PMP450i Colocated With PMP100 Timing Issue 900Mhz

Products PMP
1

Network:

PMP100 900Mhz network.  Many sites with four and six sector arrays.

Firmware: 11.2

Sync Source: CMM3 (all sites)

Scenario:

Attempted to co-locate PMP 450i 900Mhz

PMP450i Firware: 14.2

Timing coordinated using the Cambium colocation tool

Frame Alignment Mode: On (Mode 2)

Sync Source: uGPS

Issue:

PMP100 APs within signal range of the PMP450i receive interference.

Test Scenario:

PMP100 Sector facing a PMP450i sector 9 miles apart.

Running Link Capacity Test for 10 seconds on select clients.  Ten tests each

Benchmark:  Run Test with opposing sector off and idle

PMP100 Results: 99% uplink efficiency

PMP450i Results: 90% uplink efficiency

Test 1: Run Tests Simultaneously

PMP100 Results: 18% uplink efficiency

PMP450i Results: 90% uplink efficiency

Test Scenario #2

PMP100 Sector facing a PMP450i sector 9 miles apart.

Running Link Capacity Test for 10 seconds.

Replace uGPS with power port timing.  (Packet Flux Sync Injector timed via a CMM3)

Benchmark:  Run Test with opposing sector off

PMP100 Results: 99% uplink efficiency

PMP450i Results: 90% uplink efficiency

Test 1: Run Tests Simultaneously

PMP100 Results: 99% uplink efficiency

PMP450i Results: 90% uplink efficiency

Other Observations.  Reg and Re-Reg counters increment on the PMP450i when there is regular activity on the PMP100, indicating that the PMP100 is interfering with the PMP450i when using power port timing.

Speculative Observation:

There is a difference between uGPS/Aux port timing and Power port timing.

Cambium 450i with uGPS is not compatible with PMP100 with power port timing via a CMM3.

Cambium 450i with power port timing via a CMM3 source is close to aligned, but not perfect.

Solution:  None Known.

Notes:  The PMP450i needed the distance changed to get an "OK" in the colocation tool, so I cannot dismiss the possibility that a configuration tweak could help and that the co-location tool is flawed.

2

One other note:  The co-location tool represents a 5ms frame, but the devices use a 2.5ms frame.  This calls into question the accuracy of the co-location tool.

The results still show an obvious difference betwwen uGPS/Aux port timing and Power Port timing.

3

Hi, 

A few things. 

1. The PMP 100 900 MHz radio uses a 5ms frame, not 2.5ms like the other bands. Therefore you will have to use 5ms on the 450i 900 MHz as well. 

2. I see that you are using 11.2 on the PMP 100. Would you be able to upgrade the site to 13.4.1 to help with the colocation?

3. Tweaking the distance (max range) in the colocation tool is one of the ways to get PMP 100 and PMP 450i frames aligned and is not a flaw in the tool. 

Thanks,

Sriram

4

Sriram,

Shoot! My Frame size is an obvious mistake.  I just assumed 2.5ms the whole time.  Thank you so much for making this obvious to me.  So simple!  

 

I'm still confused at the obvious difference in my timing tests, which can be reproduced faithfully.  My 2.5ms frame should have been causing all sorts of odd behavior that I couldn't see.  Oh well, time for a fresh start!

I've had a case open with cambium for over a week now, screenshots, captures... I'm surprised they didn't notice that.

 

Upgrading past 11.2 is dangerous for us.  We have three neighbors running the same hardware, firmware, and timing.