Thought that it might be interesting to start a new thread and explain some of those darker mysteries attached to the old age problem of using multiple radio cards in the same enclosure.
In the last few weeks there has been yet another surge of worried users relating to more or less the same subject.
Whilst I have studied the posts one by one and mentally noted each editors complaint, it seems that there is one common point of interest that is absent, but may be most relevant to the problem.
Bare with me newbies,
In an ideal world the Source impedance Zs of the final devices (after impedance transformation) is 50 ohm resistive, and only when the Load impedance Zl is exactly the same, will maximum power be transfered.
That is in an ideal world… However, I have seen first hand, N type connectors bagged up, labelled ready to go, for microwave frequencies, that personally I wouldn’t give house room to on short wave frequencies, let alone at SHF Microwaves.
I have no idea what the schematics are for the final stages of the WLAN rf ouput devices or whether they have VSWR protection, be it, fold back or limiting.
Nor do I have access to the pcb layout.
But when there is a mismatched load and a high VSWR, either through inappropriate connectors, poor assembly, poor quality coax, bad antennas, antennas positioned too closely ( the list is endless), under these conditions the coax could radiate!!!.
Isnt it true that with a perfect power transfer, the return current flows microns beneath the surface of the outer shield (ON THE INSIDE), but with a VSWR other than ideal ( 1:1) there can be both current and voltage on the outer of the braid.
The return VSWR will travel down the coax back from the load (antenna) back and forth. And in many circumstances, when the return currents or volts are in phase with the forward volts or amps, then the voltage at these points along the coax can increase markedly. ( the same said for current, but 180 degrees out of phase)
The positions are relevant to wavelength and occur at precise points along the cable.
What we all lack is our need to know SWR or VSWR. I think I can speak for most and say that 99.99% of us do not have access to Vector Impedance measurement equipment at SHF, or Spectrum analysers with tracking generators, or VSWR bridges.
However, if either Ubiquity or Mikrotik would like to pick up the challenge, I see no reason why the bridge can not be incorporated into the design of the Wlan card, either by absorbsion method, wheatstone or diode detectors.
We only have our suppliers word that a panel antenna is “flat across the band”, and even then, I suspect that most of us are installing them incorrectly.
And when I mean incorrectly, I mean, when relative to another object that is also resonant on the same frequency. It doesn’t even have to be another antenna within the capture area, it could even be a bolt on a mast that happens to be the wrong length. Worse still a rusty bolt who’s corroded joints become a “diode detector”
I’ve seen all these things happen at VHF and the principle is the same at SHF.
For those that have concerns regarding multiple radios in the same enclosure, maybe its time to do some testing on a long bench with some 50 ohm loads plugged up.
Set up say three Wlans with the centre case containing twin radios. Do some CCQ tests.
Then on the dual radio unit, plug up the antenna port with a T connector and 2 x 50 ohm loads. (leave the room or return to the exact same operator position (reflections!!!))
Then measure the CCQ again, my bet is that with increased VSWR the pigtails may have a tendancy to radiate and proportional to the level of mal SWR.
Also I note that the substrate used in the PCB design of the WLANS whether its Ubiquity or Mikrotik is not ceramic but standard (all be it high quality) pcb material.
Would it not be an idea to swop over to ceramic substrate and have the resistors etched or laser cut.
The VSWR bridge can be switched in on another parameter called “VSWR measurement” the WLAN card could be made to step every 5Mhz, self calibrate and self test.