Color Codes and 900Mhz questions..

So we’re currently running a cluster of 6 of the model 9000APF and around 20 SM’s at the moment all on HWS. Our downlink is 70% on the AP and our Max range is set at 12 miles. Our farthest customer is 9.22 miles.
Our SM’s are running 7.2.9

That person’s statistics are:

Session Status REGISTERED
Registered AP 0a-00-3e-90-7f-4e
RSSI 1418 (-73 dBm)
Jitter 3
Air Delay 331 (approximately 9.22 miles (48657 feet))

There is quite a bit of trees in between our customers and our AP’s. In the small town we work out of in Florida, there is quite a bit of trees, but most of our shots don’t get that much interference. The RSSI is usually around 1300-1800 and the jitter is usually 2-4. We use 13dbi yagi’s for anything over about 3-4 miles. We have had pretty much no success with the integrated units over 4 miles. They would make a connection initially but after a matter of a few hours they would drop repeatedly.

After a while we realized our customers were bouncing AP to AP, so we locked them into the AP that had the best signal by unchecking all the other frequencies. I was reading another thread, and they said that wasn’t the way to go about it, instead we should use color codes and leave the frequencies open. Exactly why is this?

Another question: I don’t know whether I misunderstood or not, but I saw someone talking about using yagi antennas on their AP. Does that work at all? How much better?

Also, even with good jitter/RSSI our customers modules seem to just lose connection some times with the AP’s. Things will be running smoothly, then bam their module drops. It’s random times throughout the day/night. We have one customer who went down a total of 30 seconds in 4 days, then over the course of about 2 days she was down for about 12 hours, then it came back up no problem, a week later she’s back down now and we can’t get her to come back up so we’re going to order a 17dbi yagi antenna for better gain because it worked with another customer. We were assuming some kind of interference, but could not find anything.

Does anyone else have experience with customers just randomly dropping off? We’re trying to optimize our network obviously and I’ve found tons of useful information here but some answers lead to more questions for me =).

Thanks in advance.

Edit: Forgot to mention we have 2 multiplexed t1’s backing up our AP’s/SM’s on the bandwidth side. Our customers at 9 miles out are getting 2.3mb down and 700 up. Seems pretty good to me, does that look good to you? Now if I can just stop them from dropping signal. We use an intermapper console and it works great but like I said people just randomly drop off then come back up. Sometimes they drop for hours, sometimes minutes, sometimes seconds. Our intermapper updates in 30 second intervals and I often see customers that are down for 21 or 51 seconds. I’m assuming this is from them resyncing. Not all of our customers go down at the same time, it’s just random. Kind of frustrating.

Edit x2: The closest shot we have is:

Session Status REGISTERED
Registered AP 0a-00-3e-90-5f-16
RSSI 1767 (-55 dBm)
Jitter 2
Air Delay 59 (approximately 1.64 miles (8673 feet))

On an integrated unit.

-Turn down the power level (dbm) of your nearest customer. You want all your SM’s to read between 65dbm and 75 dbm.

-The reason it works better to assign a color code to each AP is because you want to be able to change frequency’s WITHOUT having to log into EACH SM. When you have over 100 SM’s and you need to change a freq because a local competitor just put equipment up, you are going to wish you would have done it the right way. Simply changing the color code on 1 device is easier than however many SM’s are locked onto that AP.

-Why 900 mhz customers drop from anywhere between 21 seconds and 4 hours is still a mystery. This happened ALL the time where I worked. One day they will all be stable, the next half the SM’s are reregistering.

-your download and upload speeds look normal.

-Perhaps the reason some people don’t experience this random droping of the connetion is because they are only monitoring their AP’s and BH’s. Alot of times these SM’s will drop, but their re-regs won’t change.

Sounds like some interference issues to me. An obvious question that you may or may not have asked yet, or looked into is the whole cordless phone issue. Check if your subscribers have 900MHz cordless phones. Sometimes they are fine, but other times they cause all kinds of weird stuff.


Yeah Mike that’s something that I recently realized more-so while reading these boards, especially with the vtech phones. So far I haven’t seen any of our customers with them though.

Wireless Solutions:

I’ll work on adjusting down the power level of the closest customer to see if that has any effect. Thanks. I read somewhere that Jerry said decrease the power level slowly because if you do too much at once it’ll drop completely.

I’m still not 100% clear on the color code thing. Why would you have to change the frequency on 100 SM’s when you can just change it on the AP’s? Or am I being difficult? The only thing I can think is that if I have a bunch of customers on 906 in a certain area, and then someone dropped something in there that killed that 906 signal, then I could just roll them all over to another frequency by changing the AP color code, is that correct? =)

Yeah customers dropping randomly is extremely frustrating, and it would seem like an interference issue to me, but we just can’t put our fingers on it and it’s multiple customers in multiple different locations. I just can’t figure it out for the life of me.

We monitor our customers completely, and like Jerry said in another thread, the surprise on customers faces when you call them before they call you to find out what’s going on.

Hey Chas. I am suspecting that you are running into self-interference between your AP’s.

In order to limit self interference, set each opposing AP on the same frequency.

N/S 906
NE/SW 915
NW/SE 924

You may have to adjust this for your area.

Make sure all of your AP’s are getting the same sync (if one is set to generate sync it will kill you).

Set each AP to a unique color code

Up the max range to 20 miles.

No Yagis on the AP, they are too narrow. We are using the 900 Integrated AP’s and they work fine.

Allow all frequencies on the SM’s. when you do an install, use the AP Eval page to determine what AP’s the SM can see and use the one with the best signal level/lowest jitter. Later, if you have to move frequencies, all the SM’s will follow the color code.

Adjust the power on the SM’s so that they are all within 10dB of each other, and 10dB above the noise floor.

If you don’t have a baseline for your noise floor to work from, you can figure it out by seeing what level the SM’s drop off. For example , if one SM registers fine at -77dB, but drops off at -88dB, then your noise floor is somewhere around -85 or so (Canopy needs consistent +3dB above the noise to stay registered). You lowest SM should be -75dB.

Get a mobile setup with an AP and a yagi, and go hunting for another Canopy network. If there is someone there, get them on sync one way or another.

900MHz is great for LOS, but it is not a magic bullet…

Thanks for clearing that up Jerry. I was performing a site survey and just happen to lock onto a competitor’s tower. The SM would register at 65 dbm, then reregister at 82 dbm. The reason it was doing this is because the tower had 2 AP’s that the SM was seeing. Both of these AP’s were on the same color code, therefore the SM was able to connect to both of them. Those 2 AP’s should have been on seperate color codes, then I would have simply performed a “AP EVAL DATA” to see which AP (color code) was the strongest signal/least jitter. I would then set that color code on the SM.

Thanks for the great advice, I was kind of hoping you’d pop your head in this thread =)

Couple questions though:

1) I read somewhere that the max range on your AP should be the distance of your longest shot +1 which is where I came up with 12ish. I’ll try 20 though to see if my throughput is still good + max range increase.

2) The self-interference thing is a possibility, I’ll check on the frequencies though and how we have them set up, thanks for the diagram. But if it was interfering wouldnt the AP kick off all customers on it? It’s just random customers and random different times. Would that be normal?

I understand the color code thing more clearly now, thanks. I’ll check on the noise floor etc. as well.

There is a competitor of ours about 30-40 miles away that has canopy set up, but I don’t think that would interfere with our setup?

Appreciate it =)

Chas wrote:

1) I read somewhere that the max range on your AP should be the distance of your longest shot +1 which is where I came up with 12ish. I'll try 20 though to see if my throughput is still good + max range increase.

The difference between 12 and 20 won;t hurt you, and I have noticed that my AP's perform better when the distance is higher.

2) The self-interference thing is a possibility, I'll check on the frequencies though and how we have them set up, thanks for the diagram. But if it was interfering wouldnt the AP kick off all customers on it? It's just random customers and random different times. Would that be normal?

Totally random.
The AP trains the SM. The AP wants to get the SM to register. If the communication is garbled, the SM will eventually give up and try again.

There is a competitor of ours about 30-40 miles away that has canopy set up, but I don't think that would interfere with our setup?

Not likley, but I would find out how they are generating sync and what they are doing for antennas, and what freq's they are on just for grins. If you are operating close to your noise floor it may help to know what your potential sources of interference are.

Okay so I tested most of the things you asked:

Our AP’s were already set up that way. So it’s not so much self-interference.

All AP’s are on the same sync.

Upping the max range worked somewhat, I still need to test it with our longer shots. (It’s the middle of the day at the moment, don’t want to test it when people are working on it)

We’ll change everything to a color code, we just weren’t worried about it at the moment because we only have like 20 customers across 6 AP’s

Going to work on the 10dB power variance later on as well.

Hopefully we can work these kinks out.

make sure you set the color code in the sm before changing it in tha AP…

done that one.

Not to pick on you, but that’s actually kind of funny considering you seem to be one of the most knowledgable people here =P

Question for you though: How come when I’m directly connected to the canopy unit (no matter which unit, what cable, or laptop I use) and I hit the refresh button once a second for 10-15 seconds, it eventually locks up? Even if it’s registered etc.

What software version are you running, and are you assigning public routable IP addresses to your Canopy modules?

I remember having the same refresh/reload issue back in 2002 when Canopy was first released, so that was probably back in the software version 3.0 days. But, apaprently when you assign public IP addresses to your modules the web interface of the unit quits responding, but the unit continues to pass data. They claim it is because the integrated web-server starts taking a lot of hits when it is a public node on the Internet.

I have never tested this myself, but is quite interesting. If I were to try this, I would probably just block TCP 80 coming in from the WAN all together and see if the lockup issues persisted. Or, in a future software release Motorola should just allow you specify a custom port to run the web-servers on. It would be interesting to see if when you change to a custom port, and assign the unit a public address, if the problem would persist or disappear.

Anyways, what software version and are you using public IPs?

That’s how I learned it - break it and then have to fix it… :wink:

Rather than hitting refresh over and over, use the web page auto refresh function on the radios - we set it to 3 seconds. This eliminates the freeze problem.

I don’t think it’s just canopy. If you continually reload web pages on anything with limited memory it will freeze.

Makes sense I suppose.

Our version for the SM is:

Software Version CANOPY 7.2.9 Jul 23 2005 01:49:03 SM-DES
Software Boot Version CANOPYBOOT 3.0
FPGA Version 070605 (DES Sched)

AP’s are:

Software Version CANOPY 7.2.9 Jul 23 2005 01:49:03 AP-DES
Software Boot Version CANOPYBOOT 3.0
FPGA Version 070605 (DES Sched)