2.8.26, ptp link (3km, 24dB directional antennas, short coaxial cable to antenna etc) on fast atx machines, excellent stability and throughput, nstreme with polling. But signal strength and tx-signal-strength varies greatly:
and what's up here, tx signal strength is -57 but signal strength is -74. I don't care about it - as I said, this link is excellent, just the difference between -74 and -57 makes me kinda nervous
thnx for input, but I doubt it - we tried what we could and this is not our first installation. Pigtails were changed, coaxs were changed, cards were changed, antennas were changed… nothing helped. Usually I’m within 3dB of difference, right here I have almost 20
those antennas are in the same heights, literally, horizontal polarization. Fresnel zone is clear, some metallic structures around but not too many and we know this / these conditions by heart because this is installation in one city district (we have done many of them). Link is not running over water.
I don’t understand it, we will try to aim those antennas once again. The only think I could think of right now are not so tightened screws, so one of those antennas could move a inch.
yes. As I said, this is not first or second installation of us and that link works perfectly. I’m just curious about that difference - note that not too many people on this planet would notice something like this on PERFECT link. When you have problems, you start to investigate blablabla… but not on EXCELLENT link.
You may have local interference on just one antenna. It could be caused by some reflections on the rx signal or a local omni AP that is messing up only one side as the other side can’t hear it.
antenna are symmetrical,
they reecive as well as they transmit.
so are cables. So are fresnel zones.
The only non-symmetrical things are the actual transmitter and receiver circuits.
They are different.
Why ? either because of software /config, or because of hardware
You have dealt with the hardware issue..
so over to software.
cards have a ‘default’ transmit power level. perhaps you are relying on this default level to be the same ?
perhaps you actually have a tx power setting in one end that is different to the other end
I mean, what card are you using ?? prism chipset with no adjustable power levels ?
so are the cards the same model (maybe the label is wrong? or you mixed up the model numbers … )
I had observed the same non symmetrical behavior in one g link. I replace the cards at both ends to investigate the possibility for failed tx or rx and the problem remained.
It was not a mistaken indication because the throughput behavior of the link was also not symmetrical.
After that i replaced the feeder in the side where the rssi was worst and the rssi become symmetrical, also the throughput.
So i can assume, that if for some reason the VSWR is high (antenna, cable, connector) there is a decreased received signal strength. I know it makes little sense but this is what i observed.
Cards tested where cm-9 cards and if you look inside them they have near the antenna output, two printed lines (coupler) that probably measure VSWR(?)
Antenna connections are symmetrical, but there are asymmetrical factors that can degrade signal strength (or increase signal strength) on only one side.
there may be reflectors such as metal roofs, other antennas, buildings with rf reflective material that will reflect an incoming signal and degrade or cancel it when the reflected and direct signal are received
other interference that can also degrade the direct signal
water and cable problems that can have different effects on transmitted and received signals
But, I think that most likely you have a burned PA on one side. I suggest replacing that radio.
