Thanks guys for your direct or via reflector pieces of advise. Just a remark : the 25 ohms is correct in the input/serial configuration suggested by Bryant&Bower in the article I have mentioned previously while the 100 ohms is the right one in the most common output/parallel configuration of a splitter.
On the basis of all that, I am in the process to make some cut and try tests with different cores and turn numbers using my VNA 2180 . I am planning to test the two configurations (common and Bryant&Bower ) and will advise you of my final results. More info about the need : I cannot use a commercial splitter because I am intending to use two of them in a to be constructed switching box (as small and light as possible) which will dispatch two 75 ohms beverage inputs (JA and NA/EU directions) to two radios. One radio being for example the 80m radio and the other the 160m radio. The outputs to the radios must be 50 ohms because those outputs are going to the radios via W3NQN RX band filters and we want have the filters at they best in a multitransmitter environment. In the same spirit, the switching system manages to put a 50 ohms resistor at the unused output of the splitters. To finish, each operator will have a remote switch next to his radio to choose its directions. The common mode rejection at the inputs is just a precaution for expedition places where the quality of station ground is far from optimal. 73, Gilles VE2TZT _______________________________________________ UR RST IS ... ... ..9 QSB QSB - hw? BK