Localized interference?

I seem to be running across more and more head scratchers as the weeks go on.

I have three customers on three different access points that are all exhibiting similar behavior, two of which have it happen around the same time of day. These are 2.4GHz Lite SMs, running 8.1.5 or 8.2.4.

One SM will go idle most of the day - it starts around 7 or 8am in the morning and remains unable to connect until 6 ot 7pm in the evening. This SM is just over a half mile from the AP with a perfect line of sight and no fresnel obstructions. It was recently switched to 2.4 after having the same problem with a 900 unit. Fortunately this is a branch office that isn’t using the connection right now.

The other two SMs happen to be paying customers. Both will start going idle in the early evening hours and will usually remain down until early morning (around 6am).

I find it very interesting because the noise floor on these APs is -78 or less and these SMs, when registered have signal levels of -64 or better. When the issues start their signal levels instantly lose about 20db and they begin to drop.

My only conclusion is it has to be localized interference, but the real question is what kind of equipment would cause this behavior?

Thanks in advance.

we wasted days on similar scenarios.

First scenario, the state highway department installed a MASSIVE MW unit 100 yards from the client. We tried everything we could think of on all freqs. We had to decline service to the customer.

Second scenario, the city contracted an out of town company to install moto gear all over the town. Mostly backhauls and many 1 AP to 2 SM links to tie in all the city offices. The work was poorly done: No SSA, no GPS timing, big reflectors for 1 mile links, etc. In the end, we had to change freqs and approach the mayor regarding the situation. With much resistance, we put together an ISM freq map and TRY to coordinate with others before new gear is installed.

Third scenario, we had a client who ran a wood mill. All was fine until he installed an new saw. Every time that thing kicked on, the radio started re-regging. After a day of troubleshooting, we relocated the equipment off of his tin building on onto a wooden utility pole, but by then, the customer was so mad at us (even though not our prob) he switched to satellite. Last I heard he is now worse off.

I should point out that most of our problems occurred when these units were active. So we had the opposite problem, issues at day and not at night. The power levels and jitters varied dramatically depending on the time of day.

So, look for big dishes… grab a SM and battery and scan the neighborhood, be sure to NOT broadcast your APs so competitors will be less likely to mess with you, and watch your EMI proximity and be sure to have good grounding.

Appreciate the feedback, thank you.

I installed an advanced antenna reflector at one of the clients and his modem hasn’t re-reg’d since. The non-busy unit (located at our parent company’s truck shop) has stabilized for the time being and the third client has intermittent issues with it, meaning it doesn’t consistently happen daily.

I know I’ve seen issues like this before but they’ve never actually caused this kind of havoc for the client. One that comes to mind is a seasonal client (beet piler, to be specific). Their piling site was located a few hundred feet from an RTK tower (these towers are used for GPS applications on farm equipment). The uplink efficiency got hit pretty good (around 60%) but the link never dropped and seemed to work pretty good. Interesting, nonetheless.

I’ll keep my eyes peeled for possible interference sources.

Logic (in my mind) dictates that installing big reflectors or high gain dishes means less interference. The higher the gain on the dish the smaller the V and H beamwidths are… This means less signal spewing out all over the place and a more direct shot between the two radios. Switch to shrouded antennas and you magnify this even further.


Deployment of un-synched Canopy systems is a big problem though, even more so if they didn’t take the time to analyze the RF environment properly and take the time to coordinate the frequencies. Then again, we are not license holders in the band that all of this equipment operates on so I guess it comes back to just deal with it.

I sympathize with you. I know how it is…