4 RF transmitters on 1 antenna

[Tantal]

I have 4 RF units (each in the 1GHz band) that I would like to connect to one antenna. Can I simply wire all 4 RF connectors together and connect them to my antenna? Is the resulting signal simply the accumulation of the 4 individual signals?

Or do I need to make sure that only one of my 4 RF units is transmitting at the same time?

[KJDS]

I have 4 RF units (each in the 1GHz band) that I would like to connect to one antenna. Can I simply wire all 4 RF connectors together and connect them to my antenna? Is the resulting signal simply the accumulation of the 4 individual signals?

Or do I need to make sure that only one of my 4 RF units is transmitting at the same time?

Of course you can wire them all together. They won't work but you can do it even though you shouldn't.

You will need to ensure a number of things. the first is that each transmitter needs to see a good load, usually 50 ohms otherwise it won't put out as much power as it should and won't work as efficiently as it should. If the load is a long way from the desired impedance it is possible that the transmitter may damage itself.
The second potential issue is that the transmitter may send it's output power into anther transmitter causing damage to that.

So you will either need to run them at different frequencies and use a multiplexer, that is a combination of filters that can be used to combine the different transmitters or else you will need to not only transmit at different times but also use a switch capable of handling the power and frequency of the signals.
There are further ways of approaching it, including iso-hybrids and iso-combiners but I've finished my post lunch coffee and should now do some work.

[VK5RC]

Are they all on exactly the same frequency if so you could use hybrid couplers as in picture below, splitter on input on right side, combiner on left side, 1296MHz, 20W in 300W out. Must be in phase, my feed lines are the same length and the length through the couplers is also identical electrically (different velocity factor for different PCBs).

[Tantal]

... you will need to not only transmit at different times but also use a switch capable of handling the power and frequency of the signals.

Would this be such a switch? http://www.digikey.com/product-detail/en/HMC241QS16ETR/1127-1009-1-ND/3452150

[KJDS]

... you will need to not only transmit at different times but also use a switch capable of handling the power and frequency of the signals.

Would this be such a switch? http://www.digikey.com/product-detail/en/HMC241QS16ETR/1127-1009-1-ND/3452150

That would be ok so long as you're working with less than 200mW

[German_EE]

Think about this for a moment, there are thousands of TV masts out there and each of them are driven by multiple transmitters. Here in Germany you might find ARD, ZDF, RTL, and two or three others all going up the same coax to the antenna at the top.

So how do they do this?

1) Each transmitter has on its output a VERY sharp bandpass filter called a cavity filter. Although the transmitter is operating at UHF the width of these cavity filters is measured in KHz and outside the passband there is massive attenuation.

2) The output of each transmitter is designed to be 50.00 ohms +/- j0, in other words it's as close to a perfect fifty ohm source as they can make it.

3) The transmitters are fed into a combiner. I'm pretty sure that this won't be a resistive combiner but I also don't know the technology behind it.

4) The coax (which is of very high quality) and the antenna are both matched as close as possible to the output impedance of the combiner. An SWR of 1.01:1 is probably too high because of the powers involved here.

For 1 GHz you would have to do something similar, it's still possible to use coax rather than waveguide but it needs to be high quality material. A cavity filter at 1 GHz is a silver plated thing of beauty.

[KJDS]

Think about this for a moment, there are thousands of TV masts out there and each of them are driven by multiple transmitters. Here in Germany you might find ARD, ZDF, RTL, and two or three others all going up the same coax to the antenna at the top.

So how do they do this?

1) Each transmitter has on its output a VERY sharp bandpass filter called a cavity filter. Although the transmitter is operating at UHF the width of these cavity filters is measured in KHz and outside the passband there is massive attenuation.

2) The output of each transmitter is designed to be 50.00 ohms +/- j0, in other words it's as close to a perfect fifty ohm source as they can make it.

3) The transmitters are fed into a combiner. I'm pretty sure that this won't be a resistive combiner but I also don't know the technology behind it.

4) The coax (which is of very high quality) and the antenna are both matched as close as possible to the output impedance of the combiner. An SWR of 1.01:1 is probably too high because of the powers involved here.

For 1 GHz you would have to do something similar, it's still possible to use coax rather than waveguide but it needs to be high quality material. A cavity filter at 1 GHz is a silver plated thing of beauty.

There are a number of ways to make that work. What needs to be remembered is that although the filters are a lovely 50 ohm match at the wanted frequency, in their stop bands they are very very reflective. If you combine a group of outputs together in a star formation then you have a collection of something that is a short circuit at the other transmitter's frequencies on the end of the star. It can be done but it creates a lot of issues.

One approach is to use a manifold arrangement that connects a series of paired filters that a situated between hybrid combiners. This keeps the energy of each from the other transmitters. It also makes maintenance much easier as you can swap out a section, or add to the manifold without having to retune every other element in the system.

I also think I need to comment about your statement that the bandwidth of a TV transmitter cavity filter is measured in kHz. A TV signal is a number of MHz wide. Additionally, it's quite an art to get good temperature stability from very narrow cavity filters. Aluminium has a temperature coefficient of around 22ppm. If the filter has to operate from -10 to +40 deg C then that's a total shift of 0.1%, so 1kHz of center frequency shift per MHz of operating frequency. That would cause a 700MHz filter to drift by 700kHz over temperature. There are a number of techniques to improve it, but it's a struggle to improve it by much more than an order of magnitude.

[German_EE]

Yes, you are probably right about the cavity filters for TV transmitter use as I had forgotten about the width of a TV signal. My only experience of cavities so far was on a 70cm repeater and the bandwidth there was a lot narrower.