Antenna Theory: Twisting the antennas of a Transceiver together
deef, Fri Jun 01 2007, 04:33AMHello,
My antenna theory is a little weak, and I was unable to find any information on this subject. Lets see if one of you guys can help me out.
Basically I'm working with a transceiver that has two antennas (Obviously, one for TX the other for RX). Right now, I'm running two wires (where antennas would normally be attached) into a communications multiplexer that is allowing me to have one physical antenna. This multiplexer is controlled via standard I/O pins.
What I want to do is cut out the multiplexer out and have two physical antennas on my board. BUT, to keep things so neat and tidy, I was wondering if I could twist the antenna wires together so that they were in a vertical orientation, but like a tight double helix. Right now, I'm using small-guage magnet wire as an antenna, so they would be electrically isolated from each other at all times. That's an obvious must. BUT, I'm wondering about the implications of doing something like this from an EM standpoint.
My gut is telling me it wouldn't be an issue, as the "send" and "receive" portions of the transceiver are never on at the time time, but I was just wondering if anyone had any experience on the subject. Right now, my communications multiplexer is giving me -3dB loss anyways, so there is a little bit of a margin.
Any ideas?
deef
Re: Antenna Theory: Twisting the antennas of a Transceiver together
Steve Conner, Fri Jun 01 2007, 08:32AM
If you twist the antennas together you will probably destroy the front end of your receiver by coupling most of the transmit power into it. What you're trying to do is usually done with a T/R switch made from PIN diodes. I think you can get high speed T/R switches on a chip now, for the Wi-fi market.
Steve Conner, Fri Jun 01 2007, 08:32AM
If you twist the antennas together you will probably destroy the front end of your receiver by coupling most of the transmit power into it. What you're trying to do is usually done with a T/R switch made from PIN diodes. I think you can get high speed T/R switches on a chip now, for the Wi-fi market.
Re: Antenna Theory: Twisting the antennas of a Transceiver together
deef, Sat Jun 02 2007, 06:09PM
Thanks for your response, Steve! You're a big help! I knew someone around here would have the knowledge to lend me a hand. I appreciate it.
Let me just ask a couple questions to make sure I'm understanding you correctly.
Even though these transceivers (TRF6903) are half-duplex, and the transmit portion of the chip is essentially "off" while in receive mode, it's still possible to damage the chip as you mentioned? The design guide for the transceiver suggested the possibility of shorting out the two antennas directly, and placing one antenna on the shorted junction. Obviously for cost savings (you save an antenna). They said this method was possible so long as dc-blocking capacitors were also in place for both the transmit and receive tuning circuits.
I was more worried about degradation of RF performance (eg: loss of radiated power). The thought of damaging the chip straight-out makes sense, although it didn't really pass through my mind as I asked the question.
Again, if damage as you described is possible, I imagine there would be a proximity where it would be "safe" to run two, separate antennas which would be untwisted and physically far from each other on the board. My gut tells me this has something to do with near fields and far fields.
Thanks again,
deef
deef, Sat Jun 02 2007, 06:09PM
Thanks for your response, Steve! You're a big help! I knew someone around here would have the knowledge to lend me a hand. I appreciate it.
Let me just ask a couple questions to make sure I'm understanding you correctly.
Even though these transceivers (TRF6903) are half-duplex, and the transmit portion of the chip is essentially "off" while in receive mode, it's still possible to damage the chip as you mentioned? The design guide for the transceiver suggested the possibility of shorting out the two antennas directly, and placing one antenna on the shorted junction. Obviously for cost savings (you save an antenna). They said this method was possible so long as dc-blocking capacitors were also in place for both the transmit and receive tuning circuits.
I was more worried about degradation of RF performance (eg: loss of radiated power). The thought of damaging the chip straight-out makes sense, although it didn't really pass through my mind as I asked the question.
Again, if damage as you described is possible, I imagine there would be a proximity where it would be "safe" to run two, separate antennas which would be untwisted and physically far from each other on the board. My gut tells me this has something to do with near fields and far fields.
Thanks again,
deef
Re: Antenna Theory: Twisting the antennas of a Transceiver together
Steve Conner, Sat Jun 02 2007, 08:39PM
Well, if the application note for the transceiver suggests just connecting the transmit and receive antenna ports together like you said, that implies that it already has the necessary circuitry inside to protect the front end during transmission, so you can go ahead and do this, or twist the antennas, without fear.
In fact, twisting the two antennas together should be electrically equivalent to using a single antenna with the DC block capacitors like the design guide suggested.
But one thing I'm wondering: if the design guide said you could short the two antenna ports together, why did you go to the trouble of hooking up a "communications multiplexer" (whatever that is) in the first place?
Steve Conner, Sat Jun 02 2007, 08:39PM
Well, if the application note for the transceiver suggests just connecting the transmit and receive antenna ports together like you said, that implies that it already has the necessary circuitry inside to protect the front end during transmission, so you can go ahead and do this, or twist the antennas, without fear.
In fact, twisting the two antennas together should be electrically equivalent to using a single antenna with the DC block capacitors like the design guide suggested.
But one thing I'm wondering: if the design guide said you could short the two antenna ports together, why did you go to the trouble of hooking up a "communications multiplexer" (whatever that is) in the first place?
Re: Antenna Theory: Twisting the antennas of a Transceiver together
deef, Sat Jun 02 2007, 11:58PM
Ahh, excellent point! Again, I never even thought of that but it makes perfect sense. So long as I'm using insulated wired it would be equivalent.
Why did I bother hooking up the multiplexer to begin with? I was worried about the distance I'd be able to transmit and receive signals. So, I did the original design using as much efficiency as possible. A different tuning circuit is suggested for transmit and receive. The multiplexer lets you take advantage of using separate tuning/matching circuits, but one common antenna. A little bit of gain is sacrificed by connecting the RX and TX circuitry, with shared tuning, and running a single antenna. But, I've learned there will still be plenty of range for my needs.
I'd consider my issue: solved.
Thanks again,
deef
deef, Sat Jun 02 2007, 11:58PM
In fact, twisting the two antennas together should be electrically equivalent to using a single antenna with the DC block capacitors like the design guide suggested.
Ahh, excellent point! Again, I never even thought of that but it makes perfect sense. So long as I'm using insulated wired it would be equivalent.
Why did I bother hooking up the multiplexer to begin with? I was worried about the distance I'd be able to transmit and receive signals. So, I did the original design using as much efficiency as possible. A different tuning circuit is suggested for transmit and receive. The multiplexer lets you take advantage of using separate tuning/matching circuits, but one common antenna. A little bit of gain is sacrificed by connecting the RX and TX circuitry, with shared tuning, and running a single antenna. But, I've learned there will still be plenty of range for my needs.
I'd consider my issue: solved.
Thanks again,
deef
Print this page