Go Down

Topic: NRF24L01 - Questions about creating Interference (Read 2490 times) previous topic - next topic

jamiec

Hi,

I am new to the forums but not new to electronics.
I have a project I am working on that uses an NRF24L01 and I have a strange issue that when in a certain building (a shopping mall with loads of wifi) the NRF24L01 module intermittently crashes.

I have used a Spectrum Analyser and there is so much noise in the area, however I am finding it extremely difficult to replicate this in the lab.
I was wondering if anyone had any way to use the NRF24L01 to generate just random noise and interfere with other NRF24L01 modules. Just to see if I can crash my system and then I can implement a fix.

I know its illegal to Jam signals but It would be done outside in the middle of nowhere in a faraday cage so that no external noise would interfere with my findings.

Thanks
Jamie

jremington

#1
Nov 15, 2017, 05:55 pm Last Edit: Nov 15, 2017, 07:35 pm by jremington
Quote
the NRF24L01 module intermittently crashes.
What do you mean be "crashes"?
Could you describe what happens in more detail?

It is hard to imagine that radio interference could cause a problem other than disrupting communications. Wifi works in the mall, doesn't it? More likely, there is a power supply problem.

MarkT

From what I remember the chip is a bit fussy and the real state transitions aren't quite as documented,
but I think you can always recover by powering down briefly?  This is all half-remembered from a logn
time ago and I may be thinking of an entirely different chip.

If you want to fuzz(*) a receiver like this try to set it to the most promiscuous mode (minimal preamble,
minimal address size) and select a modulation that works reliably so you know the fuzzed packets
are being read and as few as possible are being filtered out early.

I suspect you can't use a 24L01 to fuzz a 24L01 due to the rigid outgoing packet structure, a more
general piece of transmit hardware is needed, but again set to the same modulation type and parameters.

Perhaps the fact the interference isn't from a 24L01 is leading to hitting untested corners of the chip's
packet handling state-machine?

(*) https://en.wikipedia.org/wiki/Fuzzing

Personally I've experienced microwave oven interference, but in a setup where auto-acknowledgment
was disabled.  Have you tried no auto-acks?  It may bypass the issue as I've never had a lock up,
although microwave oven isn't really what you've been affected by.
[ I will NOT respond to personal messages, I WILL delete them, use the forum please ]

TonyWilk

Dunno if this helps, but having worked with some RF stuff before, the first try is to remove the antennas from whatever transmitter/receiver you are using so the working range often down to the length of your bench.

You can then move them apart to find the range where they get 'iffy'.

A slightly more scientific approach is to fit RF attenuators to one or both communicating units. If you are working on one end, the 'other end' is fitted with an attenuator (or has it's antenna removed) so your unit is still 'sensitive'. Then, if you don't have any specific test gear, use something like an active WiFi router and put it right next to your unit.

There are "rf noise generator" (google) circuits out there and you can buy RF noise generator boards - even cheap ones:
https://www.ebay.co.uk/itm/SMA-noise-source-Simple-spectrum-external-tracking-source-DC-12V-R3O4-B1O7/263270616777

which you can use to force a level of noise into a receiver (you really need a bunch of attenuators and a spec. analyser to mess about like this).

Just a few thoughts there... good luck :)

Yours,
  TonyWilk
 

Go Up