PMP 450i in heavy snow queston

We had our first heavy snow since deploying two PMP450i 900 sectors and the performance was somewhat surprising. With the 900 Mhz penetration capabilities I has assumed like the old FSK, the RSL may drop a bit but it would be fine. That was not the case.

This was a heavey wet snow with 15 mph winds. Both APs are using KP Performance Omnis. As the storm came through the RSL did drop a bit as well as the modulation levels and then we basically lost the horizontal plane across the board for over an hour with everything dropping to low MIMO A modulation.

On the plus side, we only had one SM drop drop registration, a marginal customer 5 miles out They are one of our employees so they knew it was marginal when we turned them up, but it was that or nothing.

A typical RSL drop was from a -68 to a -84 with modulation shift from 8X/6X MIMO B to 8X/2X MIMO A. Horizontal signal was NA across the board. Obviously frame utilization maxed out during the event, but customers were still on the air.

I’m curious if this is something expected with the OFDM or just something weird with this particular snow event. Possibly the OMNI’s are a contributing factor?

FYI all SM’s are using KP Performance 17 dbi yagis with all signal levels raging from -60 to -68 with 8X6X MIMO B during normal operation (except for the employee) Software release is 16.2.3 on everything.

Previous heavy rain events did not exhibit this behavior on the 450i.
The 2.4 and 5 Ghz ePMP performed basically identical to a heavy rain event, with RSL dropping a bit as well as modulation levels, but nothing this severe. We also noticed that once the storm finally did move through it took awhile for some of the SM’s to regain their original modulation level, in one case, over an hour.

This was some really wet, dense snow, perhaps laced with dust from the debris of an alien space craft or something.

In our experience, this is a normal function of Yagi’s.

With a dish, panel, sector, grid — if they get covered with frost or ice or wet snow, they’ll end up with a 1" or 2" coating on them - and you have 1 or 2" of snow in the ‘path’ of the signal.

With a Yagi that is 6" long, each element will get a 1" or 2" coating of frost or ice or snow on each element, and you’ll essentially have a 6’ long tube of snow in the path.

Typically, 900 MHz radios are less susceptible to RF fading due to weather conditions (rain or snow). However, as you mention, the antennas used in the system can have great effects on the signal integrity. It may be that the omni patterns are more severely affected, or that (as ninedd mentions), the yagi-style antennas are holding onto much more of the signal-affecting precipitation for a longer period of time. It is tough to say.

1 Like

Curious as to why you are using v/h instead of slant 45?
Slant 45 has better stability during weather events and is less likely to have build-up on the antenna elements, though if the conditions are right, they pack up just like a v/h antenna.

Ofdm does have better rf propagation characteristics but heavy driving snow just kills rf signals. We used to have good signals at all times but we would loose 2x modulation on the old canopy radios and depending on how much the path was blocked, we could see a -70 signal drop to -92 (would hold there but passing data was like passing kidney stones)

I am also wondering why a 5mi path would be marginal? Too many trees? With canopy we had clients 10 and 15 miles out still connecting at -72 ish at 2x modulation.

On the other side, KP makes some really tough and reliable antennas, We still have the old yagis and sector antennas as we are planning to add a 450i to one of our more problematic areas (too much noise in 2.4/5ghz, lots of vegetation too) and dual antennas more than 28 inches apart vertically will not have too much for adverse effects.

They are the slant 45’s, as is the KP Omni. It just identifies them in the SM a V and H, so I just tagged that Vertical and Horizontal. I had driven around during the heaviest snow and really didn’t see much accumulation on the yagi’s, but it may not take much. I’m leaning toward something with the OMNI. We have another snow event coming in tomorrow, though it will be less severe.

The 5 mile path is going through a LOT of trees and just clearing a ridge at ground elevation, also with trees, and I mean just clearing. We did a lot of testing trying to get a good signal and found one sweet spot for the -78 Since its an employee and they are not a power user, we can live the the 4X MIMO B. Honestly, we were impressed we got that. I told him to break out his chain saw come spring.

We struggled with the deciding on the OMNI deployment, but the customer base simply did not justify multiple sector 900. I suspect the sector antenna would perform quite a bit better.

On your comment about using two antennas, we have quite a few of the old KP 17 single polarity Yagi’s stashed and I actually played around with 2 connected to an SM just to see if it would work. Of course they were true horizontal/vertical but was told that the AP corrects for that. We played with the spacing and found 11 inches separation on the Yagi’s was the sweet spot. We then swapped the same SM to the dual polarity 17 KP yagi, and the newer integrated actually was 2 db better, LOL. We never planned on using the configuration but I tripped over them and thought, why not?

We still have the some old omni MTI giant horizontal panels and the monster whip stashed somewhere. We had briefly considered trying what you are suggesting, but I had forgotten how heavy the old MTI 7 foot tall monster was and the whip is no light weight. One of our installers thought the KP omni was rather heavy until we had him heft the MTI panel…

We’re actually still using an old PMP100 900 MHz integrated access point with filter. It has a 12 db integrated 60 degree horizontal panel. We used it to provide service to a single customer on a remote tower that was unreachable with 2.4 due to the tree coverage. The customer is a friend of mine and is raising her grandchildren and they needed internet access for school. We had a couple of old 900 integrated AP’s and a box of P10 SM’s, so we upgraded them to the last FSK release, 13.4.1, and rigged her up. The throughput of the old FSK gives them about 4.4 down and 400K up in 2X Advantage, which works for their needs. They will always be the only subscriber. The satellite installation and MRC was something she couldn’t afford. We even have spares, LOL It’s far enough away and pointing the opposite direction, so no interference issues with the 450i. Neither can see each other on the SA.

Faded and scratched it still got the job done, and it’s better than sitting in storage gathering dust. It was actually kind of fun playing with the old FSK, though our installers thought I was nuts. I told them it would work just fine but they didn’t really believe me. I just tagged on the AP installation with another tower climb. The entire integrated AP only weighed about 14 pounds.

The old dinosaur just works and won’t die short a lightening bolt from Olympus.


If you ever need spares for that old 900fsk, give me a shout. I have lots including a horizontal omni, a 120deg DP 12db pannel and more than enough APs, SMs, 13/17db yagi’s and integrated antennas to start another gurilla wisp!

If you decide to try the dual yagis again, try to space the antennas more than 2 wave lengths apart and ironically despite the sslant orientation, you want virticle separation. At 918Mhz this is about 13 inches. This way you dont have combined losses of the antennas magnetic fields. If you were trying to combine two to a single feed and the antennas would have an identical magnetic field then being half to one wave length would be ideal, but I found the injection losses on two 17’s with a balun splitter to give only 2db additional gain. Not enough gain for the costs but was worth playing with.

If your lone customer needs more bandwidth, I would sync two APs on an LACP port channel and two SMs on her end with a switch to complete the lacp bridge. With 3 available channels you can go up to 3 links and she could have almost 10/2 service.

We have a few clients installed in trees. Well 3/4 ready-rod through the tree and a pole clamped to the tree to get above the trees. Not ideal but on a 5.8ghz f180 at 7miles we are able to provide 98% uptime to that client with down times due to the 70mph wind storms we get.

Yeah, KP does that, it should be left and right slant or simply A and B.

1 Like

Very interesting. The employee is also our primary tower climber so want to keep them happy. An extra 2db could make the difference of them staying on the air or not during weather events.

We have used the tree option as well in a few cases. quite a few old break over poles left over from TV antennas we have taken advantage of as well…

Our snow event yesterday was about 1/2 as intense and everything performed well, with minimal drops in modulation. We upgraded one of the AP’s to 20.0.1 again today and hope the SM’s dropping registration issue has been corrected as we really like the additional modulation levels. It especially helps on the uplink on the OMNIs


1 Like