I’ve been running hAP ac2 for a few years now. Lately I’m experiencing issues with Wifi where I only get about 10-15Mbps of internet speed out of 2.4 or 5GHz wifi connection. I have 1Gbps fiber to my home, both up and down 1Gbps. I’ve checked with wired connection through my hAP ac2 and I can get the full speed that way.
I’ve since bought the hAP ax3. I’ve just turned it on with stock configuration (i.e. no changes on my part) and connected it to my hAP ac2 via ethernet. When I connect to the ax3’s any one of the two Wifis, I get insanely fast 80-90Mbps internet speeds. I’ve checked this with my 3 laptops, iPhone and a mini PC. So it looks like the hAP ac2’s wifi speeds are significantly slower.
I’m thinking that I need to replace the ac2 with ax3 now. But just out of curiosity I was wondering if the wifi on the ac2 might be misconfigured somehow to limit the speeds. Is it worth to investigate this? If yes, what do I need to check on the ac2 and adjust to have at least some decent speeds out of wifi?
NOTE: It is not simple for me to swap the ac2 with ax3. I need to recreate all the VLANs and network segments, which will take some time to implement and test. So I’d prefer to rectify the ac2 first before making the switch. I also intend to donate the ac2 to my in-laws and would like to to not suck this bad.
The hAP ac2 can route close to 1 Gbps with FastTrack. Over Wi-Fi, my units get around 400-440 Mbps from a few feet - throughput falls apart with distance and a few walls. The ax3 should get at least 50% more and have far better range. Something isn’t right with your results…
440Mbps is very very good for WLAN driver in hAP ac2.
But further away, 180Mbps is very normal. But then again the rate will lower with distance, signal quality, interference etc.
15Mbps only, something is very bad around there, or the wireless is set up for 20MHz channel width, with a single stream client, lots of connections or interference with others, or poor CCQ.
Average setup (40MHz channel width, 2S , 433Mbps interface rate) should give 180Mbps data rate for hAP ac2. http://forum.mikrotik.com/t/wifiwave2-expected-performance/160075/1
It’s down from my ISPs speed test portal - so WAN and on a recent phone. My older Apple devices get 350-380 Mbps. It doesn’t get better with these on LAN.
Settings are: 20/40/80MHz Ceee + 5180 + auto secondary ch + any installation with the MT drivers (2GHz limited at 20MHz). I get more consistent speed with the RB4011, but around the same maximum.
The Wave2 drivers gave me around 430 Mbps max when I tested 7.11.2 on the 4011.
MT doesn’t have any devices that can beat top tier consumer routers over Wi-Fi (getting over 1.6 Gbps). But you get stability and versatility.
Edit: I think it’s important to mention that I do not have any interference at my house - so I can use channel bonding to achieve better throughput. There’s a scanner in WinBox, but many apps are available to survey channel/frequency use and signal strength. Channel bonding can actually cause worse performance because it adds noise and leave you with fewer channel frequencies to pick from in a busy environment. Channel widths of 20 or 40 are much safer bets to troubleshoot. Typing /interface wireless info country-info <> will tell you which frequencies are available. Lower frequencies usually work better, but they are often the most used in crowded spaces. Auto never worked for me with the legacy drivers and many updates made my performance worse. Fixing this might take time. It this regard, the wave2 package is a lot easier to work with, like bpwl mentioned, but it can’t do miracles if there’s strong interference everywhere.
Ok so how do I begin to narrow down the issue? I do not want to end up the same with the ax3.
Normally you will not end up the same … WLAN driver in Mikrotik is like a manual gearbox in a car, wifiwave2 and other brands are like automatic gearboxes.
Optimal setting in most cases, Less control, but less needed (except for difficult cases like steep downhill)
Start in WLAN driver with Mikrotik default setting! Don’t apply tuning suggestions you find on Internet or in this forum, unless they are verified to be applicable.
Most tunings give a slight improvement if applicable, in general they destroy performance if just applied without verification if fit for that case.
Applicable in many cases:
avoid 802.11b in every device. (Any device with 802.11b active may force other devices to switch to 802.11b)
keep data rates on default. do not drop lower rates except for the 802.11b rates
enable WMM support
Disasters in many cases:
disabled slow data rates
AMPDU disabled. Or priority changed without setting the extra AMPDU enabled accordingly
any changes in HW retries, disconnect timeout,
environment with WMM priority used by other AP (they have priority above MT default priority 0). Typical case using video streaming, or fake video streaming DSCP priority
other AP closeby (they are 3dB stronger with their TPC EIRP power levels !)
Study the wifi latency, and learn how the small MSDU, MPDU and A-MPDU in Mikrotik gives low performance when other transmitters (AP or client) compete for air-time.
I played with my RB4011 5 GHz (WLAN drivers) and an old iPhone 11 to show how much little changes can make a huge difference without any bad tweaks that can make performance much worse:
From 2 feet - avg 3 runs
Default values → no channel error - radio disabled
20 MHz + 5180 (best free available) + indoor = 101 Mbps
20 MHz + 5180 + any = 118 Mbps
20/40 XX + 5180 + any = 235 Mbps
20/40/80 XXXX + 5180 + any = 390 Mbps
20/40/80 XXXX + 5180 + secondary auto + any = 445 Mbps
20/40/80 Ceee + 5180 + any = 315 Mbps
20/40/80 Ceee + 5180 + secondary auto + any = 270 Mbps (the settings that work best on my Wi-Fi 6e devices is worse here)
The frequencies you can use will depend on your region and must take into account usage from other devices in proximity.
Except for the “auto” setting, which is undefined, you should check what you get, as different frequencies will give different power and throughput.
MT ROS is not AI or Smart. It does not test for the “best” value in “auto”.
If you want to get somewhat better, don’t forget to set the priority (and AMPDU !!!) . http://forum.mikrotik.com/t/help-mikrotik-capsman-wireless-download-speed-max-200mb-but-pcs-mobiles-link-speed-is-866mbps/142677/1
Other brands use DSCP by default, but in MT/WLAN you have to enable it with a mangle rule or set it with VLAN priority
The wait time between transmissions is somewhat shorter with higher priority. Therefore this WLAN will get the right to transmit more often, if there is any competition. And overhead is a bit smaller.
Who else is faking “video” to get WMM priority in a crowded environment?
I was actually surprised by my results because I had done my homework and tested multiple clients before settling on my settings. However, it’s been my experience that even updating firmware can have a major negative impact on performance with various clients for no apparent reason. There are lots of features under the hood, so it’s not always easy getting the most out of these drivers.
I tried changing the priorities in the past, and it only lowered my bandwidth - probably my fault. Regardless, I doubt that I can squeeze any more out off my AC units at this point - so many users were reporting their rate at 1/3 vs other brands. I was getting 60-120 Mbps when the hAP ac2 was launched. Furthermore, very little information is provided by MT about these settings, and if it wasn’t for your posts, they would be even more obscure.