A little help with RF on a PC Board

[ljbeng]

In our GPS product, we install 2 antennas and feed two cables to an RF relay with 3 N-style connectors. We can then select one antenna or the other and feed that antenna to a Trimble GPS receiver. Once powered, 1 antenna is used, we don't switch back and forth. It depends on which direction we start moving. That has been working for us for 8 years.

I am now working on a new layout of this design. I want to get rid of the external N-Style RF relay and put an RF relay on a pc board. We also want to design for multiple GPS engines (Trimble, Hemisphere, etc..). These engines are small pc boards that lay on the main pc board so the electrical and the rf plug into respective connector on the pc board. I have already prototyped the pc board mounted rf relay and can successfully switch between antennas like our legacy equipment but I only built around 1 GPS engine.

My question is about the rf connectors and adding connectivity to multiple mfg cards. Can I put 4 rf receptacles (mmcx and mcx) all in series leading to the common of the rf relay? Only 1 would get a gps engine plugged in and the other 3 would be open or the connector unsoldered. Will this create too much loss or be open to interference?

Sorry for the long wind.

[jwax]

I couldn't prove one way or the other, but feel that there's no problem in doing what you propose. Several closely-spaced connectors on a board will not have a noticeable effect on the RF. IMHO-

[TuNneLvizn]

Good Day,

For closely spaced connectors, the effects on the transmission line will not be significant . The only thing that you are opening yourself up for is signal interference . The GPS signal is fairly robust in terms of interference; however, the interfering signal will contribute some increase in the noise bandwidth of the GPS receiver . This will decrease the signal to noise ratio (SNR) to what ever extent the interference signal possesses . As the GPS signal acquisition is very dependant upon raw SNR, it is critical to minimize noise influences .

If the GPS receiver is within its own enclosure (faraday cage) then the only interference will be from the circuitry within the enclosure ( e.g. local oscillator) . It would be best to use a shielded open over the unused ports if you find a degradation in the SNR .

Good Luck,
>>Tim

[dyarker]

At 1.57542 GHz even a quarter inch is a long conductor. And, you'll have about an inch leading to unterminated connectors? Lots of out of phase reflections to mess up the signal.