just another IDLE sm post :)

We have a ton of Canopy gear out there. Most is 900Mhz. We feel like we have seen it all and been to every Rodeo in town with Canopy over the last 4 years but are a bit stumped on an issue.

Shame on us but were cheating at the moment using a 900mhz SM module as a backhaul to (1) 900 Mhz AP. The subscriber module is on 922Mhz connected to the tower and the 900Mhz AP (set to generate synch) is facing the opposite direction on a utility pole with only about 15ft of seperation. We plan to populate this pole with 3 APs total and stick a 2.4 or 5.8 backhaul in once we have some customers in this area. The SM is connected at -54 and solid as a rock while the AP blaring behind it runs on 906Mhz. The SM acting as a backhaul never drops thru any of this trouble.

We have connected a customer about 1.75 miles away to the AP on the pole and his numbers are -80 RSSI 1340 jitter 4/5. Some random weirdness is going on. The customer unit goes idle at random times. maybe 30 minutes maybe 3 minutes. Pings are flawless until this happens, link tests are perfect in 1x rate. jitter & rssi are consitent always. The session count on the AP goes up every time this happens but there are zero re-regs. currenlty (after 5 hours), there are 275 sessions and zero re-reg’s).

Usually when we have a crummy link or there is interferrence, we see re-reg’s. The customer had no ethernet plugged in as this is a test site so we went over and jacked it into a switch on a private network. We read a post about the units going IDLE from inactivity. Havent found this to be true or false yet. After bringing up the ethernet, the problem persists.

The last time we had a similar problem to this we found download percentages & max distances were all mixed up on a full cluster. Even though it was synch’d on a cmm, it was doing this same thing. Setting those values alike fixed the issue. We just cant do that in this situation since its only one AP.

Is this gonna werk or should I bit the bullet and plop in a BH20? It would make a great rollout for us to go out and stick in an ap and an sm on a utility pole to serve a small development, then when revenue warrants, replace the sm with a true backhaul.

Were using a stock 900SM but have SMC and yagi’s if it would narrow things down.

Also we had a newbie tech wire my last cmm WITHOUT the power supply and explode 1100 bucks. In a 911 pinch, can you rig a synch cable in a way to daisey chain APs in synch? I read you could in a post but never found out how. I know how to make a traditional synch cable to go from AP to AP and synch 2 aps but how about 6?

Any help or suggestions would be appreciated.

Tim

Hang on there Bullrider! No shame in what you are doing.

900MHz will self interfere if they are less than 10 miles apart.

Carry the sync (RJ11 pins 1, 2, 3, 6) from the SM to the AP on the lightpole. Set the AP to rcv on timing port. This will put the solo AP on the same sync timing as the cluster feeding it.

Put the single AP on a different frequency than the AP feeding it. Non overlapping are 906, 915, and 924 so ideally the AP feeding is at 906 and the solo AP is at 924.

Set DL%, Max distance, and Control Slots the same on the solo AP as the AP’s in the cluster.

See if that helps.

thanks - I shot you a PM with a “book” sorry. Im really twisted in my logic of how I am supposed to be synching multiple towers in close proximity. I think I have a uge mess :frowning:

I, like troberts have set up similar temporary sites but typically using out of band backhauls. My question is about sync::

I have, at least in my own mind, determined that sync from the backhaul at the remote site coming from the backhaul source is not sync that is exactly in sync with the originating site. How can it be?? There is at least a small amount of latency over the backhaul that is carrying that sync. When the new remote site is in range of the original source site, I have installed CMMs or equivilant sync sources at the new remote site. Is my logic correct? What a mouthfull!!!

And since we are on the subject of sync sources; Are the non-Motorola sync sources exactly in sync with the CMM? According to my dual trace oscilloscope, they are. Any thoughts or comments here?

You are right. There is a bit of air delay so the remote units would be slightly out of sync from the first tower. An actual sync source is ideal.

I spoke with the engineer at PacketFlux that designed the sync pipes and they are designed to be in sync with a CMM

troberts@dsslink.net wrote:


We have connected a customer about 1.75 miles away to the AP on the pole and his numbers are -80 RSSI 1340 jitter 4/5. Some random weirdness is going on. The customer unit goes idle at random times. maybe 30 minutes maybe 3 minutes. Pings are flawless until this happens, link tests are perfect in 1x rate. jitter & rssi are consitent always. The session count on the AP goes up every time this happens but there are zero re-regs. currenlty (after 5 hours), there are 275 sessions and zero re-reg's).



Tim

hey Tim, I'm probably gonna make a fool of myself here, but every time I have seen this problem without the re-reg it turned out to be power fluctuations to the SM, from suspect cat 5 or Sub-par terminations. its an easy fix at least and cheap to test and doesnt take a tech more than ten minutes, give it a go :)

I’ve been away for over a year, the 900 SM’s are still having IDLE issues/Reregs? YIkes.