PMP450AP with PMP320SM?

I know that I am getting to the party late. I was told by a friend that there might be a firmware update to allow the 320sm to work with the 450ap in 3.65. Is that the case?

Nope... completely different technologies and platforms... however, you can synchronize a PMP450 AP with a PMP320 AP for collocation or migration purposes.

1 Like

@Eric Ozrelic wrote:

Nope... completely different technologies and platforms... however, you can synchronize a PMP450 AP with a PMP320 AP for collocation or migration purposes.


Have you actually done this successfully?  We tested a PMP450AP with one SM in the middle of our network (70 PMP320 APs, 1000 PMP320 SMs) and were unable to avoid a fluctuating 5dB-25dB interference from the PMP320s crippling the PMP450 sector.  After multiple tickets with Cambium, dozens of emails/phone calls, the official word has always come back to "The problem will go away once you've eliminated all the PMP320 gear, if not we'll investigate then, can we close the ticket?".  VERY disappointing.

j

1 Like

That kinda sucks.

Anyone have use for three cmm4's, 11 ap's and over 350 sm's?

No, I have not used the sync features between PMP320 and PMP450.... but that being said, there must be some people that are using it out there that could chime in.

It wouldn't hurt to try some sync and see what happens Tallen. Also, you could take advantage of the Cambium WiMAX trade out program.

I have had operators sucessfully get these working together, but it's more then just dropping in a radio.  Imagine a tower with 4 sectors and A/B frequency re-use.  To get this all working together, first, we have to be on the same frame size (5ms on the PMP450 to match WIMAX).  The rest of the attributes that normally have to match (UL/DL ratio, max distance) all have to match as well.  If you are NOT using PMP320 and specifically PMP320 with the sector antennas that Motorola / Cambium sold, you may need to swap both the front and back sector at the same time.  This is because many 3rd party antennas do not have the same front to back isolation as the ones that Cambium / Motorola supplied.  If the whole thing has been Moto/Cambium from end to end, this is all much easier to just replace one sector at a time.

So, to recap, it's all about matching frame duration, UL/DL, and max distance, and making sure the antennas have sufficient front to back isolation. 

I should further clarify that Wimax devices like the PMP320 will never talk directly to a PMP450 Access Point with any firmware load or settings.  As another user pointed out, the protocols are quite different, and the best we could to was to be able to operate them in the same space without self interference.  This gives an operator the opportunity to do a phased replacement of the WiMax gear, or operate the two technologies side by side.

Hi Alan,

I would like to know how they managed it.  Our experience has been far from satisfactory.  The 450 platform is just not capable of performing even close to the abilities of the 320 platform in NLoS situations and the 450 also even with LoS  has had less performance then what would be expected.  We currently have 3 different deployments of the 450 platform.  2 of these deployments are not on the same tower as our existing 320 platform but definitely is in range to see the interference.  We have everything running on 5ms frames and you confirmed that the settings for sync where satisfactory.  We have changed our channel plans so that there is "clear" space for the 450 with 2.5mhz+ of guard band and still the 450's performance is impacted.  At this point, we just do not believe the 450 platform is a viable solution. 


@Alan DesJardins wrote:

I have had operators sucessfully get these working together, but it's more then just dropping in a radio.  

....

So, to recap, it's all about matching frame duration, UL/DL, and max distance, and making sure the antennas have sufficient front to back isolation. 


I'm interested in this, because with 6 weeks of access to our PMP450 AP and SM, and the PMP320 AP on a nearby tower that appears to be the primary interferer, Cambium engineers were unable to improve things or offer any advice other than "once the PMP320 is all gone you should be good".

We're using an ABCD frequency arrangement, the primary source of our trouble seems to be a tower 8 or so miles distant, with a PMP320 sector on the same channel aiming generally toward the PMP450 AP.  We used the migration spreadsheet to calculate frame settings and what-not.

j

Hey Newkirk... are you using the Cambium antennas with all your PMP320 gear? Are you running legal EIRP's on all your 320 gear?

Interesting.  My experience is the polar opposite.

We're consistently seeing much faster speeds through the 450, in excess of 20 Mbps, when replacing 320s that were struggling to pass 6 Mbps.  We've been able to take down 40' guided masts and replace them with 16' poles in tripods when switching from the 320 to the 450.  This is in a very noisy and heavily treed area.  The 320 has never been a viable product for our environment, and the 450 is a godsend in comparison.

Are you using the SM on it's own, or with dishes?

Have you used the 320/450 colocation tool at https://support.cambiumnetworks.com/files/pmp450/?  It's probably worth going over the migration whitepaper there as well.

For us, we're using mostly Cambium sector antennas on the PMP320APs, with a few KPPerformance.  Transmit power and antenna gain are normal and legit for US/FCC.

I've read the whitepaper numerous times, used the calculator spreadsheet inputting our PMP320 settings (extended range enabled, 75/25, 10mhz wide channel) as supposedly have Cambium engineers when they were trying to figure out our interference.  We never stop seeing interference from the PMP320's at the PMP450AP, the SNR on the 450 fluctuates by 12-20dB sometimes, like sometimes it sees the PMP320 and sometimes it does NOT, suggesting that the frame timing might not match perfectly.  (which could happen due to a variety of causes, I presume, like sync inconsistancy - we've not seen an explanation though)

I've no complaint about the capacity of the PMP450 gear.  We started with SM on it's own, are currently using a reflector, but only have the one AP and SM pair in the field.  My complaint is that even with the frame timing calculations it seems to NOT play nice, and Cambium support has only seemed able to dither for days/weeks, suggest it will be 'fixed' if we migrate everything, and ask if they can close the ticket.  Bosses essentially consider it a failed experiment and are seeking an alternative path forward from PMP320 in 3.65GHz band.

j