The AM Forum

i want to build a directional coupler with 20dB coupled port loss for use in sampling the output of high frequency amplifiers on my test bench. i would like the coupler to be usable in the range of 1 to 30 MHz.

i plan to leave both coupled ports unterminated so that i can use different 50 ohm terminations depending on power requirements.

i am following the design laid out here: http://michaelgellis.tripod.com/direct.html
to get 20dB of coupled port attenuation i need a 1:10 turn ratio on both transformers.

i am unsure of how to select a core material. i have read that iron powder cores are more stable and superior to ferrite for high power inductors whereas ferrite cores have higher AL values and thus require less turns. i would like the device to be able to handle at least 20W to 50W, or more if feasible.

can anyone suggest an appropriate core material given the specs i am hoping to achieve?

http://www.g3ynh.info/zdocs/bridges/Xformers/part_2.html

Scroll down to core selection.

Chris

thanks chris! i'm going to read this whole article really carefully.

hi guys,

i went ahead and built a directional coupler using michael ellis' design. i used what i believe are FT82-43 cores (i bought them from the scrap yard so they aren't labeled).

i used a 5:50 turns ratio which should give me 20dB coupled port attenuation. i dont know if i really needed 5 turns on the primary and i may rewind the cores once i do some tests on them.

Port D is NOT internally terminated. i decided to do this so that i could use large sized dummy loads in case i ever test a big power amp. i read online that typically FT82 cores can handle 50W to 75W on the HF band, i hope that is accurate.

i was all ready to test the device when i realized i only have two dummy loads. i need a 50 ohm termination on my scope in order to monitor the coupled port, right?

Reading the title got me to wondering. I have a number of unmarked cores (some round and some that look like they should slip over a cable). Have no clue where they came from over the years. Anyone have a bench test design to determine what they are?

hi pete,

you probably already know this but you can figure out the Al value from the inductance per turns then compare that, along with the core dimensions, to product catalogs to try to figure out the material.