I have seen behavior like this on ePMP 1000 on two sites. It works, sees 8-9 satellites, than suddenly 0 and after random time (minutes,hours, days) it jumps back to 8-9. On one site the outages were shorter and after we set the Synchronization Holdoff Time to maximum value, it was enough to keep the AP running. On the other site it wasnt enough and after it switched itself off few times, we changed it to Flexible mode.
I am not aware of any LTE or other telephone antennas around the sites, but I believe that the problem is somehow location dependent.
I have the same problem with an ePMP 2000 Lite The next day of installing it I stopped transmitting the WiFi and lost the GPS, my distributor sent me another spare unit, but I am still waiting for the licenses of the Elevate and for the full version, more than a week ago and still are not the licenses, is Desperate, I have customers without service.
Lite is the same hardware as Full, just a license change is all.
We have NO cell carriers on the tower that is having issues. The closest cell tower is 3.13 miles away and it is Verizon and has 700mhz LTE. But this should be way too far to affect GPS for these radios.
The only thing on the site we're on besides us is some 800mhz Motorola two-way gear and some 6 GHz backhauls. Nothing else except 2.4 and 5ghz from us on this site.
From what I understand, if you have 800MHz gear there, the 2nd harmonic would be 1.6GHz, which is what the Cambium GPS operates on. I've been told that this could interfere greatly with the GPS since the pucks are passive and don't have much gain at all.
Well, my AP with the failing GPS is mounted on a customer house as a mini-pop. It is far away from any cell towers. The only other radiation near it is a Mimosa B5 Lite 5GHz backhaul radio and it is shielded from the AP. The AP is brand new and has the newer GLONAS GPS puck.
I've been tracking it since it failed. It has never gone below 6 tracked satellites since it failed 2 days ago - but I am still running in flexible mode until I can get a handle on the problem.
Not yet. It's a long climb up to this with a couple ladders... I want to give my monitoring software a week or so to track it and see if it fails again.
If the AP has a very clear view of the sky, is there a reason not to use the internal GPS chip (and not the puck)? Finding a place to get that puck to stick on a pole mount is a bit challenging sometimes.
The GPS receiver in the ePMP2000 is the same as the later revision of the 450 AP. More accurately, it's a GNSS receiver (GPS+GLONASS). I don't remember what's in the ePMP1000. Might be GPS only like the first spin of the 450AP. This is what Forrest at PacketFlux is currently using in his products. They're all from GlobalTop and are pretty reliable, but sometimes they just get confused a need a hard power-cycle to get un-stupid. Noise can definitely be an issue. Multi-path is another problem.
I agree with Chris, use sync-over-power if you can. I'm sure Cambium would like you to buy and use a CMM, but a PacketFlux SyncInjector or PowerInjector+Sync module and a SyncPipe Basic is <$300. This is the best approach because you can have the receiver down at the base of the site away from radiating devices. We have very few issues with this configuration.
It happened again. Lost all satellites for exactly 7 minutes. I can now track it with SNMP. Went from 9 satellites to 0 for seven minutes then right back to 9.
It has happened twice now and there is one thing in common... Both times it happened, it had just started to lightly rain. These are the only two times it has rained at all since the AP went up. It's been plenty cloudy so I don't think it's a reception issue. I think it is water related. However, it is still lightly raining and the GPS satellites came back to 9 so if water is causing this, it is not staying bad until it dries off.
The AP is fully exposed. It is connected to an omni with short pigtail cables. The omni is mounted to the top of a J-Bar arm and the AP is attached to the angled part of the J-Bar. The access door for the cables is facing down (on the 45 degree j-bar arm). Water can't follow the cables into the AP due to gravity. Water could potentially get in the small hole where the cables enter but could not get more than a few drops before leaking back out due to gravity so I don't think connectors in the radio are getting wet.
Here is a shot the installer took during the install. The GPS puck is on top of the J-Bar. We install a piece of metal strap across the hole at the top of the J-Bar to give the magnet something to stick to (that is the silver band you see).
We have a couple others installed the same way that have never had an issue - but the rain thing has me wondering.
your GPS puck orientation needs changed, face it upward facing so your coax is coming out of the bottom of the antenna, rather than laying flat. you will see 10 to 15 increase in GPS SNR from that change. and likely more sats tracked.
most of my APs see 4 or 5 sats above 40. If for some reason your issues are related to weather, that extra SNR would make a difference for you. or if water is some how bothering your puck, changing its orientation might resolve that as well.
I'm confused now... When I put the GPS puck in the holder at the top of the Cambium sector, it faces up - or what I call up. The part with the magnet faces earth. You're saying stick it to the pole so the cable goes down and the magnet is at 90 degrees to earth? Does it matter if the large flat part faces south, north, etc?
I will admit, the SNR for GPS on that AP sucks. However, I have 5 other APs in the area, all with pucks in the same orientation as that one and all with great SNR. Maybe a bad puck? Or the antenna connector on the AP is problematic...
If you’ve got other pucks sitting like that, with much better SNR, something is probably wrong with that puck. I’ve noticed that laying them flat, with the large surface facing skyward is not as good snr as the puck being flipped 90 degrees so the flat surface is facing parallel to earth, with the coax exiting the antenna headed downwards to the ground. That may change depending on the part of the world your in, I’m not sure, but try it with one of your APs. The GPS snr seems to update kind slow so wait 20 to 30 seconds to see the change. US Midwest is where I am, and every time I’ve flipped a puck flat, like you have, I’ve lost an average of 10points snr.
We are in Canada, the puck orientation has made a fairly large change in the number of tracked and the quality of the recieved signal on the GPS. On pole mounted AP's we use an exhaust U-bolt clamp and a small steel plate to hold the puck antenna so the cable is verticle and out the bottom (towards earth). We also found that the use of electrical tape is prefered over zip-tyes or wrapping the cable around the pole. Excess cable should not be looped as this causes issues with the gps signal. We found running the cable out and making a second pass on the pole with the cable works well as long as the cable doesnt loop.
Just my $0.02 but that has basically solved the gps issue for us aside from the fact that most of our sites use the Packetflux Syncbox and the gigabit injector, which is the way we are planning to continue to sync our sites and keep the onboard gps as a backup.
It would be nice if the ePMP guys could do AutoSync like Canopy. Prefer CMM3/4 and fail over to on-board GPS. I suppose FreeRun aka internal/generate wouldn't be a bad option to have either.
It would be nice if the ePMP guys could do AutoSync like Canopy. Prefer CMM3/4 and fail over to on-board GPS. I suppose FreeRun aka internal/generate wouldn't be a bad option to have either.
Yes, that would be very nice. They do kind of have free run already, with the Synchronization holdoff time setting... it just can't run for more than 24 hours that way.
I think there is more to this GPS issue... I just installed a second brand new 1000 AP with the new GLONAS puck. It has the same problem as my previously new install.
This 1000 AP with GLONASS puck is sitting right next to two 2000 APs with the same puck. The pucks are all within 1 foot of each other with the same orientation. Both 2000 APs are tracking 14 satellites. The 1000 AP with the same new puck is tracking 6 satelites.
My other 1000 AP with the new puck is also tracking 6 satellites, all with SNR below 30. It drops to 0 satellites tracked through out the day. No change to the 2000 APs.
My older 1000 APs with the older pucks have 12 - 15 tracked satellites and most have SNRs in the 40's and 50's.
I think the new GLONAS pucks do not work well with the 1000 APs. I now have two 1000 APs with new pucks acting exactly the same (poor SNR, low tracked sats and dropouts to 0 satellites on a clear day). I have 4 1000 APs with old pucks that work perfectly and two 2000 APs with new pucks that work perfectly.
