it, then doing the same while pinging wifi and adding a micros eval timer
Shawn A
@tablatronix
maybe my benchmarking was flawed, let me check again
Hasenradball
@hasenradball
Hello guys I have a question regarding Wire.begin()....
what is the result if I do two times
Wire.begin(); Wire.begin()?
The beground is to instabciate two classes and each will call Wire.begin().