I’ve two 4G/LTE SIM cards (with a plan of upto 100 Mbps download and upto 50 Mbps upload) from the same wireless ISP,
a MikroTik LTE6 router ( https://mikrotik.com/product/hap_ax_lite_lte6 ), and an LTE4 router that is non-MikroTik.
The MikroTik LTE router is very slow:
(WiFi was not used in the above tests, rather the test PC was attached to the router via LAN cabling)
(The tests were performed using the “speedtest-cli” program/package from the Debian Linux repository)
Eventhough the MikroTik router has LTE6 (ie. max 300 Mbps possible),
and the other router is just LTE4 (ie. max 150 Mpbs is possible), the result is
that the MikroTik router achieves only about half the speed of the other router…
Paradox, isn’t it? How come? How to diagnose and fix this in RouterOS ?
Btw, the non-MT router did cost only 1/3 of MT… , but is two to three times faster…
How should we know from your speedtest output only?
Your “1/3 of rice” device is for sure not of any help I guess. Or does this other device show connection information somewhere?
For your MT device, lets start with and post the output here (remove IMEI or serials):
Yes, a cheap rice device.
No connection details.
It has not even a log file!
The maybe only interesting setting is this:
Set Frequency Band: 1/3/5/7/8/20/38/40/41
And the APN settings, which is mostly “auto”.
For your MT device, lets start with and post the output here (remove IMEI or serials):
rsrp isn’t that bad. But your sinr is quite low.
LTE6 supports 2CA (one of the major differences to CAT4). But your device is only connected to one cell. Ao you cant benefit from carrier aggregation. But CA is done by the operator - so you cant influence that parameter.
You could use the cell monitor to find other cells around you and then use a cell lock to connect to a specific cell. Cell lock command is described for your modem, FG621-EA, in the Mikrotik LTE docs: https://help.mikrotik.com/docs/spaces/ROS/pages/30146563/LTE
Another possibility, band 20 is often combined with band 3. I don’t know which bands your provider has near your location, but instead of cell lock you can try the “band”-setting in /interface/lte menu. Maybe band 20 is heavily overloaded and thus these crappy speeds.
In the other router I had to change the country from default China to Germany.
Is there a similar setting in RouterOS for LTE modem?
(Since each country can have its own set of bands, channels, frequencies etc…)
And I wonder how “Automatic” APN settings function.
Is that possible in RouterOS as well?
What does this mean if I add band 20 and 3 there?
One of band 20 or 3 gets selectively used, or are both used at the same time?
What if I add all bands there? Ie. all of the following under “Capabilities / LTE Bands: 1,3,5,7,8,20,28,38,40,41” .
Is my said MikroTik LTE6 router ( https://mikrotik.com/product/hap_ax_lite_lte6 ) maybe lacking an important feature
that the above maybe has, that would explain the speed difference?
It’s bugging me; I just try to understand why my MT LTE6 router is much slower than the above LTE4 router.
We have this problem: the MikroTik LTE6 router achieves only half the mobile connection speed of the CPE106-E (LTE4 router).
WiFi is not used in these speed tests, rather the PC is attached to the router via LAN cable.
How to explain this illogical speed discrepancy?
Answer:
The speed discrepancy between the MikroTik LTE6 router and the CPE106-E (LTE4 router) can be explained by several factors:
Real-world performance vs. theoretical speeds: While the MikroTik LTE6 router is theoretically capable of higher speeds (300Mbps downlink), actual performance often falls short due to various factors such as signal quality, network congestion, and hardware limitations[2][4].
Firmware issues: Some users have reported performance issues with MikroTik LTE devices after firmware updates. For example, one user mentioned that “reliability, speed and signal quality all dropped after I upgraded to v027”[8].
Hardware efficiency: The CPE106-E may have more efficient hardware or better optimization for real-world conditions, despite its lower category rating[2].
Carrier aggregation implementation: While the MikroTik LTE6 supports carrier aggregation, users have reported issues with it not working consistently, which could limit its performance[2].
Antenna design: The CPE106-E might have a superior antenna design, allowing for better signal reception and, consequently, higher speeds[1].
Device-specific issues: Multiple users have reported underwhelming performance with MikroTik LTE devices compared to other brands or lower-category devices[2][4][6].
To address this issue, you could try:
Updating or downgrading the MikroTik router’s firmware
Adjusting the router’s settings, such as band selection and carrier aggregation
Improving the antenna setup or positioning
However, it’s important to note that many users have reported similar experiences with MikroTik LTE devices underperforming compared to lower-category devices from other manufacturers[6]. This suggests that the issue might be inherent to the MikroTik LTE product line rather than a problem with a specific unit.
I think the AI is right with this:
5. Antenna design: The CPE106-E might have a superior antenna design, allowing for better signal reception and, consequently, higher speeds[1].
Indeed, the CPE106-E has two big antennas (cf. img below), whereas the MikroTik LTE router does not have any such big antennas…
I was about to write that I disagree with assessment by @infabo:
RSRP at -97dBm is already bad, but SINR at 3dB is really bad. So whatever gain, provided by those antennas, will help tremendously.
As to other bands: usually mobile terminal connects to strongest cell and in idle mode MNOs usually don’t steer devices into other (worse) cells/bands. Seeing modem connected to B20 cell (800MHz) makes me believe that any other bands that might be available (B3 is 1800MHz) will have even lower signal strength. It might add something when used in CA configuration, but far from miracles.
I haven’t said -97dbm rsrp is good either. by “not that bad” i meant actually: should be good enough to achieve 9mbit/s (if sinr wasnt so low). And OP does not expect miracles, just: 9mbit instead of 3. And with 2CA I am pretty sure this could be achieved. But these LTE things are hard to troubleshoot. I think the antennas of this little hap ax are too weak. These big antennas on the Chinese device are most probably for wifi. But in the bigger housing there is maybe also a better LTE antenna.
Agree. I’ve yet to see MT device with at least half decent antennas for frequencies lower than 2GHz. If that Chinese device has antennas with 0dBi gain, then it’s likely better than MT by a few dBi.
Indeed. It helps if one knows a bit … and I have 17 years of engineering experience with 2G, 3G and 4G, from designing network, building it and optimizing/troubleshooting it. The later always involved more or less bad customer devices.
Carrier aggregation is a must to get decent speeds with CAT6.
I’m pretty sure the China box won’t do much better than the MT if properly configured.
Most built-in external antennas on 4G CPEs are used for Wi-Fi nowadays, not the LTE radio. For example, with Huawei CPEs, customers need to buy them separately.
So there are a few things to note here. In my limited experience, the hAP AX Lite LTE6 is useful only in a few scenarios:
When close to your providers tower, preferably with no other towers nearby (less than 800m away)
As a travel device
I used one when I was living about 800m away from our provider’s tower and the best I could do was around 30mbps d/l. Granted, I couldn’t see the site but there were also no other towers nearby to interfere. By contrast, an SXTLTE6 Kit mounted outside achieved the max 100mbps from the same tower. I also tried a Chateau LTE6 box in the same place as the hAP and that did markedly better, but not as good as the SXT.
I would suggest the reason for your low throughput would be one of:
You being too far from your provider’s tower, possibly evidenced by the tik choosing B20 (the strongest signal) which uses a very low bandwidth of 10MHz
There are other towers also in your area which maybe interfering with your signal, made worse if you are elevated and can therefore see multiple towers
I found by using a couple of pigtails connected to the modem (decoupling the internal antenna on the hAP device) and then connecting an outside antenna resolved the issue altogether. I only did that though because I didn’t mind voiding the warranty. The question though is what type of antenna to use in that scenario? I got away with using two simple whip antennae because I didn’t have any other towers around that would cause interference. A much more directional antenna would be advisable if the tower is far away or there are multiple towers.
My personal advice would be to skip the hAP and go with the SXTLTE6 kit or if you can extend yourself, either a Chateau or the new LHGGLTE18 kit. Just remember, if you use either kit, you’ll also need a hAP AC or AX device to deliver wifi.
I have a SXTLTE6 kit in tandem with a hAP AC Lite tower which works great (most of the time) when we are housesitting for people who either don’t have reliable internet or their provider data caps their connection (usually semi/rural folk).
@KiwiBloke, as said: SIM tested with the said 2 devices. One is good/acceptable, the other one (MT) is very bad in speed achieved.
The speed of the MT device is not even enough for streaming from my streaming provider: every 3 to 5 minutes it stops,
and reloading the page then continues the streaming for the next 3 to 5 minutes… That’s of course inacceptable…
MT should offer add-on antennas for such wireless devices for Average Joe users.
I’m not an expert in wireless stuff, so I don’t know how to extend this device with (directionally adjustable) external antenna(s).
It seems one needs even to solder something…
I’m very unsatisfied with my “MikroTik hAP ax lite LTE6” ( https://mikrotik.com/product/hap_ax_lite_lte6 ),
and I feel somehow “betrayed” since I had deliberately picked LTE6 over LTE4 as it promised twice the speed,
but actually it has only half of the speed of LTE4 !!!
So I wonder what to do with this device as it’s practically unusable in my case Should I ask for a refund?
Whenever it’s up to user to select appropriate equipment, it’s that user’s homework to find out any constraints, such as extremely low signal strength available. And then search for appropriate solutions, such as need for using external directional antennas, consequently need for modem which makes attachment of those antennas an easy task.
That’s the reality which comes with freedom of not being forced to use ISP-provided equipment … if user is forced to use ISP-provided equipment, then it’s up to ISP to provide appropriate gear.
In no way is that responsibility of equipment vendor. Yes, sometimes things could be described more clearly … but not telling about how to attach external antennas in my book is the same as saying that’s not possible. And not publishing antenna radiation patterns (which MT almost never does) is almost the same as admiting that their antennas are garbage. But it all boils down to user who has to select the right equipment for a particular use case, based on own situation and equipment assessment.
Yes, you may want to ask for refound, it’s pretty obvious that this device won’t help you solving your problems. I hooe your dealer doesn’t ask for reason for returning because device does work according to it’s specs (and reason for your dissatisfaction is something else).
I now have even switched to the development version of RouterOS,
and after that had also to install a newer “LTE Firmware” version,
but LTE speed continues to be very bad
BTW, under Tools there is also a MikroTik Speed Test tool in RouterOS.
But one has to enter a server IP address. Which server(s) are there?
And how to setup such an own test server? Ie. what port, protocol, file specs for downloading/uploading etc.
For ROS Bandwidth Test you can use Tom Jones (from North Idaho ) public test server, see: http://forum.mikrotik.com/t/public-mikrotik-bandwidth-test-server-s-now-shutdown-as-of-april-1st-2025/94863/1
Or you can setup your own on remote site with public access to Bandwidth Test ports, for eg. on CHR running somewhere on VM.
Also MT Bandwidth test Windows application (GENERAL section) can be used for such purpose, it can be act as server or client running on machine connected on different internet connection than router with more bandwidth than you have on router, ofc. ports needs to be exposed for side which will be server (better on machine with more CPU power).
In your case I don’t believe you will have better results with ROS Bandwidth Test than with any other common public speed tests.
Thanks.
Testing via the Tools / Speed Test page in WebFig.
It seems to test also UDP, but the result fields are always empty.
And what does “Loss 100%” mean?
, but is two to three times faster…
