I am experiencing a lot of problems with wireless devices connected to the RB751-2HnD.
I bought 3 units, in 2 different periods of time (2 when they became available several months ago, and one which arrived today). Local distributor said the last one is a new revision.
They are all configured as an access-point (802.11 b/g/n), and I get a lot of disconnections and very poor performances (high latency, variable between 40 and 3000ms, 100kbit/s throughput). If I manually set noise-floor to -90dB it becomes usable, but still I have reliability problems.
There are no interferences on the channel and devices work very well when connected to other access-points.
I already tried different ROS versions (5.6, 5.7, 5.9 and 5.14). All show the same behaviour on all units.
have you contacted the support@mikrotik.com about the configuration of your wireless interface?
Do you use indoor built-in antennas or you connect external antenna connected to the external connector?
What wireless devices you use to test that performance? How many clients you have on that AP? What signal levels they have?
You have to clearify better what your network setup is.
You talk about these 3 devices setup as access points and that there is no interference. Which is possible..
But then you talk about the fact that they connect well to other access points. How is that?
That is a. either not possible, or b. you work in a mesh (WDS) system.
More than 3 AP’s in 2,4Ghz band means interference since there are only 3 separated 20Mhz channels available in 802.11b/g
If the radio’s are even working in 802.11n than you already have interference with 2 units at the same time.
These unist are all High Power units, and if they are all in one room and working in ‘n’ mode it is inevatible you have interferences.
I didn’t contact support, yet, but I can do it tomorrow sending all details. I have some knowledge on wireless devices, and these aren’t my first mikrotik AP I am using. The configuration is simple: bridge between ethernet and wireless, no routing, no nat… it’s just an access point, configured as 802.11b/g/n with WPA2 PSK (AES CCM). I am using built-in antennas. I made tests using a Nokia N900, an ASUS Centrino laptop, an LG android phone. I tried also (today afternoon) with two other laptops (a Dell and an Acer). All the clients show the same behaviour.
Having only one client connected or multiple ones don’t change the behaviour.
Signal levels ranging from -61 to -75 depending on position and distance, the result is the same.
WirelessRudy: These 3 devices were used one at a time, never together, but all three show the same behaviour. If I replace the RB751 with another access-point, (I tried with DLINK and NETGEAR, 802.11n), all clients work as expected.
I have about the same results with the 751’s. Sometimes you can make them work by disabling chain 1. The 751’s would be great if they worked, but we have about a 75-80% deaf rate out of the box. RMAing the device and we still have the same % of them being deaf. Stopped using them until Mikrotik can address the issue.
I confirm there are several issues with RB751 wireless.
My setup is the following:
RB751 with ROS 5.14 configured with B+G+N, italian regulatory domain (20 dBm), very clean 2.4 GHz spectrum
2 clients within 1 m range:
A Dell D830 with an Intel 4965AGN running Linux
A Nokia N900
The Dell client, thought using high-speed modulations has unstable throughput rarely going over 30 Mbit/s to an ethernet-connected host (either single TCP or concurrent TCP sessions).
Playing with several radio parameters didn’t show any noticeable improvement however it is somewhat usable.
Pinging it from the host has high packet loss and latency:
64 bytes from 10.0.6.26: icmp_req=40 ttl=64 time=928 ms
64 bytes from 10.0.6.26: icmp_req=41 ttl=64 time=3.22 ms
64 bytes from 10.0.6.26: icmp_req=42 ttl=64 time=255 ms
64 bytes from 10.0.6.26: icmp_req=43 ttl=64 time=488 ms
64 bytes from 10.0.6.26: icmp_req=44 ttl=64 time=699 ms
64 bytes from 10.0.6.26: icmp_req=45 ttl=64 time=978 ms
64 bytes from 10.0.6.26: icmp_req=46 ttl=64 time=2.86 ms
64 bytes from 10.0.6.26: icmp_req=47 ttl=64 time=11332 ms
64 bytes from 10.0.6.26: icmp_req=48 ttl=64 time=10333 ms
64 bytes from 10.0.6.26: icmp_req=49 ttl=64 time=9333 ms
64 bytes from 10.0.6.26: icmp_req=50 ttl=64 time=8334 ms
64 bytes from 10.0.6.26: icmp_req=51 ttl=64 time=7335 ms
64 bytes from 10.0.6.26: icmp_req=52 ttl=64 time=6335 ms
64 bytes from 10.0.6.26: icmp_req=53 ttl=64 time=5336 ms
64 bytes from 10.0.6.26: icmp_req=54 ttl=64 time=4346 ms
64 bytes from 10.0.6.26: icmp_req=55 ttl=64 time=3346 ms
64 bytes from 10.0.6.26: icmp_req=56 ttl=64 time=2346 ms
64 bytes from 10.0.6.26: icmp_req=57 ttl=64 time=1347 ms
64 bytes from 10.0.6.26: icmp_req=58 ttl=64 time=347 ms
Sometimes it is completely mute and starts responding again as soon as it sends a packet by itself. If for example I start a ping FROM the N900 to an outside host, it somewhat works as long as the ping is running.
Disabiling N900 power saving seem to fix the issue, however it is not acceptable as the power consumption goes sky high. Power saving features do work with all other WiFi access point I used elsewhere:
64 bytes from 10.0.6.26: icmp_req=108 ttl=64 time=3.06 ms
64 bytes from 10.0.6.26: icmp_req=109 ttl=64 time=2.91 ms
64 bytes from 10.0.6.26: icmp_req=110 ttl=64 time=2.91 ms
64 bytes from 10.0.6.26: icmp_req=111 ttl=64 time=2.84 ms
64 bytes from 10.0.6.26: icmp_req=112 ttl=64 time=2.72 ms
64 bytes from 10.0.6.26: icmp_req=113 ttl=64 time=3.16 ms
64 bytes from 10.0.6.26: icmp_req=114 ttl=64 time=2.29 ms
64 bytes from 10.0.6.26: icmp_req=115 ttl=64 time=3.56 ms
im also experiencing poor performance. i have been testing all different configurations on this, this unit was meant to be a simple access point in a home environment indoors. It was put in purely to stream media from the lan to wireless clients and it has failed.
I have a bridge with ether1 and wlan1 on it. Simple way to reproduce is to begin pushing data through the wlan
e.g. ping from host sitting 1meter from router without anything running, response 1ms almost consistent. start a btest to the routerboard, begins to transfer - throughput is ~21mbps and CPU resources on the RB is 100%. after approx 1 minute the throughput drops to ~6mbps and the latency begins to fluctuate dramatically seemingly inconsistent.
I have just tested on the same frequency with the same hosts on a tplink 1043nd and the throughput was a pretty consistent ~100mbps with 1-3ms latency for 10+ minutes.
This unit was half the price of the 751u and has gig interfaces I may have to just use that instead and shelf the 751u
well the device i would never recomend for its wifi. I have bought and used several all ending the same way. Use them with external antenna 15dbi alfa with surge protector run fines for a few days then the receiving slowly becomes unstable and eventually unusable. clients still get stong singals as tx power is fine but rx is almost non existant. I even have 2 units which are stuck in a strange situation they power on and all ethernet ports lights are on with no wires attached no ports are working but wifi is working you connected to wifi and access winbox try to reset or upgrade the os. try the jumper or what ever it boots up to the same situation.
total 13 Units bought none worked as expected. I am running serveral rb433uah with r52hn cards they are running for a very very long time now without issues
Were running them as internal repeaters for our hotspotmservices with no problems whatsoever. Using both 751 and 751 g without external antenna ( and would never,ever use a 15 db omni on an internal device ).
So far at least 100+ devices without a single dead unit.
My advice: buy a unit, test it and make up your own mind.
I don’t have poor wireless performance, but have problems with connections between computers.
The setup is simple access point without any firewall or mangle rule. Just bridge (wireless and the ethernet port)
My biggest problem is to find some computer share in network.. Especially, the Mac OS clients cannot find Windows Share or Windows shared printers. Before I had Linksys WRT160NL, also configured as AP, and there wasn’t a problems in the same network - same configuration.
Maybe I should setup something else on my mikrotik but cannot find simple access point tutorial - most of them are for configuring NAT router.
Sorry, I didn’t have a RB751 at hand to do the tests.
I just had a fresh RB751G-2HnD delivered to me, I installed 5.17 and 6.0beta2, in both cases the problem with N900 power saving consistently showed up.
The N900 completely stopped responding to pings until it had to transmit bu itself, then the queue was flushed. As soon as it stops generating traffic it stopped responding.
64 bytes from n900.orlandi.com (62.212.12.15): icmp_req=236 ttl=64 time=3.58 ms
64 bytes from n900.orlandi.com (62.212.12.15): icmp_req=237 ttl=64 time=28869 ms
64 bytes from n900.orlandi.com (62.212.12.15): icmp_req=238 ttl=64 time=27870 ms
64 bytes from n900.orlandi.com (62.212.12.15): icmp_req=239 ttl=64 time=26870 ms
64 bytes from n900.orlandi.com (62.212.12.15): icmp_req=240 ttl=64 time=25871 ms
64 bytes from n900.orlandi.com (62.212.12.15): icmp_req=241 ttl=64 time=24883 ms
64 bytes from n900.orlandi.com (62.212.12.15): icmp_req=242 ttl=64 time=23884 ms
64 bytes from n900.orlandi.com (62.212.12.15): icmp_req=243 ttl=64 time=22890 ms
64 bytes from n900.orlandi.com (62.212.12.15): icmp_req=244 ttl=64 time=21890 ms
64 bytes from n900.orlandi.com (62.212.12.15): icmp_req=245 ttl=64 time=20891 ms
Having the same problems as vihai, but with a Acer A500 tablet & RB751G
Once connected to the rb751 there are no problems, but once the device goes into power saving mode/std-by and is awoken it will hardly pass any traffic. I have to disable the wifi on the tablet and reconnect to get any kind of data through.
Prior to the using the RB751G I was using a RB711-UA-2Hnd as a AP with no problems.
I did some quick tests with an iPhone and RB751U-2HnD RoS 5.17
First I ran the test with the phone very close to the AP. The TX CCQ was pretty low but pings responded ok.
I moved the iPhone 2 metre away from the AP and TX CCQ improved to better than 80%
FYI My RB751U-2HnD is set to 17dBm rtaher than full power. (Regulatory Mode and Country=NZ).
When the iPhone screen went black the pings stopped responding. The iPhone also “deregistered” from the RB751U-2HnD.
When I touched the control button on the iPhone to “wake” it, it registered on the RB751U-2HnD and the pings started responding again ok within a couple of seconds.
In all cases the pings varied from 2ms-100ms which was not ideal but certainly still usable.
The main thing I think to take from this is being to close to the AP or having too much power could be an issue and will cause lost packets etc. We see the same thing when testing a pair of SXT’s on the bench. If you do not drop the power they do not work well.
Hard to believe but too much power can sometimes be bad
We’ve started deploying this device when it was launched, we have about 50 of them in the field.
We must say, we had lots of trouble with this device (concerning wireless). We’ve made tons of supout’s and Mikrotik did lots of efforts to fix it. It went better when ROS version nrs increased, but it still ain’t stable as we expected. We 've ended up reconfiguring most of them without any wireless activated and added an extra ubnt access point to obtain wifi. These configurations are as good as 100% stable. Still sad it’s a no go with RB751-2HnD and wireless enabled as price is good, but we can’t afford to loose stability. Loosing stability is loosing customers and we can’t risk that. For this reason we discontinued using this device in new deployments.
I may have to just use that instead and shelf the 751u