Peplink.com Marketplace Partner

Any LTE gurus out there?

Product Discussion Pepwave MAX
#1

Hello! I have a MAX Transit with CAT18 and the Panorama WMM4G-6-60 4x4 MiMo directional antenna. There’s an ATT antenna just a pinch over a mile away and I have it pointed at it. I’m seeing wildly varying network performance. The Pepwave seems to flap between LTE and LTE-A constantly.

One speedtest I’ll see 75Mbps and the next I’ll see 5.

When it’s reporting LTE only it’s showing band 66. When it’s reporting LTE-A it’s showing band 66, band 12 in SSC1, band 66 again in SSC2.

I’ll throw some screenshots in here as well.

Are there any configuration tweaks that can be made to help stabilize and make it more consistent?

thanks a lot!

image

image

image

#2

in my experience, this is how ATT works, they have several bands, and depending on what Aggregates, depends on what it performs like. i have removed band 4 and 30 and things are better, try disabling different bands. Also, i believe the modem will display “LTE-A” is when the modem is aggregating multiple bands/channels.

2 Likes
#3

So, oddly, it’s working much better now.

I went in and started messing with disabling all but certain bands. I completely loss connectivity. I reverted back to auto and now it seems to have settled into the following:

image

I’m not sure, but as long as it stays like this I’ll be happy. Here’s to hoping when the leaves come in it doesn’t affect this too much. :upside_down_face:

#4

The short of it is that your equipment offers up a menu of bands and the carrier (tower) decides which ones to use (and how).

Cheers,

Z

1 Like
#5

@zegor_mjol is right. And, I’ll note we have seen similar behavior with AT&T in several locations. A typical scenario: LTEA followed by LTE a few minutes later – where it “stays.” We also note that AT&T cannot be relied upon to assign the mix of bands that will optimize the user experience. And, this is true with Cat 6, 12 and 18 modems.
What you are seeing is not the “fault” of your router. And, the last metrics you posted look quite good – hope they remain that way for you. :<)

1 Like
#6

Eliminate 12 and 4 from the mix. You should aggregate two Band 66s. You could very well be picking up 66 from a tower beyond the one that you think is servicing you. I do, and aggregating band 12 with 66 makes my Duo crawl. Two band 66s is optimal if your signal is strong enough - band 66 has large blocks on ATT and is often under utilized since most people don’t have the option of specifying bands and most routers default to band 12 with ATT.

And I don’t agree that ATT doesn’t optimize with band mix selection. It optimizes beautifully for me and I have both Panorama and Poynting top-of-the-line directional antennas in MIMO setup. Both brands perform the same in my experience.

1 Like
#7

I’m giving that a shot @joelbean

It required a reboot after changing the bands, but now the base network is 66 and it will grab another 66 sometimes. I believe I found a software bug along the way. If you have IP passthrough enabled, the required LAN static route goes away upon reboot and requires re-adding.

Is there such a thing that it will only aggregate during demanding situations (a speed test, for example)?

If I sit and overanalyze it I see the second band drops off when idle but reconnects during a speedtest. I get a little more packet loss than I’d like, but hey, thank goodness for TCP retries. :sunglasses:

image
image1629×827 210 KB

image
image1004×153 20.2 KB

#8

What the router selects for connections is independent of what you use the bandwidth for. It is not on-demand so to speak. It either connects, aggregates, or not.
It appears that your signal strength (RSRP) is only fair on one of the 66 bands. Your SINR is very good though.
There is a way to know exactly what tower you’re connecting to and then see the antennas/bands that the tower serves.
Open ‘Details’ for the connection, then choose ‘Detailed engineering data’ in the link with the “?” at the top right corner. If you tell me the “UTRAN Cell ID” I can tell you about the tower/s your connecting to. You can do the same on cellmapper.net by inputting the UTRAN number into eNB ID Calculator under Tools, then get the tower ID it returns and put it into the “Tower Search” box on the left-side panel. It’ll show the tower in the map on the right with all of it’s detailed info about bands supported, etc.
I’ll do it for you if you just send me the UTRAN Cell ID.

#9

Thanks for the tips! I confirmed that I am pointed at the tower providing band 66.

image

image

#10

Yes. That tower has great band 66 coverage, strong and wide. Band 12 is a turtle and narrow.

Are you using MIMO 2x2 or 4x4? In other words, cat 12 or cat 18 router?

Depending on how far you are and what direction you are from this tower, you should be able to sustain two aggregated band 66s fine. I believe this will be your best performer. If you are very close to the tower (I think I read that you are), try aiming the antenna slightly down from the tower. This will help the diversity since the earth bounces signals quite well. You may be able to stabilize your second band 66 this way. It’s worth a try.

#11

I don’t understand why your signal strength isn’t better at a mile away. Perhaps you are in a dead zone for band 66. In this case, aim your antenna/s offset from the tower and see if you can improve the signal/s. Although, the Panoramas and Poyntings (directionals) have a quite wide beam and ample gain to pick up the tower signals, so I don’t understand the marginal RSRP at 1 mile. You should be getting 70 - 85 RSRP with this tower on band 66.

#12

I have a CAT18. Distance is just over a mile, however, forest. :evergreen_tree: Lots of trees in-between.

I tried pointing down a bit which didn’t help, but will also try offsetting. I’ll report back after more tinkering.

thanks again!

#13

I think I have it the best it can be. On a single 66 band I can pull 50Mbps down and 6 up with pretty minimal packet loss and respectable latency/jitter numbers.

My only disappointment is that I’m pretty sure I overspent going with a CAT18 setup because it won’t utilize it. Oh well. Live & learn I guess…

#15

5gstore, who I bought this gear from, had a pretty good troubleshooting tip. Throw the included antennas onto the Pepwave and re-check statistics. I had actually forgotten that it included little antennas.

So inside my house with the little antennas is very close signal-wise to the Panorama on the roof pointed at the tower. I actually had a little less packet loss as well and an improved SNR. :thinking:

Small antennas indoors:
image

Panorama on rooftop:
image

#16

I know. It doesn’t make sense, does it?

#17

The paddle antennas don’t have any signal loss from cable. If you could expose your Panoramas outside with a couple of feet (only) of good cable, you would probably report better signal metrics than even the paddles. It’s always a trade off when cable is involved. Most vendors put 195 with SMAs and the signal loss can be significant.