Hello. I have bought mAP lite but its performance seems too poor.
The AP is provisioned from a RB750Gr3, 4 different SSID (4 VLANs), powered by Cisco active PoE switch.
if I move it 5 cm from the mAP lite, I get -30 dBm signal which is excellent.
if I move it 1.5 m from the mAP lite (still in the same room), I get -65 dBm signal, which is still good but quite a drop
if I move to a room separated by brick wall, keeping door open, with the mAP lite right behind the wall, signal goes down to -85 dBm with poor speed (<5Mbps) and phone connecting to LTE instead.
Looking at frequency usage, I don’t see interference as a problem but I did try channels 1, 6 and 11 without much difference. Therefore I wonder: is mAP lite performance really meant just for close range? Even 1 wall can degrade the signal so much?
I tried TPLink old Wifi router which covered the whole floor, even passing 2 brick walls with the phone still getting full internet speed (50 Mbps).
There shouldn’t be such a degradation in signal strength if you have default wireless settings. Can you export your wireless settings with hide-sensitive=yes here?
What is strange is that even with 7ms ping to the internet anywhere where I get any signal and reported 130/144Mbps client connection, I only get 10-30 Mbps real speed for download and 5-8 Mbps upload. The signal is worse on the same floor than is the floor above/below. Like if the signal has no issue passing through the floor but the wall is a problem. I added a second mAP lite and the funny thing is that clients on the 1st floor connect to the AP on the 2nd floor and vice versa, unless the client is in the same room and within 4 meters from the AP. I tried changing channels, which only decreased signal between floors but didn’t improve it within the same floor.
I know that I could buy 3-4 mAP for each room on a floor for the price of one Unifi AP which I used until now but there aren’t that many Wifi channels to cover 2 floors (6-7 mAPs) without channel overlapping. Perhaps someone can recommend a better mikrotik AP (2.4Ghz, 802.3af PoE) on the same performance level like my old Unifi AP AC LR?
I just found out what’s the thing the mAP lite doesn’t like the most. It’s me! Here’s what happens: my iPhone SE 2020 gets -92 dBm signal, very poor, considering I’m just one wall (with door open) from the mAP lite. However, as soon as I put down the phone, the signal grows to -72 dBm! This also translates to speed - holding the phone, I get 20 Mbps max, mostly nothing and phone switching to LTE. With the phone in the exactly same spot but not holding it and moving 1m away, it gets 50 Mbps download.
So it seems the iPhone SE wifi is horrible but I still didn’t have such issues with TP Link or Unifi. It just became more of a problem with mAP lite…
Wifi radiation is a complex thing. There are simulation programs, but I know none with a persons body modelled.
There was for the mAP a RBmAP2n (old/single chain) before the RBmAP2nD (dual chain), is this mAP Lite a RBmAPL-2nD, with dual chains?
Wifi 2.4 GHz is “microwave” frequency. All what heats up in a microwave oven will absorb signal. (water, grease, metal, …) will absorb/reflect the signal.
The little mAP 2nD Lite in your hand can be completely enclosed by absorbing material.
Did you have a look at the CCQ in the registration table? It is a good indication of signal distortion by wall’s or of occurring destructive (adjacent channel) interference.
Co-channel interference cannot be seen on MT while using the channel, AFAIK.
I don’t see CCQ in the registration table. Just on the AP in the wireless interface status. And it’s between 20-30% (with client on the same floor, distance from AP = 7 meters, room across the hall, door open).
Overall CCQ is not specific enough. The per connection CCQ is in the registration table (select the CCQ column to be displayed). With CAPsMAN it could be in that registration table, not local.
20-30% is rather poor. Every packet has to be sent 5 times on average. Consumes a lot of airtime, and I assume with this CCQ level the interface rate is very low (expecting close to 6Mbps, with 5 repeats, and 50% overhead … this will get very slow) and that again takes more airtime. ALL on channel 6 (2437) will be slow if you start data-transfers because of all the airtime consumed.
Surprised to see 28dBm as transmission power. This is way stronger than the client device. Can be done indeed in the United States, but this AP is not strong enough to deliver that. Previous TX power was 20dBm, there was also “b” speed enabled, down to 1Mbps. Well actually I don’t understand “mode 20gn” with “basic rate CCK 1-11”, but OK the config says b/g/n what explains the “b” rates. But why has it a status of single stream (1x HT:0-7) ? The mAP Lite is a “nD” device, 801.11n dual stream. Well, I’m not using CAPsMAN so just odd to me to see those numbers.
On your power drops … RF power drops 6dB for every time you double the distance. 6 dB drop is 1/4 of the power as the surface of the globe around the AP is 4 times bigger.
If you have -65dBm at 150cm distance , then 6 times halving that distance gives an increase of 36dB in free field, or -30dBm at 2.3 cm. Given the fact that the RF signal around the AP within 30 cm is far from constant, and the fact that -30dBm probably saturates your receiver, and as you have indoor reflections, the number you see are far from special.
5 cm is not a safe distance to stay near the AP.
What happens through the wall is another story. We see RF signal as light, but actually the transmission is quite different. Losing 20 dB is normal, a drop of 20% in the CCQ is the painfull part for the wall passing.
I’d just like to post an update. I still have issues with mAP performance but I think it might be a temperature problem. The surface is qiute hot. I didn’t pay attention to this at first but recently I mounted the unit using the supplied metal plate but after about a day, the adhesive tape started to melt and the metal plate shifted on the surface. It’s holding pretty firmly when I remove the mAP so the glue is fine, it’s just the mAP is so hot, it melts it. So even Mikrotik didn’t expect it to be so hot.
Should I try returning these and hopefully get my money back? What’s Mikrotik policy on this?
By the way I now got an hAP lite and it works so much better. A single unit cobers both floors.
after snapping the mAP to the supplied metal plate and checking again in couple of minutes, the temperature is even higher. Seems enough to start melting the glue. Is this really OK and expected?
Measured temperature in Celsius!
