I have been tinkering with building a low cost media converter that would convert a single-ended 75 ohm coax serial (60 Mbs/8b10b/2Vp-p) interface to a UTP cable construction and back to coax at the receiver.
My objective with this is keeping it simple (low parts count/no power requirements etc...). What I came up with (see pic) is a simple dual channel Fibre Channel transformer(s) wired according to my schematic.
I was expecting to experience lots of errors relating to impedance discontinuities, signal reflections and attenuation etc........ The result was actually surprising to me...it works, and works well. I have not looked at the signals with any TDR scopes or the like...yet. The equipment to which it is attached has not complained...yet.
I contacted two manufacturers that make the transformer, and I got two distinctly different answers as to my intended hookup. One manufacturer said my circuit would absolutely not work because the four coils don't share a common core...fair enough. The other manufacturer actually built my schematic in their lab and confirmed that at least my hookup diagram was theoretically and practically possible.
I would like to have my schematic picked apart and please tell me why this shouldn't work.
Thanks for your time.
-Eric
My objective with this is keeping it simple (low parts count/no power requirements etc...). What I came up with (see pic) is a simple dual channel Fibre Channel transformer(s) wired according to my schematic.
I was expecting to experience lots of errors relating to impedance discontinuities, signal reflections and attenuation etc........ The result was actually surprising to me...it works, and works well. I have not looked at the signals with any TDR scopes or the like...yet. The equipment to which it is attached has not complained...yet.
I contacted two manufacturers that make the transformer, and I got two distinctly different answers as to my intended hookup. One manufacturer said my circuit would absolutely not work because the four coils don't share a common core...fair enough. The other manufacturer actually built my schematic in their lab and confirmed that at least my hookup diagram was theoretically and practically possible.
I would like to have my schematic picked apart and please tell me why this shouldn't work.
Thanks for your time.
-Eric
