Chateau 5G ax 4G internal antenna

[nsmith1979]

Hi all,

I've recently purchased a Mikrotik Chateau 5G ax, to replace my Zyxel NR5101, as my home 5G router. I've got an external antenna already setup (Poynting v3 5G) and get a good 5G signal but the 4G is not so good. 

With the Zyxel router it would connect to my local mast (350m away) on band 20 (Vodafone) and give a reasonable signal but the Chateau does not provide ENDC when it connects to the same mast / band so I don't get a 5G, which seems a bit strange. The other 'best' option is connecting to band 1 from a tower that's about 850m away (which is also the 5G mast in my area) but the signal isn't great. 

5G Signal:
RSRP: -84dBM / SINR: 23dB / RSRQ: 11dB

4G Signal (band 1):
RSSI: -69dBm / RSRP: -111dBm / SINR: 2dB / RSRQ: -16 dB

I want to try and improve the 4G signal so was thinking of putting 2x SMA ports in the spare openings on the back of the Chateau and connecting them to A0 A1 on the modem, then using the 'bunny ear" antennas (the ones that come with it) to see if it improves the signal (moving the router would mean not having the 5G antenna connection).

I'm after some advice on whether this would be worth doing or not (would the internal antenna be better).

[Seán]

I suggest doing a modem firmware check by running the following command in the Winbox Terminal just in case the issue is with older firmware:

/interface lte firmware-upgrade lte1

From a user that ran modem ports A0+A1 to the external SMA ports, the rabbit ear antennas provided 3dB to 5dB improvement over the internal antennas going by its post. This could also be due to the polarisation as usually rabbit ear antennas are positioned in a V shape (roughly +/-45 degree polarisation), where as the internal antennas are horizontal and vertical. I assume 5G is coming in on band n78 as the modem antenna configuration is different for 5G on the lower bands, typically A0+A1+A6+A7 for 5G on bands n1, n3, etc.

This would also let you test how well the outdoor antenna performs on band 1 by connecting it to A0+A1 temporarily to see if it's worth getting a second outdoor antenna. You can also try a compromise by connecting the outdoor antenna to the SMA ports running to A0 and A4 internally, connecting the rabbit ears to the other two SMA ports. This should provide a higher upload speed as A0 and A4 mainly handle the uplink.

[nsmith1979]

Thanks for the reply Sean,

The modem is running the latest FW (RG502QEAAAR13A03M4G_02.001.02.001), 5G band is n78.

One thing I don't understand is how the Zyxel router was able to provide 5G NSA on band 20 (or band 1 & 7) from my local mast but the Chateau doesn't, they use the same modem?

I've ordered the U.FL - SMA pigtails so will try connecting A0 & A1 to SMA ports and do some testing.

[Seán]

I just thought about a similar issue posted a while back in this thread and on MiroTik's forum where a few users could no longer get 5G NSA after upgrading their firmware to R13A02 from R11A##. Although we're now on R13A03, I wonder if there is something in this firmware that could be causing this as it's quite possible your NR5101 is running older or vendor customised firmware on its modem.

With my Chateau 5G, the biggest change I experienced was a few years back when I upgraded from R11A03 to R11A04, the band 3 carrier RSRP suddenly dropped down to <-120dBm. It turned out that the firmware changed the antenna port configuration for 4G upload CA to ports A6+A7 which I had on internal antennas at the time. This meant I had to get a second outdoor antenna and do internal rewiring to connect to modem ports A6+A7 as there is no way that I'm aware of to disable upload CA, even on the latest firmware.

I suggest raising a support ticket with MikroTik mentioning that your NR5101 with the same modem operates fine on 5G NSA with using the same 4G band and cell ID.

[nsmith1979]

I've raised a ticket with Mikrotik, I'll see what they come back with.

I've installed the extra SMA connectors onto A0 & A1, the bunny ear antennas didn't really improve things too much so I used a pair of SMA splitter pigtails (2x Male to 1x Female) to connect all four ports to my Poynting directional antenna (not ideal I know). The 4G signal has increase by over 10 dBm, all the signal parameters are much better for 4G. I've lost about 3 to 4 dBm on the 5G signal but it's still 88 to 90dBm, SINR and RSRQ haven't changed.

Download is 300Mbps (+) Upload is 45Mbps (+) so not too bad.

There isn't much room for the SMA connector for ANT 1 on the back on the router inside the housing, the modem sits right in front of it so the pigtail is bent quite a lot, I've ordered a 90 degree U.FL to SMA pigtail which will hopefully fit nicely into the space. I'll see if it improves things further on 4G.

[nsmith1979]

OK - so I got a pair of better 'bunny ear' antennas until I get round to installing another external antenna for the 4G (I have a Poynting wideband omni antenna I'm not using so will mount it soon). The 4G signal is better then it was and I'm still get 300Mbps D/L and 35Mbps U/L.

Mikrotik were no help with the ENDC connection, just said it will be down to the operator.

One other thing I've noticed which seems strange, I have ANT4 & 5 as the 5G external SMA connections (as it comes out the box) but if I swap them over the signal values change by quite a lot. It will go from RSRP 85dBM to 110dBM if they are switched over, same 5G cell, surely that isn't right?? And also the 4G signal parameters change as well, they get much better when the 5G gets worse?

I'm wondering if the latest firmware could of changed the port mappings (not sure if anyone else has noticed it). I've asked Quectel to send me the latest port mappings for the module so will see if anything has changed. Or, I guess it could be a faulty module maybe.

[Seán]

If you mean swapping the 2 coaxial leads between the SMA ports that run to A4 and A5 internally, it's possible 5G is coming in stronger on one polarisation over the other. However, I would not expect a 15dB difference, like something is obstructing the signal in one polarisation. The modem does use A4 for one of the second MIMO pair (A0+A1+A4+A6 for 4x4 MIMO), so swapping the leads would change the polarisation provided for this connection.

Pity to hear about MikroTik's response, I thought they would have at least tried testing older modem firmware.

From looking up the internals of the Zyxel NR5101, unfortunately its RG502Q-EA modem is non-removable to try swapping.

[nsmith1979]

Hi Sean - So, I've installed a 2nd XPOL-5G-V3 antenna on my pole so I now have:

A0+A1 - Antenna #1
A4+A5 - Antenna #2

Antenna #1 is about 60cm lower down the mast than antenna #2, signal strength is OK with D/L 350 - 390Mbps & U/L 40 - 60Mbps.

Would I be better using A4+A6 for the 5G connection?

   

[Seán]

It's worth giving A4+A6 a test run as A4+A6 is the MIMO1 pair for 5G and the MIMO2 pair for 4G, i.e. this should provide 4x4 MIMO in 4G and 2x2 MIMO in 5G.

This may improve your download speed as the main 5G will be getting cross-polarised signals on its primary MIMO instead of combining each MIMO pair with an internal antenna, i.e. A4+internal A6 for MIMO1 and A5+internal A7 for MIMO2. The upload speed may drop depending on whether it was using port A5 for upload as 5G n78 can transmit on both ports A4 and A5.

[nsmith1979]

OK - so swapping A5 for A6 has increased the download speed and the upload hasn't changed, in fact it's a little better staying around the 60 - 65Mbps mark. Download is 420Mbps so definitely an improvement.