Hello,

I'm hoping to get some direction on building a multi-band cellular system.

My cell tower has two active bands for use with my provider.

Band 25 uses 1850-1995mhz
Band 26 uses 814-894mhz

My cellular device is capable of communicating on either band. What i wanted to do is get a high gain yagi for the lower band 26 frequencies, and a high gain wifi grid antenna for the higher band 25 frequencies. The antennas would connect to a single run of low loss 50 ohm coax through a 2-way splitter/combiner. This would allow me to change bands on the cellular device without physically swapping antennas.

My questions:

Any issues combing antenna s of different frequency ranges like this? It seems like it should be fine from a receive standpoint.

Will I loose 50% of my transmit power? If I'm transmitting on the high frequency band, is the low frequency antenna get half the power?

Would the use of high, low, band pass filters between the antennas and splitter correct that issue?

Suggestions welcome

 

What you want to do is not feasible with just antennas and cable. The coaxial cable you need to use for this applications is called "hardline". It is expensive, difficult to work with, and the connectors are hard to come by. The "2-way splitter/combiner" is likely to involve multiple sets of cavity resonators. And the dollars just keep piling up. Then there is the test equipment required to verify the construction, setup, and tuning. In short, I think the odds are against you succeeding in this project without a great deal more engineering.

Along a run of coax, every impedance discontinuity, will result in part of the signal being passed, and part of the signal being reflected back to the source. Controlling losses and reflections will determine the operational efficiency of your scheme.

 

Thanks for your response Papabravo. I should have given the below information, but I didn't want the original post to be too long.

Yes.. I already have the "hardline" cable.

I have a 125 ft run of Draka RFA 1/2 50 ohm coax feeder cable. The attenuation is less than 4dB@2.4Ghz for the entire run with connectors. The 50 ohm cable is terminated with N-Male connectors. The splitter I was contemplating on using is terminated with N-Female connectors and has an insertion loss of 0.5db. On the equipment side of the cable, I have a polyphaser tsx-nff lightning arrestor (less than 0.1db insertion attenuation) that is grounded to the home common ground, and then a small length of LMR 400 ultraflex low loss coax that connects from the lightning arrestor to the equipment. This setup works wonderfully with a single antenna. The entire attenuation at 1.9Ghz frequency is less than 5db from antenna to cellular device. You are correct, this cable is over 1.2" in diameter and hard to work with, which is why I use a small 4ft piece of more flexible cable on the equipment side.

Cost wise-- I have about $400 into it that also includes a 24dBi 2.4Ghz grid antenna (which gives me about 19/20dBi @ 1900Mhz). It's definitely not cheap, but you do what you have to do when you live in the middle of nowhere

So my original question pertains to issues adding a splitter and second antenna for the 800-900Mhz spectrum.

My goal is to be able to switch bands on the fly depending on the situation (without climbing up a 30ft tall antenna tower to physically swamp antennas).

1) is one band more congested than the other?
2) atmospheric conditions
3) band 26 (lower frequency band) has better propagation than band 25 which is useful during summer when tree foliage can attenuate signal.
4) Band 25 has better gain and generally lower congestion.
5) Is one band having technical issues.

These issues are important to me. I do on-call support from home, and it's an hour drive to the hospital if I have connectivity issues.

I think the downstream signal should be fine. Is there a difference between combing the signal of two frequency specific antennas vs. connecting a single wide band log periodic antenna? Besides the fact that the wide band log periodic antennas generally top out at 11dBi (probably less depending on the portion of the frequency spectrum used).

The upstream transmission part of the equation is where I am second guessing myself.

Pretty specialized situation. I need to pick the brain of a wireless engineer for about 30 minutes I live over 5 miles from the tower so every dB is precious.

 

As I see it the best solution is a remote switch box which selects the proper antenna. Using a single feedline means the two antennas are in parallel. In this configuration is is hard to get a proper match with acceptable return loss for both. Such a remote switch is often used when multiple antennas are placed on a single tower.

DX Engineering makes lots of good stuff -- like this
http://www.dxengineering.com/search...stem-device-type/remote-feedline-relay-switch

I know you need both receive and transmit so dig deeper and ask for a catalog.

or you might try something like this
http://www.birdrf.com/Products/Comp...ion-Coaxial-Selector-Switch.aspx#.VO552vnF_KE

 

As I see it the best solution is a remote switch box which selects the proper antenna. Using a single feedline means the two antennas are in parallel. In this configuration is is hard to get a proper match with acceptable return loss for both. Such a remote switch is often used when multiple antennas are placed on a single tower.

DX Engineering makes lots of good stuff -- like this
http://www.dxengineering.com/search...stem-device-type/remote-feedline-relay-switch

I know you need both receive and transmit so dig deeper and ask for a catalog.

or you might try something like this
http://www.birdrf.com/Products/Comp...ion-Coaxial-Selector-Switch.aspx#.VO552vnF_KE

Thanks,

Those are interesting solutions, and honestly an option that I had not thought of.

While they eliminate some problems that I was working around, they introduce some other issues.

Automatic switching will require DC power to be introduced into the antenna system which is not compatible with the cellular device. The lightning arrestor is DC blocking, so power could be introduced on the antenna side, but that would be outside the home in the elements. The switching mechanism itself would also be outside on the weather elements. The only way to do this and have the switcher inside the house would be to purchase another 125ft run of hardline.

Tricky stuff.

What I think would work is a diplexer. Unfortunately I haven't found one with the correct frequency range. Wilson (no WeBoost) actually makes one, but the frequency range is off just enough that it probably wouldn't work. It's actually made for this exact scenario, and the reason I was thinking if frequency filters would work. Isn't a diplexer a device that has multiple band pass filters in one housing?

Example:
Low frequency antenna would have a 1000Mhz low pass filter
High Frequency antenna would have a 1500Mhz high pass filter

 

Maybe, but the construction and tuning of filters in those frequency ranges is not for the faint of heart. You might need to rent some expensive equipment to test and adjust the resonant cavity filters for the job.