Community Forums

turbopilot · 02-11-2019, 07:35 PM

Quote:

Originally Posted by NealC

The past two days I've been doing extensive and exhaustive (exhausting) testing.

Thanks for the testing. So many variables.

I am not an antenna expert but I have closely followed Dr. Andre Fourie's YouTube channel for years. He is associated with Poynting an antenna maker in South Africa. Dr. Fourie has an excellent three part lecture on cellular RF performance. I will link to those videos below.

Here are three screen shots taken from that lecture dealing with the issues we have been discussing here: antenna type, antenna cable attenuation and MIMO antenna separation. They speak for themselves but he covers each slide in detail in his lecture.

Right or wrong much of my input on this thread has been based on Dr. Fourie's excellent and detailed information. If you go to his YouTube Channel and look at all of his presentations over the years it sure appears that this guy is an expert.

Anyone planning on purchasing an advanced cellular modem should watch this lecture. Poynting antennas are hard to find in this country. Some models are available on Amazon.

Here is the lecture in three parts:

https://www.youtube.com/watch?v=gup3E0ahk5g

https://www.youtube.com/watch?v=hlp7dR5mTho

https://www.youtube.com/watch?v=y-Bm-Na1MpQ

NealC · 02-11-2019, 08:14 PM

I just don’t think cellular is consistent for all the reasons we know especially with the unknowns of postpaid vs prepaid, over subscribing, meteorological conditions, time of day and tower saturation on and on. Even when I connect one time on 700 MHz then later 1900. Hard to really draw firm conclusions but I do feel good about how I’m setup. I just feel the cellular space is so inconsistent. Drive 10 minutes down the road and it’s a completely new situation.

turbopilot · 02-11-2019, 08:54 PM

Quote:

Originally Posted by NealC

I just don’t think cellular is consistent for all the reasons we know especially with the unknowns of postpaid vs prepaid, over subscribing, meteorological conditions, time of day and tower saturation on and on. Even when I connect one time on 700 MHz then later 1900. Hard to really draw firm conclusions but I do feel good about how I’m setup. I just feel the cellular space is so inconsistent. Drive 10 minutes down the road and it’s a completely new situation.

Yup, sometimes the theory does not translate well to real life. I suspect much of the instability with cellular relates to reflection, refraction and multipath. Somewhere in Dr. Fourie's videos he talks about these problems. Very same problem we have getting reliable OTS TV in a motorhome.

I have done most of my testing inside a steel aircraft hangar with one wall open using a steel door. I have seen the LTE-A signal degrade to LTE or a channel disappear from the television when an airplane just taxis by the hangar. I image the same stuff happens as a cellular signal moves through an RV park with big metal vehicles moving around.

jharrell · 02-12-2019, 06:13 AM

Quote:

Originally Posted by turbopilot

Yup, sometimes the theory does not translate well to real life. I suspect much of the instability with cellular relates to reflection, refraction and multipath. Somewhere in Dr. Fourie's videos he talks about these problems. Very same problem we have getting reliable OTS TV in a motorhome.

Dr. Fourie's videos definitely have good information but he does seem to over generalize on the antenna categorization. Whats missing is real world test data and comparisons among a manufacturer's and designs which is very difficult to find.

The only place I have found this is Rvmobileinternet.com and they only have the resources to test a small number of the ones out there, but their field testing results are very interesting. Some antennas that you would think would be inferior based on cost or design consistently outperforming the ones that based on their specs and outward design. Modern cellular is extremely complicated so many little things add up to your final performance, this will only get worse as time goes on with 64 antenna MIMO and beyond that the modem and antenna array will just have to be engineered together for maximum performance.

I wish I had the resources to just buy the various antennas and cables and experiment with them in various situations, but for now I have to rely on those who share what information is known based on their experience which is why a thread like this is so valuable.

NealC · 02-12-2019, 06:50 AM

I wouldn't trust anyone's tests that are single location results. With cellular and testing antennas I want to hear about results over time, i.e. multiple locations and testing results. We all know one tower, one area, congestion, prioritization, etc. all come into play and then prepaid vs. postpaid. There are just a lot of variables, we all know this is not the same as speed testing wifi in your house which is a constant at the source.

jharrell · 02-12-2019, 06:58 AM

Quote:

Originally Posted by NealC

I wouldn't trust anyone's tests that are single location results. With cellular and testing antennas I want to hear about results over time, i.e. multiple locations and testing results. We all know one tower, one area, congestion, prioritization, etc. all come into play and then prepaid vs. postpaid. There are just a lot of variables, we all know this is not the same as speed testing wifi in your house which is a constant at the source.

Right which again Rvmobileinternet.com are the only ones that have this info. They have field testing results going back a few years in spreadsheet form with multiple antennas, carriers, modems and boosters run at various real world camping locations.

Its still a relatively small sampling of combinations but far better than any other resource publicly available that I have been able to find.

turbopilot · 02-12-2019, 08:07 AM

Quote:

Originally Posted by NealC

I wouldn't trust anyone's tests that are single location results.

Agree.

But I look at the trends not any single test.

I began my search for the best internet when I bought my NA. Here is my history at the same location.

1. Installed PepWave MAX (version before current LTE-A) in my new NA on purchase. PepWave was located in cabinet with rabbit ears. Poor "DSL like" performance. Very disappointed.

2. Installed "cheap" wide antennas on roof with long RG174 cables. Performance increased about 20% download but big increase in upload speeds. Around 7 up and 12 down.

3. Last Spring I got the PepWave MAX LTE-A version. Same antennas. LTE performance about the same but occasionally it would latch on to LTE-A and speeds significantly jumped up to 20 down and 15 up.

4. I added the SureCall antennas after Summer season with shorter antenna lead lengths a saw a small pump in LTE only download speeds but I started to see LTE-A pop on more often with 20 down speeds.

5. A few weeks ago I moved the PepWave to the roof in a box with 12" antenna leads. Now I am consistently seeing down load speeds around 10 mbps down without LTE-A and 20 mbps down when LTE-A is on. Biggest difference with PepWave on the roof is that I see LTE-A engage much more often.

All of this testing in the same location with everything stable except the occasional airplane passing by.

So I am making progress trying to apply the theory to the practical. As Neil said the frustrating aspect is the inconsistentcy in results. Makes really hard to tell if you are making progress, but I clearly have since I started the project.

NealC · 02-12-2019, 08:17 AM

In my December travels with Pepwave I not once needed Weboost which I often did. This in itself tells me this is a definite improvement. But the micrometer testing (pun intended) with the antennas etc. just can't be done in one location. The fact my new antenna latched to a high mhz band yesterday made me believe it is worth it so I'm sticking with it and will test in travels, although over night I lost the band 2 and went back to 700 mhz so I remote switched it over to AT&T which is routinely LTE-A in this area.

To the readers other than the geeks participating here:

The era of WiFi as your Internet source is behind us. Sorry but save your money on WiFi ranger, going forward you need a dedicated device with cellular modem. Jetpacks, Mobley, etc. are all good, at times, but better antennas with MIMO (dual external antenna even if rabbit ears) is going to be better. Some have purchased them for the MiFi devices but it's expensive if you can find them. It seems going forward Peplink/Pepwave and Cradlepoint are the vendors to solve our problems. Call 5GStore.com and speak with them, they will help guide you. A pepwave as Turbo found for us sitting on your dinette with your favorite sim(s) is going to work far better than small hotspot devices. It has wifi which is powerful for all of your devices, it has cellular built in, it even has GPS to help track your coach if it were stolen, etc. The tech space is evolving rapidly, sadly WiFi at campgrounds is not but cellular is, this is the way to go from this point forward IMHO.

NealC · 02-12-2019, 09:13 AM

Here is the Pepwave Load Balancing doc if anyone is interested. There will be a test

NealC · 02-12-2019, 09:15 AM

Here is IP passthrough.

turbopilot · 02-12-2019, 01:33 PM

Quote:

Originally Posted by NealC

Here is IP passthrough.

Thanks.

Been testing "Outbound Policy" in the PepWave. Nominally with two WAN's you would like the router to pick the WAN with the fastest download speed.

In my preliminary tests using "Fastest Response Time" algorithm it picks the WAN with the fastest ping time. In my case locally I have my LTE-A modem as one WAN and a WiFi as WAN connection to a DSL modem as the second modem.

The DSL modem has the slower download speed but a faster ping time than my LTE-A Verizon modem. Verizon modem ping is around 60ms and the DSL WiFi as WAN is around 40ms. As a result it picks the WiFi as WAN source.

So given the available algorithms I don't see a way to set up the PepWave Outbound Policy to select the WAN with the best download speed. Maybe I am missing something. Testing continues.

NealC · 02-12-2019, 01:37 PM

Quote:

Originally Posted by turbopilot

Thanks.

Been testing "Outbound Policy" in the PepWave. Nominally with two WAN's you would like the router to pick the WAN with the fastest download speed.

In my preliminary tests using "Fastest Response Time" algorithm it picks the WAN with the fastest ping time. In my case locally I have my LTE-A modem as one WAN and a WiFi as WAN connection to a DSL modem as the second modem.

The DSL modem has the slower download speed but a faster ping time than my LTE-A Verizon modem. Verizon modem ping is around 60ms and the DSL WiFi as WAN is around 40ms. As a result it picks the WiFi as WAN source.

So given the available algorithms I don't see a way to set up the PepWave Outbound Policy to select the WAN with the best download speed. Maybe I am missing something. Testing continues.

Are the Health Check settings on the source details area used? If so try a different health check?

turbopilot · 02-12-2019, 02:08 PM

Quote:

Originally Posted by NealC

Are the Health Check settings on the source details area used? If so try a different health check?

Health check is active. Don't recall what method but I know it works by failing up and down the WAN's.

jharrell · 02-12-2019, 04:28 PM

Quote:

Originally Posted by turbopilot

So given the available algorithms I don't see a way to set up the PepWave Outbound Policy to select the WAN with the best download speed. Maybe I am missing something. Testing continues.

The Pepwave doesn't know host fast your wan connections are (I think they are adding some kind of speed test in firmware V8) but I believe you can tell it what you think they are manually then use Overflow to get somethign closer to what you want:

https://forum.peplink.com/t/understa...d-policy/15156