Strange problem with 900Mhz towers

Here’s the problem we’re having, we have two existing 900Mhz towers.
One to the north (Tower A), one to the south (Tower B), they are about 13 miles away.

In between the two towers there has been a “dead area”. We put 3 integrated APs (Tower C) to cover the southern area of Tower A and the northern part of Tower C.

The problem we are seeing is that customers that were connected to Tower B, and that should have a great connection to Tower C, won’t connect to Tower C. However when we turn off Tower C, they go back to having a marginal signal to Tower B.

The only thing I can think of is that customers that are south of Tower C that are connected to Tower A, are actually drowning out the signal from Tower C.

Does this make any sense to anyone? Do I need to put a full 6 AP cluster on Tower C to fix this problem?

Do you have color codes set? Are all frequecies enabled on the SMs? Are you causing self interference between your towers?

Also, what is your frequency layout on all 3 towers? Do you have the same Distance/Control Slots/Downlink Per. on all 3 towers?

These are all pretty simple questions that could really help the community give you an informed answer.

All the towers use the same color code.

They are all set using 906, 915, & 924 with their directions recommended in the manual.

Dowlink, control slots, & downlink % are all the same also.

They are all also on GPS sync.

Well Color Codes should be at least different at each tower location to prevent self-interference. Tower A (CC=101) Tower B (CC=102) Tower C (CC=103). Granted they don’t have to be those exact numbers but you definitely need that set.

Once set, you log into the customers radio and you can do an AP Eval that will tell you which Color Code is getting the best signal for that customer.

With the way that you have it now, the SM is basically so confused it can’t connect.

We’ve never had a problem with different color codes though, we have over 13 towers in that area all with the same color code. 1600+ customers in that area and have never seen this before.

different color codes is just good sense.

Even if it is good sense, it still doesn’t explain why it sees such a horrible signal from the closer tower. Changing color codes doesn’t affect the frequency at all.

Color codes won’t cause self-interference, they will simply help the SM distinguish between APs.

If your SMs can see two APs broadcasting on the same frequency, that may be the problem. When we first branched off our main broadcast site (using 2 180degree MTI sectors) I had both APs set to the same frequency using different color codes. Any of my clients that happened to be hitting “in-between” the two antennas could not connect. Changing the frequency on one of the APs solved the problem.

How many APs are on Towers A and B? It might have something to do with the way the frequencies are laid out in your setup, especially if the other tower sites have a different number of APs.

Every tower I have, all have different Color Codes (3-120 sectors). Any tower cluster (6-60s) I have a different Color Code on each AP making it impossible for the customer to change between APs.

I understand the reasoning for having different color codes, however that would not cause the problem that I am seeing.
Having separate color codes is not a requirement for tower setup.

I guess the question I am looking for, is having a 3 AP cluster, only covering 180°, going to cause a frequency problem when it is sandwiched between two towers, assuming all the settings are the same between the 3 towers.

If the other towers are 6-60’s and you have a tower in the middle with 3-120’s then the answer is yes.

pcpolo wrote:
I understand the reasoning for having different color codes, however that would not cause the problem that I am seeing.
Having separate color codes is not a requirement for tower setup.

I guess the question I am looking for, is having a 3 AP cluster, only covering 180°, going to cause a frequency problem when it is sandwiched between two towers, assuming all the settings are the same between the 3 towers.


Yes, if you have a different number of APs at the other two sites.

If you have 6 APs on Site A, 4 on Site B and 3 on Site C, you're not going to be able to follow the Canopy recommendation for frequency separation.

There are 6 APs on Site A, 6 APs on site B (which is 13 miles south of Site A), and sandwiched between the two is Site C with 3 APs point NW, N, and NE (basically the top 180°).

So am I to assume that SMs that are south of Site C, which are connected to Site A, are causing the problem with SMs north of Site C that are tryign to connect and that if I put 3 more APs to complete the cluster of Site C, that that SMs north of Site C should connect better?

pcpolo wrote:
There are 6 APs on Site A, 6 APs on site B (which is 13 miles south of Site A), and sandwiched between the two is Site C with 3 APs point NW, N, and NE (basically the top 180°).

So am I to assume that SMs that are south of Site C, which are connected to Site A, are causing the problem with SMs north of Site C that are tryign to connect and that if I put 3 more APs to complete the cluster of Site C, that that SMs north of Site C should connect better?


More likely it's the APs causing problems with each other. Your North and South clusters are affecting the cluster in the middle, and the middle cluster affecting the North and South, hence why you are seeing SMs drop their links when you power it on, and come back up when you turn the tower C POP back off.

So the fix? complete Tower C’s cluster?

pcpolo wrote:
So the fix? complete Tower C's cluster?


That would be my suggestion.

ok, thanks, that’s the only thing we could think to try next. Now to find 3 more APs…

pcpolo wrote:
ok, thanks, that's the only thing we could think to try next. Now to find 3 more APs......


If you look at the AP eval pages on the SMs when pointed towards site C and are seeing adequate signal levels, than there's no reason to believe that adding the extra APs wouldn't solve the problem.

From what you've been describing it really just sounds like self-interference due to the mismatched cluster sizes.

Another thing - if you haven’t already, you should probably at the very least assign separate color codes to each cluster. If you have an SM that can see multiple clusters, it will try to connect to both of them.

I’ve had problems with that in the past where my SMs were connecting to the “wrong” APs when I’ve power cycled something. It doesn’t matter if the signal level on one AP is -60 and on the other is -82. If it connects to the one at -82, it will stay registered to that AP until it reboots or the link drops.

Yeah, that’s something we’ve been considering.