Hello,
I just want to share my experience with others regarding new 5mhz channels. It is for sure that using 5Mhz channels we get better strengh due of focusing the power of card to smaller channel … but THIS KILLS the WIRELESS CARDS. It works perfectly for a period but after the card transmits very low. For example if the signal were -60 now we have -90.
I would say to Mikrotik staff that when the add new features, to better test them.
All cards are R52 with RB511 and Integrated antennas, lattest version of MT.
It happend to 3 units till now and just to the Basestations that are with 5Mhz band.
I’m curious why if i’m setting the tx-power-mode to default,
and i’m doing /interface wireless monitor wlan0 as example
it shows that default power is very high ?
eg : -26 dB for 6 MBps
Aren’t that value to high and can damaged the card ??
If it’s any help, I just had an interesting experience while trying narrow band mode that sounds similar to what kujtimhajredini describes but appears to be reversible and definitely not lightning-related.
Layout is an 2.4ghz MT client and AP situated about 42km apart, bridging a group of customers down in the jungle. This is a legacy arrangement that we’d like to clean up with 5ghz gear but in the meantime we thought we’d try an interstitial channel at 5mhz bandwidth so that we could get this long link off of useful local access frequencies.
Before switching to 5mhz bandwidth, the link was running at about -77dBm. After switching to 5mhz bw we saw a pleasing increase to about -68dBm RSSI. Throughput was decent. However, after a few minutes of operation the RSSI dropped on both ends to -90dBm and communications with the client became impossible. Fortunately we had scripts running to do fallback to conventional 2.4 g-mode operation if the link failed.
This episode did not appear to do any harm to the cards. After the scripts recovered the link, our RSSI values were exactly where they were prior to starting the experiment.
There’s no fundamental reason why concentrating output power from the radio into a narrower bandwidth should harm anything, so I wonder if this is some kind of bug in the wireless firmware?
Budi-a’s remark about power levels is kind of scary. We may allow the blue smoke inside the chips that makes everything work to escape! Maybe I should redo this with the power level set down, manually.
Edit:
Upon further examination, it appears that when narrowband mode was first invoked the tx power setting on our sample units were spontaneously fixed at 20dBm instead of using the normal power/speed curve. This exceeded the capacity of the cards (max 18dBm) and presumably caused some of the essential blue smoke to begin leaking out of the cards, leading to a drop in RSSI.
This spontaneous change does not appear to happen after the first time narrowband mode is invoked.
May i ask you what mode of TX-power that do you use ?
default
all-rate-fixed
card-rates ?
And in your conclusion :
a. first time invoking 5 mhz channel change tx power from 18 to 20 ?
b. second time invoking, third, fourth, and so on doesn’t invoking
change in tx-power ???
PS :
Please share your script
I’ve problem with strange fading in MT based system,
as example :
After we set-up the link we got RSL about -68 dBm,
after a few time (days, weeks) the signal fading to -78 dBm,
we thought maybe its a hardware / alignment problem,
but after we doing something that resetting the card like :
change setting (frequency, periodic-calibration, and so on)
but after that we change back to its previous values,
the rsl came back to -68 dBm.
Any of you have script that could be used for that kind of problem ???
We’re using manual tx power, with the cards dialed pretty far back from the maximum output for a given rate. That was the case prior to the units volunteering to change their mode to fixed.
The spontaneous change of tx power happened only once, the very first time we made the change to 5mhz bandwidth. Rebooting the router does not cause the same problem to happen again.
It also appears that the cards only began to overheat or lose sensitivity when we began running bandwidth tests on the link (local time of day when testing was about 5:00am so the radios were not in much use by customers).
I’m wondering if this problem only happens when a radio that was previously not running “wireless-test” or a later version of firmware with standard support for narrowband is upgraded with the later firmware supporting narrowband and then configured for narrowband operation. In our case both routers were previously running 2.9.5 and were upgraded to 2.9.13. Neither had ever been used with the old “wireless-test” package. The tx power weirdness did not appear on the upgraded radios until we invoked narrowband operation.
The script is extremely simple (simple minds here, simple scripts ), . It just changes the band mode to the desired setting, waits a bit to allow reassociation to happen as well as whatever monkeying around we need for synchronizing settings/testing on the other end, warms up with a few pings to the target on the other end of the link, then does a test ping. If the ping fails, it drops back to a “standard” operational band:
/interface wireless set wlan1 band=2ghz-5mhz
:delay 300
/ping 10.0.8.14 count=5
:if ([/ping 10.0.8.14 count=1]=0) do {
/interface wireless set wlan1 band=2.4ghz-onlyg}
Other more clever methods are certainly possible, heh!
One feature here I don’t really understand is that after the script fired and took us back to conventional mode, communications were improved enough to get things squared away tx power-wise, even though the tx power was still screwed up according to the MT’s output. Hmmm.
Have any of you notice TX Power problem and have any
occurence of mini pci card damages, when you using
default tx-power-mode ??
I’m a little afraid with MT tx-power-mode default because
when we’re monitor using /interface wireless monitor wlan0
it said its tx-power from 6 Mbps a/g mode is 26-27 dB ?
Aren’t its too high for CM9 or its clone ? AR-5213 based ones ?
We’re worry of possibility damaging the card for long time
operation.
Any of you see the same problem like mine ???
Any fact / statement from authorative figure would be very appreciated.
eg :
If we set tx-power-mode to default,
what’s the actual tx-power in dB will the card operated at ???
), . It just changes the band mode to the desired setting, waits a bit to allow reassociation to happen as well as whatever monkeying around we need for synchronizing settings/testing on the other end, warms up with a few pings to the target on the other end of the link, then does a test ping. If the ping fails, it drops back to a “standard” operational band: