The tower guys cell phone?

Ken Hohhof wrote:
At this tower with FSK, we knew there was another WISP (let's call them ABC) with an AP 9 miles away LOS on 915 MHz. And we always got the best results if we also went on 915 MHz and matched their timing parameters, rather than trying to avoid them.

After experimenting with the 450i equipment, I am convinced that another WISP (let's call them XYZ) is operating 900 FSK APs in our area with AP Eval Data turned off. I know we have a competitor who thinks it's cute to do that. I think that's a douchebag thing to do.

Anyway, I'm also convinced that while ABC and us have our APs pointed at each other, we may be seeing XYZ's SMs as much as their APs. Coexistence in 900 MHz is tough because free space loss is low so LOS signal carries a long distance, plus NLOS signal gets through trees and even scatters off them, plus even the highest gain antennas don't have narrow beams. You can't count on SMs using narrow beam antennas pointed only at their own tower.

So while smartgrid and RTK are certainly contributors to the noise level, I think we have XYZ using the bottom and top of the band but in stealth mode. Probably they see ABC on 915 and avoid them. So we are able to coordinate timing parameters with ABC and coexist in the traditional manner on the same channel.

It is possible that we are seeing SMs on other ABC sectors pointed away from us, but we are only 135 feet AGL, and I don't think their subs would be high enough that we would see them 10 miles away, given earth curvature.

So the next hurdle was setting the timing parameters on the 450i to match ABC's PMP100. For some reason there are checkboxes for what firmware revision the PMP100 is running, and whether it uses timing port or power port sync. We did not previously have to know this much detail, I wonder why. Anyway, I finally hit on a combination that worked: 5 ms frame, numbers from the Cambium compatibility tool, and Legacy Framing Mode off. At this point, I was able to get the SMs to register at 915 MHz and run MIMO-B in a 7 MHz channel. Actually, I started by putting the FSK timing parameters into the compatibility tool (75%, 10 miles, 3 slots) and it said OK without any futzing around.

Some other observations:

1) You can power the AP with power port sync off a CTM-2 in 56V OFDM mode using an adapter cable with pins 5 and 7 swapped, as described in the manual for a CMM4. I assume the same is true for a 430 SyncInjector.

2) According to the CTM-2, the AP is drawing 10 watts (we are not using the AUX port to power anything), so don't get scared by the 30 watt number.

3) The AP and SM are capable of up to 25 dBm xmt power, but with 12 and 13 dBi antennas it will be slightly less due to regulatory limits. Check the AP config to make sure it is set to the max you are allowed, I think mine started out set to 22 dBm. Many WISPs no doubt are using ~17 dBi yagis or grids without turning down the EIRP to 36 dBm, if so, a legal 450i system will be at a disadvantage by comparison although antenna gain doesn't always help through foliage.

4) If you leave all the channels and widths enabled, expect it to take forever for an SM to register. You need to uncheck some of the boxes.

5) If you do an SA when you first install the AP and see -35 at 900 MHz, it may be your tower guy's cellphone.

6)  The AP has no internal GPS.

7) 16 inch coax jumpers will be a little tight, 18 inches or longer will be better (assuming LMR240).

8) The AP comes with one gland which looks like the LPU grounding gland but it has a rubber insert which is already slit and you can get a pre-terminated cable through it. It doesn't come with a second gland for the AUX port.

9) The sector requires lots of different wrench sizes, I think the list is 10, 12 and 14 mm.

10) Cambium has made installing the CPE equipment a piece of cake. Basically take the pre-assembled antenna out of the box, snap on the SM, attach the SMA connectors and slide the boots down, the only assembly is the U-bolts. It's very nice.




Reply via email to