Combining (phasing) amplifiers

The following discussion about how to combine (phase) two amplifiers took place in the Moon-Net reflector. I found the subject so interesting that I decided to put together all the messages in this page.

On 1-August-1999 4S7CCG wrote:
Currently I have only 100W on 2M.  Import restrictions due to the civil war here preclude importing an amplifier, and enquiries for power tubes and components within the local ham population have not been productive, however there are several VHF 100W "bricks" available.

I have tried simple coax power dividers on the input and output to run 2 of them in parallel.  There was success driving two with a power divider on the input and getting 100W from each into separate power meters and dummy loads, but I don't get 200W when the outputs are combined with a coax power divider.  Some phase problem I guess?

Can anyone help with suggestions on how to combine VHF solid sate power amplifier "bricks" for more output?

On 2-August-1999 VE7BQH wrote:
You have run into the common problem with trying to use standard power dividers. At best you have only 6 dB isolation and a ton of interaction!. A Wilkinson divider would probably work better BUT trying to find a high power non inductive resistor would be a problem.

The ONLY way to couple two amplifiers together successfully is to use a true Hybrid ring this gives you near 40 dB of isolation between ports.

Attached is what you need to build a Hybrid Ring. You need two, of course, one for the input and one for the output. If you do not have a dummy load capable of half your total outpower, make one of a LONG roll of RG 58 with a small load on the end. While he used plumbing fittings, you can get inventive considering you are a resourceful guy!! You could make the input one of RG 58 and RG 59. RG 58/59 might get a little warm on the output.

On 2-August-1999 SM5BSZ wrote:
Combining solid state brick amplifiers is a very good idea. Just connecting them in parallel like you did may be dangerous to them because if the phase is incorrect they will se a very large SWR.

You may well use a normal power divider at the input. If both amplifiers are near 50 ohms they will not cause any problems. At the output you should use a hybrid. It can be a made from some lengths of coaxial cable.

I guess you will get the description of a "rats race" from several sources, so here is an alternative that you may find easier to make:

First, with a power combiner only you have:

          Tx 1---------------------
                                   |
                                   |
                                   |A
                                   |
                                   ------------- out
                                   |
                                   |
                                   |B
                                   |
          Tx 2---------------------

The transmission lines A and B have the impedance required to transform the two 50 ohm sources (tx1,tx2) to a single 50 ohm source. They are both one quarter of a wavelength.

As a consequence the combined line A+B is a half wave transmission line that causes a direct connection between the two transmitters. If they are exactly equal in phase and amplitude that is ok, but if they are very different both transmitters will see the power from the other one as returned power due to SWR.

To avoid the problem you should connect one more transmission line parallel to the A+B line, and that line must be one half wave longer. The impedance does not matter, a cable that is n*(half wave) does not transform the impedance.

There will be no power transmitted through the extra cable if tx1 and tx2 are in phase with the same amplitude. If the phases or amplitudes differ there will be power flowing through the extra cable.

By making the extra cable lossy, connecting a load at one quarter wavelength from one end, both routes between the transmitters can be made equally lossy so the transmitters become isolated regardless of their phase.

          Tx 1---------------------
                                  /|
                                /  |
                              /    |A
         dummy load ---------|     |
                            /      ------------- out
                          /        |
                     -----         |
                    |              |B
                     ------------\ |
          Tx 2--------------------\

There is no need to make the dummy load properly matched. When everything is running ok there will be very little power going into it anyway. Use some small extra antenna, a simple dipole, a ground plane antenna, anything for the dummy load (if you have bad luck with the phases the dummy load has to take 200W so a small resistor is no good....)

Place a power meter in the line to the dummy load. In case the power meter shows there is actually power flowing in the line, try to make the cable lengths shorter or longer for one tx channel. Just make a small change and watch how the power ratio between output and dummy load changes. It does not matter if you change on the power side or the drive side. Once you have seen if it improved or degraded it will be easy to change cable lengths until there is no power lost in the dummy load. Then all the power goes to the antenna and you will have 200W.

Once the phases are correct you do not really need the extra cable and the dummy load but it is a good idea to leave them in place in case something happens that changes the phase or amplitude relation between the amplifiers.

An extra line at the input power divider will isolate the amplifier inputs from each other. It should of course also be one wavelength long and loaded one quarter from one end. Use a resistor - an antenna at the drive side may convert your power amplifier arrangement to a big oscillator!!

Perfect isolation will be with a 25 ohm resistor if you use 50 ohm cable or with a 50 ohm resistor if you use 70 ohm cable. Not very critical. (With A, B  and the extra cable all from 70 ohm line you have the "rat race")


(Note: I'm not showing any E-Mail address here in order to avoid them from being collected by SpamBots. You can possibly find the E-Mail addresses of the above OM at QRZ.COM.)

This discussion is still not closed. If you have any opinions or additional related information you would like to be published here, just send me an E-Mail

Read other Moon-Net discussions

 
Share: Twitter   Facebook   Home page Site map Radio Sherlock search engine Terms of service  Privacy  Cookies