Besides hAP ac^2 router, I also have wAP ac. The test results of wAP ac are as follows:
Based on the figures presented, wAP ac is much better than hAP ac^2 higher and stable throughput and with lower Retr value. The Retr count in hAP ac^2 is almost ten times higher than of wAP ac. Also the download speed is around one third of wAP ac.
Listing 1: Download wAP ac v6.41.4 (File server ->Bridge->WLAN->Laptop PC)
@yhfung, the wAP AC provides three chains under 5G so its not really fair to compare that band to the hAP ac2 — the hAP ac2 provides 2 chains for the same 5G band.
My tests for the hAP ac2 also confirm your findings for both bands … very poor performance on either band. I am hopeful that MikroTik and Qualcomm the wireless chip provider come up with a solution to this poor performance soon.
yhfung,
The main difference between wAP and hAP ac2 is CPU [arcitecture]: while wAP is based on QCA9533 (well known MIPS), hAP ac2 is based on IPQ-4018 (ARM - new to Miktorik).
Actually I am not able to respond the questions, and to be honest I still owe the requested response to mikrotik because I was not able to find a suitable time to do that. On the other side it is clear that mikrotik knows the problem and works on the solution so we can expect a progress soon. Hopefully.
The 3 or 2 chains on the ap do not make any difference when only one client with 1 or 2 chains is connected for testing purposes only.
Since the hAP ac2 provides 2 chains for 2G and 2 chains for 5G — a Qualcomm MIMO system — does anyone know if BEAMFORMING is supported**?** is Beamforming turned on be default and where would I find the settings for Beamforming**?**
The IPQ4018 SoC supports beamforming (802.11ac TxBF), but that would require a phased antenna array design, which is highly unlikely. Basically, the chip is far more advanced than this specific device can fully support.
i confirm the weak performance at hap ac and cap ac devices. too bad i didn’t read the forums yet.
as i said earlier in a different thread: the hardware specs are part of my wet dreams. they’re exactly what i waited for. i really hope this gets fixed soon.
Finally, I am able to receive a formal reply from MikroTik about the problems found in hAP ac^2, as follows. However he said that he was not able to replicate the problems that I had. Also he could get very good results of 300Mbps via the WiFi. When I have time, I shall follow his guide to see whether I can or cannot obtain the results as he mentioned.
How about yours guys? Are you ALWAYS able to get a stable and high data rate at 300Mbps over a long time of time, eg. 5 minutes?
YH
The words in red were coloured by me on purpose.
------ message received from MikroTik Support ------
Hello,
Thank you for contacting MikroTik support and investigating this!
Please try these steps:
Update router to latest RouterOS release candidate as well as update firmware in /system routerboard.
Reset configuration with no defaults:
/system reset-configuration no-defaults=yes skip-backup=yes
Set basic configuration for wireless like:
band=5ghz-a/n/ac channel-width=20/40/80mhz-Ceee mode=ap-bridge country=(your country)
Test speed with different frequencies (5180, 5260 etc.) and different channel widths.
Test speed with two different setups - one where you bridge all interfaces, other where you have a local interface with IP, DHCP-server, NAT mascarade.
Generate supout file at the moment of download test and send it over.
Describe your test setup (location of device and distance, network setup, client device etc.) and let me know about your findings.
We need some more information. We cannot reproduce this issue. Testing with MacBook Pro 2016 on distance 10m in an office environment with neighboring APs hAP ac^2 performs stable and reach up to 300 Mbps.
In the past I was an electronic design engineer, good product performance should be easily able to get by most users. How comes there are so many users having the same WiFi problems when the product was launched? Performance was improved by the Release Candidate but not high and stable enough. Also how comes the Retr count is so high as compared with other MikroTik WiFi routers?
So everyone should buy MacBook Pro 2016?
Why didn’t they perform reliable tests with more than one Wi-Fi card, especially mentioned previously:
Intel AC-7260
Intel AC-8260
?
On my 100mbps (Virgin Media UK) internet connection I get a burst of 50mpbs which goes down to 6mpbs. It’s like being on adsl 10 years ago! This is with the stock firmware.
With the new RC firmware, things are much better, but still not great, it hovers between 70mbps and 35mpbs.
With the Virgin Super Hub 3 router, i get 110mbps sustained.
On my hAP ac2 the default config has all interfaces bridged …. so how do you un-bridge OTHER where you have a local interface with IP, DHCP-server, NAT mascaraed ? [EDIT] OK just learned how to remove the bridge from each individual port — will run some more tests.
Waiting patiently for good news before making the plunge. Should I instead start a crowd fund me page to buy some other products ubiquitous with satisfied customers?
I have already sent my reply to MikroTik Support by including the test results of single frequency of bridge mode since I do not have much time to do. Based on the readings shown in the log below, it shows that the download rate is not stable and the value Retr 402 (packet retry count due to errors) is extremely high for 20-second test. For good WiFi router, this should be close to zero for just 20-second test.
My logic is like this. If the release candidate firmware is good, then the majority of us would be happy with the results by general settings made on our hAP ac^2 routers, right?
