Hi,
what’s about the WLAN radiation pattern? Elevation 1 and 2 shows the radiation from horizont (0°) to azimut (90°). Ther’s no pattern to -90°. Therefore I think radiation below the antennea is weak. Added to this is the aluminum roof skin, and 5 GHz is shielded more than 2.4 GHz.
In our mobile medical examination vehicles (RV’s), we connected the WiFi antennas to the router inside and not to the roof antenna.
A bad example from our experience:
Truck with driver’s cab (metal). The antenna is installed on the roof of the driver’s cab and the router is installed in the driver’s cab. WiFi antennas are inside connected to the router.
The examination cabin, also with a metal shell, is separate from the driver’s cab (approx. 30 cm - 11.8’ - between the cabins).
Only -85 dB at 2.4 GHz reaches the examination cabin.
An attempt to position the two WiFi antennas between the driver’s cab and the examination cabin resulted in only minimal improvements: -78 to -80 dB.
Another tip is to get a WiFi analyzer app - I like WiFi Explorer which runs on macOS, but there are similar tools for Windows.
It scans for other WiFi signals and makes graphs of channels vs. base station ID and signal strength, which can be invaluable for diagnosing weird slowdowns.
@snoack The cadillac solution would be to get two Peplink AP One Minis, put one in the front of the RV and one in the rear, and set them to fixed 2.4GHz and 5GHz channels that don’t overlap (which you may need to alter when you get to a new RV park if it’s crowded).
Another trick: If you don’t care about FAST internet, but more about RELIABLE, then you can use the “secret” channel 165 as discussed here:
In short, it’s a 5GHz channel that can only be 20MHz wide, which means it’s often unused.
I ended up ordering another MAX BR1 Pro 5G and did some tests with the router replaced as well as using the paddle antennas vs the 42G installed on the roof.
| Router | Antenna | Location | Laptop (link quality) | Android (RSSI) |
|---|---|---|---|---|
| Old | 42G | Couch | 57–59 % | −57 to −66 dBm |
| Old | 42G | Bed | 25–35 % | −79 to −85 dBm |
| Old | 42G | Dinette | 45–50 % | −63 to −75 dBm |
| Old | 42G | Outside | 30–35 % | −72 to −81 dBm |
| Old | 42G | Antenna / Roof | — | −43 to −51 dBm |
| Old | Paddle | Couch | 65–75 % | −49 to −54 dBm |
| Old | Paddle | Bed | 47–49 % | −62 to −73 dBm |
| Old | Paddle | Dinette | 69 % | −44 to −56 dBm |
| Old | Paddle | Outside | 46–50 % | −67 to −69 dBm |
| Old | Paddle | Antenna / Front | — | −26 to −41dBm |
| New | 42G | Couch | 60–67 % | −55 to −65 dBm |
| New | 42G | Bed | 29–44 % | −76 to −85 dBm |
| New | 42G | Dinette | 54 % | −62 to −69 dBm |
| New | 42G | Outside | 37–40 % | −72 to −76 dBm |
| New | 42G | Antenna / Roof | — | −44 to −47 dBm |
| New | Paddle | Couch | 75–84 % | −46 to −49 dBm |
| New | Paddle | Bed | 42–54 % | −65 to −77 dBm |
| New | Paddle | Dinette | 74–84 % | −47 to −53 dBm |
| New | Paddle | Outside | 40–62 % | −67 to −70 dBm |
| New | Paddle | Antenna / Front | — | −20 to −47 dBm |
I only tested on the 5GHz band (channel: 48, channel width: 40MHz), everything else set to default settings (max. output power, no boost, etc.).
- The data suggest that the router is indeed fine as replacing it doesn’t seem to improve the signal meaningfully.
- Switching to the paddle antennas seems to improve the signal strength significantly in all locations around the RV, by an average of ~10dBm.
- I also tested with an AP One AX Lite as dedicated access point which seems to further improve the signal in all locations.
I think what @Philip_Hendrickse said makes sense that the 16’ long antenna cable is dropping the signal too much for 5GHz WiFi. The signal strength on 2.4GHz when using the 42G seemed fine though but given my environment, I assume the 2.4GHz band is just too congested to provide a reliable connection.
I’m curious though if the cable length can cause issues for cellular signal too?
Depending on the frequencies used by the provider (800 MHz in rural areas, up to 3 GHz (or even 6 GHz) in cities), the cellular signal is affected in the same way.
A look at the data sheet for the cable used lists the loss for different frequencies—the higher the frequency, the greater the loss.
Every 3 dB halves the power.
As listed by Philip Hendrickse:
Cable loss of CFD200 at 5 GHz is 0.87 dB/m ~ 0.265 dB/ft = 4.24 dB/16 ft
Calculation:
Cable Loss at 16 Feet
The cable loss scales linearly with length. Multiply the per-foot loss by 16 feet for total attenuation in dB.
Loss Calculations
At 5 GHz:
Loss = 0.265 dB/ft × 16 ft = 4.24 dB
At 2.4 GHz:
Loss = 0.1676 dB/ft × 16 ft = 2.6816 dB
At 1.8 GHz:
Loss = 0.1402 dB/ft × 16 ft = 2.2432 dB
At 5 GHz:
Attenuation = 0.265 dB/ft × 16 ft = 4.24 dB
Power loss factor = 10^(-4.24/10) = 0.378 (37.8% of transmitted power reaches antenna; 62.2% lost).
At 2.4 GHz:
Attenuation = 0.1676 dB/ft × 16 ft = 2.6816 dB
Power loss factor = 10^(-2.6816/10) = 0.543 (54.3% reaches antenna; 45.7% lost).
At 1.8 GHz:
Attenuation = 0.1402 dB/ft × 16 ft = 2.2432 dB
Power loss factor = 10^(-2.2432/10) = 0.600 (60.0% reaches antenna; 40.0% lost).
Summary Table
Frequency | Attenuation (dB) | Power at Antenna (%) | Power Lost (%)
5 GHz | 4.24 | 37.8 | 62.2
2.4 GHz | 2.6816 | 54.3 | 45.7
1.8 GHz | 2.2432 | 60.0 | 40.0
Thanks for the explanation, that is very insightful! I think I will leave the antenna for now for cellular connectivity.
If I ever decide to reinstall the antenna, can I just cut the cable an re-terminate the SMA connectors myself (with tools that don’t cost a fortune)? Or is it possible to connect a new cable harness directly to the 42G (it seems the cables are hardwired on the antenna end)?