I have two LHG LTE6 devices and both never exceed about 30Mbps, both for up and down connectivity. It almost seems like there is a hard limit somewhere.

When I noticed this in one location, I thought it was because of the poor LTE signal (it's a remote location). But then I installed another device which is roughly 300m from the tower, and got the exact same result. Also, this is with two different network operators, and my iphone can achieve significantly higher speeds.

Am I doing something wrong?

I upgraded the firmware in the LTE modem, which didn't change anything.

This is the interface info (for the location with good signal):
Any advice appreciated!

With my R11e-LTE (without 6) I reached until 60Mbps using LtAP.

Regards.

Rather than doing the carrier aggregation, try choosing only one band. I manage to get 100+ mbps If I set my LtAP LTE6 to use only band 1. If I have the device do the aggregation using band 1 and band 20, it never gets better than 20 odd mbps. Doesn’t make sense to me but that’s how it works. I initially did a scan, identified each band that was discovered, and then tested each one individually.

No it can do a lot better...
What is your RSRP, RSRQ, SINR and RSSI values ?
What is the Bands ( and CA band ) your device uses ?

Rather than doing the carrier aggregation, try choosing only one band.

He's right. I restrict my failover link to B3 because when it does CA2 with B20 (5MHz) + B3(15MHz),
it will transmit very poorly over B20 first and then start speeding up with constant use.
I'm using an LtAP LTE6 kit so the only difference in our links is the antenna sensitivity (yours is 17dBi and mine is 5dBi with MTAO-LTE-5D-SQ).
I can get CQI 13-14, RI 2 even with low power levels on B3 only primary at 15MHz compared to CQI 4-9, RI 1 using CA2 with B20 primary at 5MHz and B3 at 15MHz.
When it idles using CA2, this drops to the same slow coverage to the point that DNS/DoH drops out and it becomes unusable.
I think this is either due to the tower's preference or the radio's preference in band transmission preference but either way, it's an impediment to a stable failover link for me.

The only point CA makes sense is when you have good signal across all expected bands, otherwise it makes sense to lock it to something for a more stable connection.
In my case, the failover went from 5Mbps up to 17Mbps on CA2 on to 43Mbps - 75Mbps with B3 in use only.
Ping maxes around 45ms but usually gets 17 - 23ms which is very decent for an LTE link.

I don't use cell lock because sometimes the cell ID's change either due to maintenance or whatever the MNO is doing.
Best to keep the entire band visible for any cells operating on that frequency.
Keep in mind your setup is more for directional, last mile connectivity so using CA2 seems like an inefficient use of this.
As always, YMMV so test, test again. See what works.

Also, the master of cellular MikroTik stuff here is SiB.
They have a very authoritative post on everything LTE.
Try the logger scripts to get a better idea of what your network is doing.

You can find the post here: viewtopic.php?t=178312

There are multiple reasons as to why you might experience low bandwidth in any type of network connection...
Personally before disabling the CA band i would check the RSRP, RSRQ, RSSI, and SINR values separately for each band...
If those values are not within the expected range, i would try to improve them... That might be changing position of the antenna, firmware upgrade etc... Then making tests again...

Those 30 Mbps can be simply because of interference, bad aiming, wrong band and more...

Those 30 Mbps can be simply because of interference, bad aiming, wrong band and more...

Correct or your link's CA2 may be flapping between primary nodes, causing low overall link quality due to the time factor of backing off the negotiated band and renegotiating primary/secondary with a different cell, a scenario where cell lock can be useful.
You have to evaluate the problem or the next time you have it, you won't know what has changed.

Maybe yes... maybe no...
However i would first check other parameters before disabling the CA Band... That's why they exist in the first place... so that you can evaluate your connection...

Hmm, interesting suggestions — I thought CA was an obvious "keep always on, because it is always better" setting.

For those asking about bands used, RSRQ, SINR, etc — all those are in my post (just scroll the "code" section down so that it shows the entire info output). As for firmware updates, done already. As for aiming the antennas, I did plenty of work on each one, spending many hours finding the best position.

And again, this is something I'm seeing in two different places, with two different network operators.

As for speeding up during use, that is something I'm seeing as well — even when using the "speedtest" app, after 3-4 seconds you can see a definite increase in speed, but then it seems to hit that 30Mbps limit and never exceeds it.

As for speeding up during use, that is something I'm seeing as well — even when using the "speedtest" app, after 3-4 seconds you can see a definite increase in speed, but then it seems to hit that 30Mbps limit and never exceeds it.

As much as I wish I could diagnose that to my earlier problem, I think what Zacharias said is important.
Try disabling all bands and test one at a time to get average RSSI, RSRP, SINR, RSRQ and optionally CQI + RI under load.

Sometimes a lower power (low RSRP but good RSRQ) band will be more responsive due to less crowding.
B20 (800MHz) is usually quite saturated but of course this depends on the area the tower is serving.
B7 (2600MHz) is normally for short range/high density coverage so if this is negotiated as your primary CA2 band then you might have slow throughput until traffic is shifted onto B3 (1800MHz) either by tower decision or the LTE6's firmware decision.
Or it could be that your B7 signal is just at the threshold with served density + distance giving you poor throughput.
That's the scenario I experience with B20 + B3 CA2 aggregation and I've verified the throughput issue by using a USRP E312 to watch traffic gradually decrease over B20 and increase over B3.

Also, your output indicates an RI of 1. An RI of 2 here would be ideal for CA2.
To be highly reductive of the LTE standard , an RI of 1 means UE is receiving downlink data from eNB in TX Diversity mode.
An RI of 2 means UE is receiving downlink data from eNB in MIMO mode. This is the only case where CA2 really shines.

There's a lot to consider here but for stability reasons, I'd try testing each band independently and looking at both speed test throughput and overall responsiveness.
Maybe you're just negotiating the wrong CA2 band pairs and you can test this by disabling all but the two bands of a pair you want to test and then testing it.
CA2 can double throughput under ideal conditions but it can also halve it under less than ideal conditions.

As always, YMMV. Test, test again. See what works!

Edit: clarification on what I meant by lower power band, specific to my scenario.

This source summarises it better than I did. https://www.sharetechnote.com/html/Handbook_LTE_RI.html

RI (Rank Indicator, Rank Index)

To have clear understanding of RI, you have to understand the detailed concept of MIMO, Channel Status Information Matrix and mathematical definition of Rank of a matrix. Depending on instruction from the network, UE may periodically or aperiodically measure RI and report it to Network. Refer to CQI,PMI,RI Report section for this.

But I will explain the practical/intuitive meaning of RI here.
In simple words, RI is an indicator showing how well multiple Antenna work. What do you mean by "how well the multiple Antenna work ?". We usually say "Each of the multiple antenna (e.g, each antenna in MIMO configuration) works well if the signal from each antenna has NO correlation to each other". "No correlation" implies "no interference to each other".

Maxmum RI value is very closely related to the number of Antenna. Maximum RI is same as number of antenna on each side if the number of Tx antenna and Rx antenna is same. If the number of Tx and Rx are different, the one with less antenna is the same as Max achievable RI.

Max RI means "No Correlation between the antenna", "No interference to each other", "Best Performance".
For example, in case of 2x2 MIMO, the RI value can be 1 or 2. When the value 2 in this case means "No Correlation between the antenna", "No interference to each other", "Best Performance". If the value is 1, it implies that the signal from the two Tx antenna is percieved by UE to be like single signal from single Antenna, which means the worst performance.

Ok, so I finally found some time to test this.

First, I don't seem to be able to turn off CA completely — even after locking the modem to a specific EARFCN/cell-id pair, the ca-band appears in "interface lte info lte1", at least initially. It does seem to disappear after a minute or so, not sure what to think of that.

Second, after spending some time trying a lock with all available bands, the results are very consistent: I am always limited to around 28Mbps, with the exception of one band which seems to be crowded and I get speeds of 14-18Mbps.

In other words, no amount of cell-locking improves my situation. I seem to be hitting a consistent hard limit of ~30Mbps. I now reverted to the default configuration, which usually settles to:
At least I get the flexibility of automated switching in case of network reconfigurations/failures. But I am still limited to 30Mbps.

Why would that be?

When I noticed this in one location, I thought it was because of the poor LTE signal (it's a remote location). But then I installed another device which is roughly 300m from the tower, and got the exact same result. Also, this is with two different network operators, and my iphone can achieve significantly higher speeds.

Some network operators limit the bandwidth when they detect that the SIM card is installed on a PC/tablets or router. They give full bandwidth only if the SIM card is installed in a phone. (They can differentiate both with the modem IMEI)
That would explain why you have better performances with your phone.

For instance, Vodafone Portugal limits the bandwidth for pre-paid card installed on PC/tablets/routers to 40Mbit/s DL/UL, but allow full speed if installed on phones.

The above statement is true. Particularly, when the mobile operator provides unlimited data, they introduce restrictions such as the above.

Are you still using Orange Poland?
If you change the TTL does this increase the speed: https://wiki.mikrotik.com/wiki/Manual:Interface/LTE

I am seeing this limitation in two locations, one with Orange Polska (this is where I am 300m from the tower, all the numbers in this post are from this location), and another with Plus.pl (this is a remote location, so I wasn't sure if the limitation isn't due to weak signal).

In both cases there is a very precise limit just short of 30Mbps, usually shows up as ~28Mbps in SpeedTest measurements.

I can't change the TTL, because the LHG LTE6 works in passthrough mode, the routing is done by Ubiquiti UDM Pro routers.