Hi!

TL;DR:
When two MikroTik routers are bundled together, wireless 2.4 GHz traffic breaks down even when on separate non-overlapping channels. Why?

Long version:
Not the best title, I know. Anyway, I have a hAP ac2, quite neatly (IMO) set up as a travel router. I can connect to a network via wifi or ethernet. When connecting via wifi, I use the 2.4 GHz radio, while keeping the 5 GHz for the LAN side of things. I use a mangle rule to be a bit naughty with TTL to disguise that a whole network is connected and once I have cleared the captive portal of the wifi/network I'm connecting to, with a client device (my phone for example), I can just press the "Mode" button on the router, which is programmed to run a script that connects to my home network via WireGuard and then routes all traffic via WG. Seems to work pretty well! However, the fact that the 2.4 GHz radio is occupied with the WAN connection, means I have no means of connecting any 2.4 GHz only devices to the router. Now, almost all stuff I have can use 5 GHz, but I have a few exceptions. So my plan was to configure a hAP lite I have lying around, as an AP for any 2.4 GHz stuff and literally bundle it to the hAP ac2 with a few cable ties and then power it from the USB port on the ac2, so I would have a neat little router package with a minimum of cables that could be powered from a single power supply. However, I discovered the hard way that when they're close to each other, the 2.4 GHz traffic basically grinds to a halt, even if they're on separate non-overlapping channels. I thought that interference wouldn't be a problem, as long as non-overlapping channels were used..? What is causing this? Is it normal or is the hAP ac2 somehow extra sensitive to such interference? I don't know much about RF really...
I know you could set up some virtual devices and use the same 2.4 GHz radio for both the WAN connection and any 2.4 GHz LAN devices, but I don't have the best experience with that and the channel for the LAN devices is then dictated by whatever channel of the WAN connection.
So I guess my last question is if there is a way to achieve what I wan't, without having the router and the "AP" a meter plus apart from each other and cables between them on the floor.

what protocol settings are you using? is NV/NV2 deactivated so both APs only use 802.11?
if not, try to lock all wifi interfaces to only use 802.11

The problem is power pre-amplifier (PPA) in receive path and its automatic gain control. It gas to amplufy analog received signals so that they enter the analog-digital converter at certain level. The problem is that PPA doesn't know the exact frequency used and amplifies the whole 2.4GHz band ... including the strong "interferer" on adjacent channels. The interferer, however, defines the gain figure which means that wanted signal (from farther wireless peer) doesn't get amplified enough.

Solution to the problem is to use same radio for both station mode (upstream connection) in master mode and AP mode (slave) ... which has a few gotchas but on device with 5GHz radio they are not that important most of time.

Wouldn't some tin foil between the two devices (RF shielding) have the same effect of having the devices a few meters apart?
It costs nothing to test, and if it works one could experiment with a very fine metal mesh (mosquito mesh) that could be simply glued/fixed to one side of the hAP Lite.

what protocol settings are you using? is NV/NV2 deactivated so both APs only use 802.11?
if not, try to lock all wifi interfaces to only use 802.11

I'm pretty sure I haven't touched anything named NV/NV2 (can't really check right now), but for protocol I believe it was set to 802.11 on the hAP ac2, but on the hAP lite it was set to "any". I tried both "any" and 802.11 on both devices, but I'm not sure I noticed any difference. But am I right in that under normal circumstances only 802.11 is required?

The problem is power pre-amplifier (PPA) in receive path and its automatic gain control. It gas to amplufy analog received signals so that they enter the analog-digital converter at certain level. The problem is that PPA doesn't know the exact frequency used and amplifies the whole 2.4GHz band ... including the strong "interferer" on adjacent channels. The interferer, however, defines the gain figure which means that wanted signal (from farther wireless peer) doesn't get amplified enough.

Solution to the problem is to use same radio for both station mode (upstream connection) in master mode and AP mode (slave) ... which has a few gotchas but on device with 5GHz radio they are not that important most of time.

Wow, thank you! That's a brilliant explanation, which perfectly explains what I'm experiencing!

Wouldn't some tin foil between the two devices (RF shielding) have the same effect of having the devices a few meters apart?
It costs nothing to test, and if it works one could experiment with a very fine metal mesh (mosquito mesh) that could be simply glued/fixed to one side of the hAP Lite.

That's an interesting idea! Somehow it feels it wouldn't be that simple, but I'm sure going to try it! Feels like it would make the devices rather "directional". But if it works, I'll cut out a piece of sheet metal att work, on which I can mount them. If I'm not going mkx's solution anyway (but if I do, I might need some help to set it up properly without risking any security problems).

I have a similar problem this summer-vacation-time. Lots of visiting people are bringing their own travel router (used as repeater , or used as VPN gateway for things like Netflix with their home account). It must be a hype this year!

Problem is that we place wifi in each holiday residence with a hAP ac2 indoor, and wAP ac outdoor.
Problem is the indoor hAP ac2. I get connections at -10dBm and stronger.

Users do not realise that routers/AP (even travel routers/repeaters/hotspot) have a 20dBm radiation power (or even higher as they like to fiddle with the settings).
Dutch (Netherlands) camping guides advise the travellers to set very strong power, and that with omnidirectional antenna, to connect and extent the Camping AP.
Where 30 cm distance from the hAP ac2 might be OK for a smartphone/tablet/laptop, then that is NOT far enough for that travel router/repeater/hotspot.
No problem on the distant AP on the camping, or even with the wAP on the wall. But they do set their device just next to the hAPac2

(Access list in hAP ac2 rejects those connections ... they set it even closer then .... but as internet fails they focus on the cabling of the hAP ac2 )

Sounds like exciting times ahead for you!
People are getting more and more savy, especially the young amongst us who seem to be influenced by tech channels to VPN-UP etc for the pron.
As well as people that want to VPN home due to ease of use functions like BTH!

@bpwl
That seem like one of the few cases where a cAP might be more useful than a hAP if you mount it on a ceiling or high on a wall.

@acrophobic
Sure, if you have the possibility a sheet of aluminium some (say) 2 mm thick would be light enough and have the mechanical resistance to hold the two devices.
The problem is more about whether the shield needs to be larger than the size of the devices (or needs to be bent similar to a box around the corners/edges of the hap lite).
Radio waves shielding is tricky, they may dtffuse even through the (power) cable, in a different context (faraday bags/boxes for forensics) that has proved to be an issue.

I have a similar problem this summer-vacation-time. Lots of visiting people are bringing their own travel router (used as repeater , or used as VPN gateway for things like Netflix with their home account). It must be a hype this year!

Problem is that we place wifi in each holiday residence with a hAP ac2 indoor, and wAP ac outdoor.
Problem is the indoor hAP ac2. I get connections at -10dBm and stronger.

Users do not realise that routers/AP (even travel routers/repeaters/hotspot) have a 20dBm radiation power (or even higher as they like to fiddle with the settings).
Dutch (Netherlands) camping guides advise the travellers to set very strong power, and that with omnidirectional antenna, to connect and extent the Camping AP.
Where 30 cm distance from the hAP ac2 might be OK for a smartphone/tablet/laptop, then that is NOT far enough for that travel router/repeater/hotspot.
No problem on the distant AP on the camping, or even with the wAP on the wall. But they do set their device just next to the hAPac2

(Access list in hAP ac2 rejects those connections ... they set it even closer then .... but as internet fails they focus on the cabling of the hAP ac2 )

That is some interesting insight, it doesn't sound like a nice problem. I find it a bit funny that camping guides encourage visitors to beef up the transmit power on their devices.
What I want is to connect and preferably get around any restrictions on the network. What I DON'T want is to mess up anyone's WiFi. Would you recommend to actually lower the transmit power of my router then and if so what should I set it to?
On the last line you write "Access list in hAP ac2 rejects those connections". Does this mean the visitors routers are rejected/disconnected based on their transmit power?

Sure, if you have the possibility a sheet of aluminium some (say) 2 mm thick would be light enough and have the mechanical resistance to hold the two devices.
The problem is more about whether the shield needs to be larger than the size of the devices (or needs to be bent similar to a box around the corners/edges of the hap lite).
Radio waves shielding is tricky, they may dtffuse even through the (power) cable, in a different context (faraday bags/boxes for forensics) that has proved to be an issue.

I'm leaning more and more towards just a bit of cabling and placing the hAP lite a meter or two away from the hAP ac2, WHEN I need it, even if it's not as neat. But I will try the tin foil later or tomorrow, just to see if it has any effect.

I find it a bit funny that camping guides encourage visitors to beef up the transmit power

Sorry for the confusion. It is in the community-blogs of people going to campings and holiday houses ("Gites de France" style).
Camping owners hate those EZ repeater/ EZ Hotspot Extenders (based on Engenius and other strong-radio AP's), they even block them on MAC addresses.
These (omnidirectional, high transmit power) devices destroy the wifi experience for others, and even for themselves.

My worst device interference, that even still bites me today) is with wAPac and SXT SA5 ac or SXTsq 5 ac.
It generates false radar detects in SXT and wAP.
Even with wAP on 5500MHz/Ce and SXT on 5660/Ce, I had it again after a power fail last weekend.
(Hmmm, wAP ac eventually moved to 5180/ce , only then the repeated false radar detects stop. But 5180 is not allowed outdoor in ETSI region)

My other problem is with a "Mikrotik PL7400 powerline AP" https://cdn.mikrotik.com/web-assets/pro ... 200241.pdf
Had some of these as spare devices. It is now not used as powerline extender, but just as wifi to ethernet converter (station-bridge) with wifi uplink to AP-bridge hAP ac2
(PWLN could be used as repeater in 2.4GHz also)
PWLN has a very compact form factor. It's powerline interface is compatible with the hAP Lite. Device is discontinued by MT.
Problem here is that the wifi signal is always much too strong. Even with TX-power set at 0dBm, and in this older 6.44 release with also antenna gain of 20 (reducing allowed TX power with 20dBm)
It still gives -25dBm on the hAP ac2. (Ron Touw learned us that some chipsets do not follow power settings below 2dBm)
Moving them further apart will be the only solution. Reducing hAP ac2 power is not what is needed here.

FYI only. My experimental "refuse too strong transmitters" access list.
As can be concluded from MKX excellent contribution above, that this cannot solve the problem as such.
There is also no informative feedback to the user.

(Dutch comments, sorry, yes I'm that lazy here. Text can be used in "deepl.com/translator")

@bpwl
That seem like one of the few cases where a cAP might be more useful than a hAP if you mount it on a ceiling or high on a wall.

Yes would be better placed. Even hAP ac2 to the wall and not on furniture is better placed to avoid closeby devices.
cAP ac however is not as good at penetrating thick walls, as the hAP ac2 performs well.
Shielding ... The cAP XL ac has a built in reflector, projecting signal somewhat in a halve-globe shape. viewtopic.php?t=182762&hilit=cap+XL+antenna#p909850

FYI only. My experimental "refuse too strong transmitters" access list.
As can be concluded from MKX excellent contribution above, that this cannot solve the problem as such.
There is also no informative feedback to the user.

(Dutch comments, sorry, yes I'm that lazy here. Text can be used in "deepl.com/translator")

Thank you! Well, I'm lazy too, so haven't translated the comments. But looking at the rules, I can understand the principle anyway

Wouldn't some tin foil between the two devices (RF shielding) have the same effect of having the devices a few meters apart?
It costs nothing to test, and if it works one could experiment with a very fine metal mesh (mosquito mesh) that could be simply glued/fixed to one side of the hAP Lite.

I tried with some aluminium foil (regular kitchen aluminium foil folded a few times), but it seemingly did nothing Fun try though

Yep, as said RF shielding is tricky business, strange that you did not notice some differences.

Well, at least we tried ...

An alternative (that you won't like) could be modifying the hAP lite, add an external antenna connector and use an external flat (directional) one:
viewtopic.php?t=106818

Small flat directional antenna example:
https://it.aliexpress.com/item/1005006119868330.html
(these are often used on remotes)

Some versions - it has to be seen - do sport a "test connector" that can be used (improperly) without needing to solder:
https://apps.fcc.gov/eas/GetApplication ... id=2542112
(as long as you can get your hands on a (rare) pigtail with a proper connector that seems like being a MS 156 (i.e. connector and RF switch), example:
https://it.aliexpress.com/item/1005007225091883.html

Of course, no guarantee whatsoever, only take this path if you think it will be a fun experience, besides whatever results you may get.

The interference is likely due to the close proximity of the two routers, even on non-overlapping channels. Try increasing physical distance, lowering the transmit power, or using shielding to reduce RF interference between the 2.4 GHz radios.

Thanks for the link.