Help optimizing small Point-to-Multi-Point link (unstable / bursty connection)

[Elvis1991]

Hi everyone,

I’m trying to optimize a small Point-to-Multi-Point wireless bridge between three buildings and I’m running into stability issues.
I’m a complete beginner with wireless bridges, so I probably did several things sub-optimally and would really appreciate your feedback.

Topology

• 3 buildings arranged roughly as a scalene triangle

• One master / AP bridge in the central building

• Two stations / spokes in the other buildings

• Angle between the two spokes as seen from the master is about 18 deg

• Distance:

  • Master ↔ S1: ~46 m

  • Master ↔ S2: ~450 m

All devices are mounted outdoors with line of sight. The link works, but the connection feels unstable and “bursty”:

• Sometimes latency and throughput are fine

• Other times browsing and remote access feel choppy, with short drops or freezes, even though basic link parameters look acceptable

What I see

From my (very limited) understanding, the basic wireless stats seem OK for this kind of link, so I’m not sure where the real bottleneck is.
I’m wondering whether:

1. My assumptions about distances / angles and device choice are simply wrong for a PtMP setup

2. There might be issues with:

   • Noise / interference

   • Wrong channel width or frequency choice

   • Misconfigured data rates, distance settings, or ACK timeout

   • Antenna alignment / mounting (especially for the 450 m link)

3. Or whether I should be using different hardware for a more reliable PtMP installation instead of my current “amateur” approach.

Questions

Given this scenario:

1. Are these distances and angles reasonable for a stable PtMP link with typical MikroTik gear, or is the geometry itself problematic?

2. Which parameters would you check first to diagnose the “bursty” behavior?

3. Is there any obvious configuration best practice for small PtMP setups that beginners usually miss?

4. At what point would you say, “this is the wrong equipment / setup for PtMP, use X or Y instead”?

I can share exports, screenshots (wireless tables, registration, logs) and exact device models / configuration if needed, but I wanted first to ask if my basic assumptions about this PtMP design make sense at all.

Thanks in advance for any suggestions and for your patience with a newbie trying to build a small radio bridge

.

[BartoszP]

Hi,

Check the CCQ on the third picture. It's 25%.
According to Wireless Troubleshooting - RouterOS - MikroTik Documentation it seems that the connection is not as good as you think.
What is CCQ and how are the values determined?

Client Connection Quality (CCQ) is a value in percent that shows how effective the bandwidth is used regarding the theoretically maximum available bandwidth. CCQ is weighted average of values Tmin/Treal, that get calculated for every transmitted frame, where Tmin is time it would take to transmit given frame at highest rate with no retries and Treal is time it took to transmit frame in real life (taking into account necessary retries it took to transmit frame and transmit rate).

[jaclaz]

@Elvis1991

At what point would you say, “this is the wrong equipment / setup for PtMP, use X or Y instead”?

Right at the beginning.

The 46 m can be bridged with anything.
The 450 m are not so easy.

Or, if you prefer, the 46 m link should be - with your used hardware - "rock solid", the 450 m link hardly so.

The SXT2 has a 60° beam width, so your two stations are well within its cone.
BUT the Sxtsq Lite 2 are also 60° (please read as "very wide" for a station).

If both links, the short and the long one, are unstable, the most likely reason is interferences, the 2.4 GHz is often very populated.

Since they have such a wide beam, it is improbable that antenna aligment would matter, while the placement (fastened to a wall vs. at the top of a wobbly pole) may make some difference.

More recent devices tend to be 5 GHz or 60 Ghz.

There used to be (now MIA since the latest site update) a page that suggested devices as function of frequency used and distance:

Selection guide for PtP links

https://mikrotik.com/products/ptp

but I doubt that it considered the 2.4 GHz for more than a few hundred meters.

[BartoszP]

NV2 is in use so contemporary devices are out of list.

[Elvis1991]

thanks so much for the hints about CCQ and the NV2 limitation!

From your comments I get two main points:

-a CCQ around 25% means the link is objectively bad, not just “a bit bursty”,

-using NV2 ties me to legacy wireless hardware.

Current setup:

1 × SXT 2 as AP at the central building (NV2),

2 × SXTsq Lite2 as stations:

S1 at ~46 m,

S2 at ~450 m

All three are 2.4 GHz with ~60° beamwidth, in a noisy band. The S2 unit is on an old TV antenna support which moves a bit in strong wind; however, the unstable / bursty behaviour appears even with no wind, so I assume 2.4 GHz interference is a major factor.

I’m considering this redesign:

Keep S1 as a short 2.4 GHz PtP (effectively what it is already).

Replace the S2 leg with a dedicated 5 GHz (or 6 GHz) PtP link using narrower-beam antennas and more modern hardware, to get out of the 2.4 GHz noise and reduce what the radios “see”.

At the central building both radios would simply terminate into a bridge / small switch, and from there I’d use a single 802.1Q trunk towards the core router to carry all existing VLANs transparently. The remote ends would also be simple bridges, just extending the VLAN trunk.

My questions:

1. From your experience, is this hybrid design (2.4 GHz legacy + 5 GHz PtP, both bridged, single trunk carrying all VLANs, etc) technically sound, or would you avoid mixing like this and rather redesign both legs at once?

2. Any specific gotchas in RouterOS / NV2 when bridging an NV2 link with a 5 GHz PtP and passing VLAN-tagged traffic (STP/RSTP, HW offload, etc.) that I should be careful about?

If the consensus is that this “partial” upgrade is still not worth it, I’ll plan for a full move to 5 GHz on both legs and drop NV2 entirely.

[jaclaz]

Not from direct experience, only what I would personally do:

1. buy two Sxtsq5, at $65 each they are well affordable
2. try them on the longer link as PtP
3. if the link is fine, buy a Mantbox Ax 15s (179$), replace the Sxtsq5 on the central bulding with it and use the Sxtsq5 on the shorter link too.

All in all a $300 or so investment, if everything works as expected.

In theory, you could get the Mantbox (which Is dual band) and only one Sxtsq5, keeping on the short link your existing Sxtsq lite 2 but you would need to use 802.11 on that link, saving 65$.
Or use two PtP links, 4 Sxtsq5 in total, saving some 50$.

For this kind of money, It seems to me not worth It to keep the old 2.4 GHz device

[Elvis1991]

@jaclaz

function

Thanks a lot for the straightforward advice.

I'm considering LHG 5 or LHG XL HP5 as spokes instead of SXTsq 5 ax, since I found used ones at the same price. The idea is to leverage the narrower beamwidth for less interference and better margin on the 450m link (and maybe the short 46m one too), still in PtMP with a central sector AP. But I kindly say if it has any sense in my setup.

In my setup (central AP seeing two sites at ~46m and 450m, over roofs with some branches in between), do LHG models make sense over SXTsq 5 ax as CPEs, or are they overkill and too alignment-critical for this small PtMP?​

Before buying, does it make sense to run some quick interference tests (like scanning channels with a tool or checking spectrum on current gear)?​

[igorr29]

there is NO need for lhg (let alone the xl version) for 450m link. you’ll overload them with such short distance.

also don’t go the ax route. get an sxt sa ac (still in production, it’s level 4 licence and can act as AP), and get 2 x sxtsq5ac as clients. this should give you pretty decent results.

[jaclaz]

Yes, the LHG's sound like "too much" for those links.

The Sxt sa ac + 2 SXTsq 5 ac is a very good suggestion, with $119+2x$65=$249, i.e. saving some 50 bucks, you get good, proved to be working, hardware that will be more than "good enough" for your links.

And there are several reports that links through the "good ol'" AC devices are more stable than newer Ax ones.

[Elvis1991]

Thanks all, I will try with the 5Ghz solution and I hope to solve the issues. I will keep you updated.

[Elvis1991]

The hub (mANTBox 15s) successfully establishes the PtMP link and traffic is transparently bridged through the hub.
All spokes can access the upstream network without issues.

However, I’m now facing a different problem:

• Traffic passes through the hub correctly.

• Spokes and devices behind them have full network access.

• The hub itself and devices directly connected to the hub do NOT have connectivity (no access to the rest of the network).

So effectively:

• The hub behaves like a transparent bridge for PtMP traffic.

• But the hub’s own IP and locally connected devices are isolated and cannot reach upstream resources.

This looks like a bridge / VLAN / hardware offload issue, where forwarding works, but local access on the hub is broken.

If anyone has seen this behavior on mANTBox 15s (or AX / QCOM-based devices) or can point out what to double-check (bridge settings, VLAN filtering, PVIDs, hw-offload, etc.), I’d appreciate it.

RouterOS version, bridge config, and interface details can be provided if needed.

Attached the configuration of the devices.

cpe_home.rsc (1.8 KB)

CPE_attached_router.rsc (1.8 KB)

HUB_baita.rsc (1.1 KB)