I have recently replaced my existing LocustWorld Mesh software with MikroTik RouterOS. During this process I also had to replace the existing Prism2 based cards with Atheros based CM9 radio cards, other than this the hardware has not changed, neither has the cabling nor antenna (8dB Omni).
Since migrating to RouterOS what was a stable Mesh system with good signal power and SNR has become unstable and I have signal power varying during the 24 hour cycle. During the day the AP in question is reporting around -67dB, but in the evening this is dropping to -77dB and the CCQ is dropping from the high 90s down to under 50.
The base unit is indoors so there should be little variation in the system’s operating temperatire.
The two clients connected to this AP are only around 400m away.
The system was running on channel 11, (2.462GHz). I have today switched to channel 6 to see if that makes any difference.
What is causing this? Is there anything I can do to resolve it? I’m based in the UK so installing a higher power card is not an option as I’m limited to 100mW (20dBi) EIRP.
Regards
Chris Macneill
Hi Chris
What sort of physical environment is this equipment located in? The CM9 is a very popular radio due to it’s low cost, but bear in mind that it was designed as a card for laptops and notebooks. We have had issues with CM9 radios in noisy environments - particularly in 2.4GHz. I suspect this is largely due to the amount of consumer-grade crap on the 2.4 frequencies these days (Wireless cameras, baby alarms, bluetooth, wireless toasters etc).
We’ve had good results from the Ubiquiti cards - others regard the R52 highly. Can you go to 5.x GHz? It’s far less noisy in my experience.
HTH
As the antenna is external we’ll be picking up any old rubbish from the surrounding houses.
As all the CPE equipment is 2.4GHz 802.11b only, the P2MP links have to be 2.4GHz. I have migrated one P2P link to 5.8GHz.
I’m just surprised at the vast drop in signal beween day and night, 3 to 6dB I might have expected, but not 20dB.
Thing is it’s the absolute power which is varying, the noise floor is staying constant. The SNR is only dropping because the absolute power has dropped. If the problem was due to extraneous noise I would expect the RX power to remain constant, noise floor to increase and SNR decrease proportionatly.
It’s almost as if the wireless card doesn’t like working in the dark!!
I bought some Senao cards that are equivalent to the R52, but I had a lot of problems with these crashing the router. I suppose I’ll have to buy a few R52s and try them.
In this day and age with all the standards that equipment is supposed to adhere to, why is there such variation in quality and performance between different products?