Forum

We are having a problem at one of our customer sites, which I was hoping someone may have some idea about...

Ever since a big wind storm this weekend, this customer has been receiving only -87 dBm signal. Originally, he was getting somewhere in the -70's. I think it was around -78, but I'd have to pull his paperwork to be certain.

This is a 2.4GHz setup on channel 11. Our end is broadcasting at 36 dBm. He has a clear line of sight.

http://www.kayandee.net/photos.php?folder=gallery/Wireless&image=gallery/Wireless/customer_site.jpg&width=640
(Ignore the fact that this is on my wedding website.)

The big silo in the background is what our AP is on.

We replaced his radio (tried 3 total). The radios all have an integrated 19 dBm antenna. He is about 1.5 miles from our tower, as the crow flies. We replaced his PoE injector, and we verified the Cat5e cable up to 100Mbps using a Pentascanner.

We have other people on this AP that have no issues. In fact, we happen to have someone about 1.5 miles further out on amost exactly the same line of sight. That person is using the same kind of radio with a built-in 19 dB antenna and is receiving a -72 dBm signal.

I used an Avcom PSA 1727B spectrum analyzer with a 2 dBi yagi and received a -74 dBm signal at the top of his antenna tower for channel 11. I also scanned for channel 11 signals using an 802.11 protocol analyzer to make sure that I was really seeing our AP on the PSA. I saw two other access points on channel 1 and another on channel 6, but nothing else on channel 11.

We also turned off his cordless phones, removed their battery packs, shut down his wireless mouse, turned off his wireless router (which is on channel 6), and shut off his computers that have wireless cards. On the ground, the PSA only saw channels 1 and 6 (coming from the neighbors' houses).

To eliminate the possibility of a multipath problem, we tried moving the CPE to the other end of his roof. That didn't help either.

Does anyone have any other suggestions?

Being that this is a weird problem, some of my answers may be as equally weird.

- First, replace the Cat5. Even if you just run it through a window. When you tested the cable, I'm sure it tests on all 4 pairs, but the throughput only tests the green and orange pair. Its a cheap thing to check.

- When you say you are replacing the radio, I'm assuming that includes the pigtail and antenna. If not, replace everything

- Move the antenna back to where it was. Its good to get back to the baseline

- I figure that it is an omni at your AP. Can an antenna be damaged in such a way to lose some of its omnidirectionalness? :) I told you some of my answers would be weird.

Let us know what you come up with.

By the way... where are you located... if you don't mind updating your profile.

Thanks for the suggestions, but they don't really apply.

First, the cable tester we used does not just test pinouts. It tests near-end and far-end cross talk, impedence, etc. It actually certifies the cable as capable of transmitting data at 100 Mbps. It has been fully certified up to the Cat5e standard at 350 MHz.

Second, the antenna is integrated into the radio. There is no pigtail or external antenna.

Third, we did not leave the radio at another location. We just moved it to another location temporarily in order to eliminate the possiblity that a multipath signal was causing a nullification effect which could have been responsible for obtaining such a low signal.

Any other suggestions out there?

Here's a stab at this anomaly.

Go back to your source AP at the Silo.

BTW, What kind is it and what kind of antennas are they using, Are they internal?

Check the positioning of that AP and antennas as it relates to the original position when it was first optimized.

Perhaps the wind storm shifted it a bit off course.

for peace of mind sake realign it

I actually have someone going to check on that today. The APs are internal radios that are in a NEMA enclosure with a small heater for the winter and a good vent for the summer. They run coax up to a 3-sector omni (120 degrees per sector). Theoretically, if the antennas turned, they should receive the same coverage. But I have someone going to make sure the alignment hasn't changed for (like you said) peace of mind.

On Thursday, if the problem is not solved after inspecting our access points, we will be taking our bucket truck out to the site (since our only licensed driver will be back on Thursday) and we'll put an atenna up in the air at about 45'. That's higher than the little trees seen in the picture I posted the link to. We'll hook that antenna into our PSA 1727B spectrum analyzer and take a signal reading. Then, we'll slowly lower the antenna and watch to see if the signal changes. We'll also try another radio on that pole at that height to see if it gets a better signal than on the antenna tower. Then, we'll move the truck over to a different location on his property and make another measurement. Then we'll move the truck straight East, out to the main road, and take another one. I'm hoping this will shed some light on what is going on, but if anyone else has any ideas, I'd love to hear them. I'm quickly running out of ideas.

Man... that would have been great to know that your antennas were sectors and not omnis...