Hi All,
I'm a proud owner of Mikrotik stuff (RB5009 + CRS309 at the moment) and would like to implement my network to try solving some issues I have with other brand equipments.
I'll attach my house scheme at the end so it would be easy to understand what I need.

JUMP THE BOLD PART IF YOU DON'T CARE ABOUT MY WIFI ISSUES STORY
The "other brand equipment" are actually a unmanaged switch that connect the eth cables going in every room which it's connected to the router and a set of 3 TP-Link Deco M4 access points (all connected with eth cables).
The source of all my problems are the access points. Ok, they provide a wide range and stable network and more or less the mesh (I know it's a wrong commercial name) works smoothly...the problem is that I have no control over their behaviour. If I reboot the router in those 2minutes they decide to act as DHCP and start assigning IPs all over the network...not to mention loop packets send over the networks...this are the main issues, there are some others, but this is not the place where to talk about.

So, I decide it would be better to change those stupid APs with something else I could have control over, and because I like Mikrotik it would be nice to use their products.
Actually I have about 20/30 wifi devices (most of them are smart plugs or CloudCams, so they don't need wide band but simply a stable and strong 2.4GHz network).

Oh, I forgot to say that if I turn off all the APs except the one in the middle of the house I have no issue in about 90/95% of the house but I don't reach very good all the devices in the boundaries.

As far as I understood, I should buy ax devices to take advantage of Wave2 so first of all I identified possible candidates in the hAP ax2 and the cAP ax. (I looked mainly at the Antenna gain dBi declared and the prices)
Now, if you look at the house map, I placed a green "O" where there is an access point and a red "X" where I have critical wifi devices. They need absolutely the stablest signal as possible. The one that actually have the worst signal, but still enough to work without issues is the one on the right balcony. It usually connects to one of the two APs that are approximately equidistant (dotted arrows).

My first thought was to simply substitute the three APs with 3 others but, given that I don't produce money , I would like to understand if I can save something reducing the number of APs without losing wifi coverage.

Somy first three questions are:
1) do I really need Wave2 devices to have a working and smooth roaming like I have now with Decos? I read something about how to configure roaming in routerOS and understand that having Wave2 is better. Am I wrong?
2) How does the signal come out from the cAPs? I know they're ment to be installed on the ceiling, so I imagine they spread the signal 180°, but mostly underneath, like if they'd be directional antennas. Am I wrong?
I was wondering if I could place two cAP vertically in opposit directions (left-right), in the middle of the house (where I have the router now) and obtain a strong omnidirectional antenna.
3) In your opinion will I have at the end a better or worst wifi network?

Please, if you heard something by a friend of a friend of a friend of yours about the products I'm talking about but you have never used one of them, don't clog up this topic (sorry if this was too rude).
Thank you

Looks like a pretty small area to cover. Depending on you bandwidth needs and expectations this could be covered by a single AP as well. Depends on the walls (brick, drywall,...).

How is the signal from the middle X, to the O to the right of it/ (If you turn middle O off)
I would start with one CAPAX in the current middle position and see if the service is adequate at all locations.
If not, then consider moving the central one to the far right hand position and get a second capax for the left hand position.
(assumes you can wire to all three positions)

You will need capsman to make the most out of the roaming experience if using two devices. If only using one no need to use capsman.

Well, actually I can get 200 to 400 Mbit all around the house (probably the limitation is the device used to do the speedtest. Wired devices goes up to 750. I have 1000\300 fiber that goes 750\275 effective).
My minimum needs are 20\30Mbit per devices (3 or 4 at the same times) but sometimes I want to access big files on the NAS, so the 200 to 400 Mbit I get now are ok (but I usually do this in areas well covered).

The walls are made of bricks, the perimetral wall are double the width and I have metal grill on the windows plus anti-mosquitos metallic meshes. So the last mile to get WiFi outside is the hardest.
Not to mention I have many others WiFi around to interfere mine.

Leaving on only the middle AP, the right balcony'a cloud cam recording start to be discontinuous.

All the three APs are already wired.

Does the cAP ax has good "radial" performances? I saw some picture of the interior and seems it has four 45° angled antennas. I can't imagine how the signal propagates from it

Then I would start with two capaXs on the outer positions and see if that satisfies the requirements.

How do you suggest to position them? on the ceiling? vertically on a wall or maybe on a shelf facing upward?

If there is possibility mount them on the ceiling.

It is not clear at all (at least to me) which "shape" is the coverage of a cap Ax, I have found an old post stating that the older cap (I think it was about the cap XL) that has the similar 4 small antennas had an almost spherical coverage (as opposed to the doughnut like shape of a common omni antenna).
If this is the case there shouldn't be that much of a difference between ceiling vs. high on the wall, it all depends on how much work it is to extend the ethernet cable to get to the devices, but in any case signal on the back side of the device (besides there being a wall or ceiling) should be by design 0 or next to 0.
So, for the "left side" cap AX, the 0 signal on the left won't be a problem, for the right side it would mean that the X spot in the small balcony will remain as is or get worse.
For the (hypothetical) middle one the ceiling mount is instead needed.
But (no idea how well or bad your current AP's perform) it seems to me that two (or three) cap AX's are overkill for the size of the house and the placement as in the drawing is "unfortunate".
According to the Mikrotik (as always rather scarce) documentation the cap AX should be in the center of a room (without intermediate walls) and then it will cover a very large area.
The corresponding placement on your floor plan is the "middle" one and it should be capable - notwithstanding the walls (that seem, at least on the plan, to be internal, thin ones) of covering all the apartment.
About the small balcony on the right, if all that is needed is a connection for the IP camera, I would think of experimenting with *any* el-cheapo wireless AP/repeaters, the kind that are only a little bigger than a power adapter, plugged in a socket near the area and "dedicated" to the camera only.
Then, I would try with a single cap AX (on the ceiling in the middle position) or if the ceiling is a no-no for any reason, maybe with a little more money an Audience would be a more suitable device.
What remains not well covered is possibly the large "front" balcony.
That might be the real issue, in theory a "directional" AP (with an angle of 90° or 120°) placed on the left side spot would be ideal, but it has to be seen which Mikrotik device would be suitable (and there is anyway WAF possibly involved), otherwise a second cap AX mounted on the wall might do, while still (IMHO) sub-optimal.
In the right side spot, should coverage with the two above AP's not be sufficient, I would rather experiment with a hap Ax2 (again esthetics and mounting possibilities are relevant here) that might even reach the small right side balcony.

In a nutshell, I would in first instance buy only one cap AX, and test which coverage it has in the three spots, and then proceed based on the results, but before that, I would think whether an Audience would be more suitable (to be placed in the center spot), even if it a bit more expensive it should be easier to place in a non-ceiling position.

It is not clear at all (at least to me) which "shape" is the coverage of a cap Ax, I have found an old post stating that the older cap (I think it was about the cap XL) that has the similar 4 small antennas had an almost spherical coverage (as opposed to the doughnut like shape of a common omni antenna).

I assume you meant "semi-spherical", and anyway I expected this kind of "shape".
What I would like to know is if the signal has the same power on all the equidistant points of the vertical section (placing the cAP on a horizontal plane) of the semi-sphere, or if being right in front or on the side of the cAP make any relevant difference.
I'm asking this because while I could place the cAP on the ceiling at about 2,9m height, the cam on the right balcony is at 2,3m and I'm afraid it wouldn't be reached by the same signal would receive if it was at 1,5m.
This is the reason why I was supposing to place two cAP vertically "on a pole" in the middle room (it's kind an office, I already have a 12U rack in it and I could put them on top of it withouth being uglier than now ) to have the real spherical effect of an omnidirectional antenna.

About the Audience, looking at the bare specs, it seems to me it has half the antenna gain in 2,4GHz, even a bit less than what hAP ax2 has, and it should have better performance only in 5GHz. Then, if I understand correctly, it has roaming dedicated antennas to communicate with other Audience, but that could be something I don't need.

I think the option of a tiny repeater (what about mAP lite?) near to the little balcony isn't stupid at all! Because to be honest I have an ethernet plug on the opposite side of the same wall of the balcony and it may solve brilliantly that issue.

At this poin I think the best scalable test it would be, as some of you said, to take one cAP ax and see how it perform with different mounting solutions to be able to make correct hypotesis on how (and if) to implement

EDIT: my camera app tell me the RSSI value of the cameras, so I can tell the signal strenght of the cam on the right balcony when it connects to the right or the middle AP and I see pretty near values, just a little better from the right AP (-62 vs -72)

No, I meant spherical, this is the thread, I remembered, but that for *some reasons* I couldn't find earlier
viewtopic.php?t=183719
but it doesn't make sense, as in the back of the devices the signal will be 0 or much lower, as there are (AFAICU) reflectors behind the antennas, so semi-spherical seems more plausible.
Still, your hypothesis of a "mounted high" device not receiving much signal from a ceiling mounted device seems to me correct, I would expect the angle (when viewing a cross section with the AP mounted on the ceiling) to be not 180°, but more like 150° or so, in practice cutting out anything at the same height as (or slightly less than) the AP..

The map lite, is a nice, cute, little device, and actually very cheap for what it offers (it is the form-factor that it seems to me not-so useful in your use case, as it has the separate power supply, needs a support/mounting bracket, etc.) , but I was thinking more of a (I will probably be flamed for this) el-cheaper device, like (example only):
http://www.pix-link.com/page88?product_id=144
I have two of them at home (the WR09 models I have are all white, and apart the little blinking lights they are almost invisible against the white plug and wall), if I remember paid them like 15 or 16 Euro each, and though they are of course not very fast, if their usage is limiited to the camera it will be enough, and they just work.
There are similar more powerful/faster devices:
http://www.pix-link.com/Product?product_category=24
and of course many similar devices, from better known brands.
The idea is to have a "dedicated", minimal AP only for the camera, you don't really-really need it to be Mikrotik or to "integrate" it in the rest of the network, it would be more like a PtP (Point toPoint) bridge.

Anyway, the map lite is designed for portability, for a fixed installation I would rather prefer the only slightly larger map.
Both can be powered at 5V, so it would be possible to replace the wall plug with one integrating a USB power supply, get a short USB white cable and have the whole thing much less visible.

P.S. If the middle room is an office / has no particular esthetic limitations, than an Ax3 could also be considered, it is a much more "flexible device" that can be easily reusable in some other setups, though I have seen on the forum contrasting opinions on its real world wi-fi capabilities, it seems to me better value for the money than a cap Ax.

No, I meant spherical, this is the thread, I remembered, but that for *some reasons* I couldn't find earlier
viewtopic.php?t=183719
but it doesn't make sense, as in the back if the devices the signal will be 0 or much lower, as there are (AFAICU) reflectors behind the antennas.

exactly, I have the same idea. it would have been better a verbose explanation

The map lite, is a nice, cute, little device, and actually very cheap for what it offers (it is the form-factor that it seems to me not-so useful in your use case, as it has the separate power supply, needs a support/mounting bracket, etc.) , but I was thinking more of a (I will probably be flamed for this) el-cheaper device, like ....

....The idea is to have a "dedicated", minimal AP only for the camera, you don't really-really need it to be Mikrotik or to "integrate" it in the rest of the network, it would be more like a PtP (Point toPoint) bridge.

Anyway, the map lite is designed for portability, for a fixed installation I would rather prefer the only slightly larger map.
Both can be powered at 5V, so it would be possible to replace the wall plug with one integrating a USB power supply, get a short USB white cable and have the whole thing much less visible.

I have both ethernet and power where I would put the mAP lite and it wouldn't be visible, so whatever the installation method would be, it will be ok I just don't want (anymore, bad experiences) unmanaged devices able to do things they weren't asked for

Anyway...the strange signal pattern that cAP ax seems to be able to offer changes a bit my understanding of how would go my tests, but I'll give it a try. I didn't find decent specs for my actual access points so I haven't either anything to compare. It could end up in "just one cAP ax is much more powerfull than three TP-Link APs" with the same probability as "three cAP ax perform less than a third than the previous APs"

P.S. If the middle room is an office / has no particular esthetic limitations, than an Ax3 could also be considered, it is a much more "flexible device" that can be easily reusable in some other setups, though I have seen on the forum contrasting opinions on its real world wi-fi capabilities, it seems to me better value for the money than a cap Ax.

well, in first place I considered the ax3 too, but looking at the specs it seems they just added more juice to the 5GHz steling it to the 2,4GHz. I have the impression that because of my needs, it would cost more for having less

Ok!
I took one cAP ax and configured it last night, so I have still nothing particular to share with you about the range etc...except the fact that placed near the old AP (they're still both powered for now) and standing near them, a simple speed test from my phone (in 5GHz) shows bit better results, but as I said before I'm more interested in the widest range possible...so 2.4GHz

Now I have a question for the signal-nerds reading: is there something I can configure in the channel settings to squeeze every additional meter from the AP? I did a search and didn't find nothing that goes deep in the frequency and bands configuration (the official documentations is very concise about aspects that in my opinion would need pages (or hours of video) to explain.

Well, after a couple of days I understood something:
the cAPax placed in the middle room it's enough but also not
5GHz is ok everywhere in the house (mainly thanks to the fact there are only 3 or 4 other AP on that frequencies, so there's room to find a free slot), near the AP there's almost no difference with wired devices, in the furthest place i still download at 200/300Mbit.
It's not the same on 2.4GHz, it's a bit crowded here, so the signal for the critical devices stays between 60 and 65 dBm but it's too much subject to the environment. Closing a couple doors or shutting down the roller shutter raise the signal to about 70 or bit more and this seems to be the limit under which it becomes unstable.
And this is what I obtained changing the country from Italy to Panama, which seems to be the one with the strongest Tx power. In fact with wifi analyzer on a tablet placed in the same room of the AP there's a 10dBm difference between the two countries on the same channel. And in fact using Italy the furthest devices have really instable network.
Now I'm going to try placing the AP in different rooms or different position in the same room to see what happend so maybe I could go with two APs in the two remote rooms.

do you have any suggestion on how to configure the AP?

For the experiments, I would leave settings as they are, to not introduce variations, and fine-tune only once you have a somehow satisfactory setup.
The 5 GHz works fine, so leave it as is for now.
For 2.4 GHz it all depends on which channels are busy/have noise, the generic recommendation of using channels 1, 6 and 11 may be not ideal because of the interferences, at the end of the day the only way to know is trying the various channels, see also:
https://superuser.com/questions/443178/ ... 786#776786

About Tx power, it is only part of the story.

WiFi is a bad beast.

By raising Tx power you may be able to "reach" user devices (that might have poor reception) but you won't enhance the reception from the same devices (that as well might have lower transmission capabilities) for this latter you would need a higher gain antenna (in reception, only on those devices that allow replacing the default ones) that however would mean to lower Tx power.

Since you seem to be based in Italy, [italian] è un cane che si morde la coda[/italian].

Ahahah sembra una barzelletta, cosa ci fanno due italiani in Lettonia?
Ok, let's continue in English so someone else will understand too

I can't change clients antennas so the only possible solutions are: tuning settings, moving the AP or using more than one of them (after testing different positions of course).
About the channel, the AP already chooses the better possible. Yesterday was channel 5, in this moment moved on 4, and probably I'd have done the same, cause the stronger other WiFis I see on WiFi analyzer are on the higher channels, while between 1 and 9 there are only things under 80dBm. I didn't checked yet, but I suppose that where I need coverage (the little balcony on the right) there wouldn't be more interferences than in the middle room because everyone here have their modems on the opposite side of the apartment

I can't change clients antennas so the only possible solutions are: tuning settings, moving the AP or using more than one of them (after testing different positions of course).

Yes, but I was referring to the (Mikrotik) AP's.
Of course you cannot do anything on a CAP or on another similar device with fixed antenna(s), but there is an ongoing debate on the forum for the Ax3 (which could be a suitable device for your "middle" position).
In a nutshell about the Ax3 (which has replaceable and orientable antennas) there are two "parties", one of which experienced same or worse wi-fi coverage (particularly in 5 Ghz, I don't remember seeing posts about poor 2.4 GHz) than other similar and in theory "inferior" devices (by Mikrotik and other manufacturers) and the second that experienced a very good results, and there are ongoing experiments to try and understand why this happens, the relationship between specific country settings and (higher and lower channels), antenna gain, antenna orientation and what not.
Nothing definitive has yet come out of those threads, though.

As I said earlier, for roughly the same money the Ax3 seems to me to offer more than a Cap Ax in performance and flexibility (it is essentially a good router besides and before the Wi-Fi and as a matter of fact a few expert members use one of them as "pure router") if there are not esthetics considerations, and it might allow some "fine tuning" or "experiments" (like using more directional antennas or higher/lower gain ones) that it is simply not possible on the CapAx.

If you don't mind me drawing random circles over your house , if the Cap AX in the left and middle position can cover the area in the bottom drawing (circles of roughly 7.5 m radius) another Cap Ax on the left will cover the rest or viceversa.
If, on the other hand, it is more like the top drawing (circles of roughly 5 m radius) and you will need 3 devices, personally I would buy and try an Ax3 in the middle position, as, if it can deliver *something more*, you may be able to avoid buying the second Cap Ax for the right side (and if needed use an el-cheapo AP for PtP of only the outdoor camera, you could have also the option (cannot say if possible) to "move" the antenna(s) in other positions using antenna extensions (again if the esthetics in the middle room are not that important) .
If it doesn't, the right side Cap Ax would cover anyway the right side and again you may be able to play with positioning the Ax3 in the middle in various positions and/or using extension cables and other antennas to attempt to make its coverage less circular and better cover the large balcony at the front.

I'm from Italy too (I say that since country settings do matter) and I own an AX3.
First, I must say that I discovered that some IoT devices do not work well when I tried to play with country settings: greater tx power, but inconsistent behaviour. Maybe it was the asymmetricity (the client was able to hear the AP shouting, but it was not able to shut itself back to the AP), maybe it was some kind of country/area lock. I was not able to debug since, as usual, this kind of devices does not provide a lot of info (and, unfortunately, also the ROS wifi debug log is not very informative). However, I switched back to Italy and playing with channels helped a lot (much more than changing country). In my case, 5500 is a very sweet spot.

As an AX3's owner I can say that, in my case, the router provided me better coverage than the Vodafone Station Revolution (VSR), both in 2.4GHz and (especially) in 5GHz. It is closed in a cabinet, but it covers a greater area than the VSR and also the (nominal) rates are better than before. I still need a 2nd AP (I chose an hap lite ax since I wanted to replace also an unmanaged switch) for covering the second floor.
Your mileage may vary, obviously.

I can't change clients antennas so the only possible solutions are: tuning settings, moving the AP or using more than one of them (after testing different positions of course).

Yes, but I was referring to the (Mikrotik) AP's.
Of course you cannot do anything on a CAP or on another similar device with fixed antenna(s), but there is an ongoing debate on the forum for the Ax3 (which could be a suitable device for your "middle" position).
In a nutshell about the Ax3 (which has replaceable and orientable antennas) there are two "parties", one of which experienced same or worse wi-fi coverage (particularly in 5 Ghz, I don't remember seeing posts about poor 2.4 GHz) than other similar and in theory "inferior" devices (by Mikrotik and other manufacturers) and the second that experienced a very good results, and there are ongoing experiments to try and understand why this happens, the relationship between specific country settings and (higher and lower channels), antenna gain, antenna orientation and what not.
Nothing definitive has yet come out of those threads, though.

As I said earlier, for roughly the same money the Ax3 seems to me to offer more than a Cap Ax in performance and flexibility (it is essentially a good router besides and before the Wi-Fi and as a matter of fact a few expert members use one of them as "pure router") if there are not esthetics considerations, and it might allow some "fine tuning" or "experiments" (like using more directional antennas or higher/lower gain ones) that it is simply not possible on the CapAx.

If you don't mind me drawing random circles over your house , if the Cap AX in the left and middle position can cover the area in the bottom drawing (circles of roughly 7.5 m radius) another Cap Ax on the left will cover the rest or viceversa.
If, on the other hand, it is more like the top drawing (circles of roughly 5 m radius) and you will need 3 devices, personally I would buy and try an Ax3 in the middle position, as, if it can deliver *something more*, you may be able to avoid buying the second Cap Ax for the right side (and if needed use an el-cheapo AP for PtP of only the outdoor camera, you could have also the option (cannot say if possible) to "move" the antenna(s) in other positions using antenna extensions (again if the esthetics in the middle room are not that important) .
If it doesn't, the right side Cap Ax would cover anyway the right side and again you may be able to play with positioning the Ax3 in the middle in various positions and/or using extension cables and other antennas to attempt to make its coverage less circular and better cover the large balcony at the front.

I considered the ax3 and decided for a no, because of the lower antenna gain declared in the specs for the 2.4GHz. Reading the forum I understood (and maybe it's just a matter of what threads I landed on) that it could have performed under the expectation (I'm always talking about 2.4GHz only)...but as you said maybe it's the result of the two "parties" but now I know that is not still clear how it really perform.
About experiments with antennas I'm a bit sceptical because it could become a long term and costy experiment. for just 20€ difference and the possibility to spent hunderds in antennas, considered I don't need another "good router too", it's again a "no".

for now I'll try different positionings and give you feedbacks

I'm from Italy too (I say that since country settings do matter) and I own an AX3.
First, I must say that I discovered that some IoT devices do not work well when I tried to play with country settings: greater tx power, but inconsistent behaviour. Maybe it was the asymmetricity (the client was able to hear the AP shouting, but it was not able to shut itself back to the AP), maybe it was some kind of country/area lock. I was not able to debug since, as usual, this kind of devices does not provide a lot of info (and, unfortunately, also the ROS wifi debug log is not very informative). However, I switched back to Italy and playing with channels helped a lot (much more than changing country). In my case, 5500 is a very sweet spot.

As an AX3's owner I can say that, in my case, the router provided me better coverage than the Vodafone Station Revolution (VSR), both in 2.4GHz and (especially) in 5GHz. It is closed in a cabinet, but it covers a greater area than the VSR and also the (nominal) rates are better than before. I still need a 2nd AP (I chose an hap lite ax since I wanted to replace also an unmanaged switch) for covering the second floor.
Your mileage may vary, obviously.

Ciao Paesà!
Luckily I have no issues with my devices after setting another country, what was unstable became a bit more stable and nothing more...so at the moment seems to be a win. At the moment I really don't care about the 5GHz side...because the environment is so free it has never been a problem, always good performances in every area of the apartment, that's another reason why I wasn't particularly interested in the ax3 against the cAPax