Recently I have been reading a lot about near or ajancent interferance. I did a quick test of self interferance by installing a XR2 and a CM9 card on the same rb. I had no pigtails installed while running the test. Scan from the CM9, I saw a 75db from the XR2. Scan from the XR2 I saw a 85db from the CM9.
According to several documents I have been reading this situation will cause a significant reduction in performance of the receiver. I have entered a couple of links if you are interested in reading more.
OK, I am in way over my head but from my simple understanding of what I have read the receivers see rf from other channels as noise. That being the case it seems to me that we are raising the noise floor to 75db and we need to not install more then 1 card per band in a case.
I would have been more interested in the test results if the outputs of the two cards were terminated in non-radiating dummy loads rather than left unterminated. Being unterminated will increase free radiating RF within the enclosure. (I’m presuming the tests were carried out in an RF free environment with the unit installed in it’s original metal and earthed box?) I would also be interested in the results of rerunning the test after swapping over the two card positions.
Also the “self induced” interference one gets from nearby channels on your own site also depends on the aerial types you’ve used and their positioning with respect to each other on the mast / pole, also coax type(s) used etc.
The tests are meaningless to me as they stand at the moment but if you can redo them with the changes as suggested, it might be very interesting to see the results again.
Not looked at the references you’ve quoted, but will take a look.
I asume the 79 and 80 is the db you got from card 1 to card 2 or visa versa. When you did you link speed test what was you noise floor and you signal level from the other end.
This was neven intended to be an accurate test. It was a quick test I ran after reading the docs in the links I mentioned in my post.
What I plan to do is to select 2 towers as described below. We are going to select the water tower and tower based upon having similar noise floor levels. All antenna, radio cards and router boards and os2.9.x are the same and the same version. The 1st is a water tower sites with sectors mounted on the hand rail and 1 radio in eack box. The second is a tower with 2 boxes, 1 with 1 2.4 AP and 1 5g BH and the second box has 2 2.4 AP cards. The sectors are mounted on the legs of the tower at the same elevation and only 36" back to back.
Any suggestions to make the tests more accurate would be appreciated.
My Spectrum analyzer shows some interference in 10 Mhz near channels So i made a rule that there are only 2 non-overlapping channels. i.e i always assume that the channels are 20 Mhz and 10 Mhz interference on both side.
I always use Metal enclosure properly grounded and the power supply should have a high filter coil at the output.
Sometimes instead of using Direct PoE , I use PoE Injector and D.C Power splitter on the AP and directly power the Unit through the D.C Port. This has helped me many times.
All Connectors should be properly sealed by weatherproof material.
The tower ground should be isolated from other electrical equipments.
Always use STP Cable whose conductor jacket should be properly grounded on both ends.
I use that amazing G.E. Outdoor silicone too…the big kind you apply with a gun.
i just put that everywhere after sealing and taping…its great.
ive put NO tape and NO weather tape on N connectors in Salt water environments (as a test, i have alot of stuff on beaches and near them) and just coated the entire thing in silacone…amazing…after like a week you can just tear the silicone off…and after around 9 months i tore it off…looked like the connector was BRAND new… i was shocked.
I too use heat gun with RTV Bars. Those are good … atleast it keeps the connectors look like new. but i am unsure how it is effective in keeping the interference out.
Gotta see the electrical characterics of my silicon bars.