a rb 532 w/ nmp 8602 card to a 5.5 db rubber ducky omni.
to
a RB112 which has 2 nmp 8602 cards in it (I’m going to use these things as repeaters and are what im testing here.)
The 532 is just an access point for my cable line, for my house. AP-Bridge, DHCP server, Masq Nat.
I have the rb 112 set on wlan 1: as station and wlan2: ap bridge with its own dhcp-server.
everything works great, i can connect any of my 4 laptops to the WLAN2 of the repeater and surf/ping GW just fine.
the only issue / question i have is why when both sides show -50 signal, bw test 10mbits up or down to each other, -99 noise floor, have a disconnect every hour or so.
it reconnects just fine, its just i have laptops connected to this Access point that will stack up 10s of days of uptime, and this radio to radios setup, with MUCH better signal strength cant even keep up an hour?
as for the log messages, its honestly every type of wifi disconnect: reassociating, no beacons, decided to deauth: group key handshake timeout. different things different timies.
any ideas? both sides are MT 2.9.3 (host) and 2.9.26 (rb112)
EDIT: I’m not using WDS, since i used to have a WDS connection in the field and it would never be stable…(this is different equpiment btw.)
rb532’s work with 8602’s a lot better than a rb112 will. Go read the specs for the power available on each RB, then read the specs for the max power required for a single 8602. You’re lucky you haven’t burned out the voltage regulater on that rb112, and it sure isn’t going to be stable.
You can have it cheap or you can have it right, pick one.
The RB112 can easily provide enough power for a single SR2, SR5, SR9, and NMP-8602+
And there has been known issues with some NMP-8602+ and RB532s, just search the forums.
I have for example had NMP-8602’s that causes an RB532 to reboot whenever its at 5ghz but they work fine in RB112s. As a rule I only use SR5s whenever operating at 5ghz.
As far as the disconnect issues in my testing I have been able to repeat odd disconnect behariour by a few different methods.
Interference - I have had a customer with a baby monitor knock both his internal wireless off and his CPE off. Whenever he turned it off or on it would pulse across the entire spectrum and disrupt the signals long enough for it to cause a disconnect. His signal strength was -52. I’ve had the same issue with numerous poor quality 2.4Ghz phones.
Good speed, good signal, but antenna is getting reflections off of other objects - A good example of this was a problem I was having with wireless inside a customers house. He was using a new type of insulation on his outer walls that was lined with foil, and he had a metal roof put on. Speeds were great, signals were excellent, but every 2-3 hours all his computers would disconnect for approx 5-10 mins then all reconnect. The issue was that the signals could not ‘escape’ his house and would just continue reflecting and bouncing around until there was so much old RF ‘noise’ that the connection would not stay stable. Once all the computers disconnected and stopped transmitting, the old RF ‘noise’ would slowly dissipate during the 5-10 mins that the process would start all over again.
thats true regurading noise…but here is why i think that is flawed:
I have 4 laptops that are in further parts of the house and they are generally plugged into electical sockets and will rack up DAYS and DAYS of wifi uptime as read @ the MT box…while the RB112 in Station mode will only last 5-30min.
. reguarding the nmp 8602:
I have TONs of experience with these cards.. TWO 8602s will work fine on RB 112s…
532s will give you tons of random problems with the 8602..read around i have a huge thread on the 532 + 8602 where eugnene finally admited there is a 8602 / 532 specific issue…
infact if i have a card that causes the 532 to continously reboot i will use a rb 112 to reset the cards or set the settings to what i want, then put it back in teh 532 and it will work fine…as long as i dont change any settings
My ‘garbage’ is easily proven. Take a spectrum analyzer, a few wireless devices, an AP, put them in a some completely RF sealed room and watch what happens on the SA. Just because I used a ‘real world’ example does not mean I didn’t look at the science behind it.
This is not really a multipath problem, multipath refers to how the 2.4ghz propagates from radio to radio. And yes that can be a problem as well, however this problem has to do with the amount of background noise that is generated over time.
For example when radio A transmits to radio B, and after radio B has acknowledged the receit of the transmission, where does that radio wave go. It doesn’t stop dead, it will keep bouncing around until the microwave energy has become so low you can consider the wave ‘dead.’ In most cases the wave will bounce away or into materials that absorb the energy, thus killing the signal fairly quick, but in some cases for example in small area that is effectively rf sealed. The signals can remain around long enough to start causing background noise problems.
I have done both extensive research and ‘real world’ testing to prove this. If you don’t believe me fine, these forums are for helping and I am just suggesting something that ‘could’ be causing a problem.
Radio energy dissipates within microseconds and does not build up after time. Otherwise should keep the microwave oven closed for an hour after you heat your hungry man dinner to keep from frying your eyeballs.
Radio energy dissipates within microseconds and does not build up after time. Otherwise should keep the microwave oven closed for an hour after you heat your hungry man dinner to keep from frying your eyeballs.
Except that your hungry man dinner absorbed the microwaves - that’s why it’s hot!
And, after all - after waiting for an hour before opening your micro wave, you could just start it all over again, as your hungry man dinner will be cold again after that hour.
i have similar problem with
[configuration1]: SR9 rb112 and 8km link with no line of site. signal strength is 85-87dBm but we have no link actually. the station and bridge sit connected all time, but ping is bad. noise floor 97, 2427 2GHz-10MHz, antennas 15dBi yagi.
IMHO On all my 532ap+SR9 to 112+SR9 clients, I must have at least a -70 link before I will even consider it stable, Even at that strength I am losing association if the noise floor drops down to -85 from its normal -95. The best investment I have made is noise filters. They do help, but the links will still drop for a split second from time to time.
