On 4/20/22 10:09, nick wrote:
Thanks for the feedback do you have any part numbers of those parts?
Hello Nick,
KF5IDY did some simulation to determine the differential output impedance of the AX5043, given the matching network shown the datasheet. His results are here:
Assuming the network is optimized for the frequency specified in the data sheet, the unbalanced impedance that yields minimum loss ranges from 62 Ohms to 149 Ohms, with the middle two entries at 100 Ohms. So, I'm going to assume the IC impedance for the unbalanced pins is 100 Ohms.
He may be persuaded to share his Qucs simulation file with you.
It looks like a straight 1:1 diff-to-unbalanced solution would work. I might look into these parts:
Mini-Circuits NCS1-521+ Johanson Technology 0430BL15A0100 STMicroelectronics BALF-112X-02D3 ( integrated harmonic filter, more research needed to see if it is a reasonable match to the AX5043 output impedance )
The nice thing about the AX5043, versus the little brother part AX5243, is the separate TX and RX paths. One can use the single-ended path with the proper harmonic filter (which has loss) for TX and then use the differential path with a simple balun for RX so that the LPF loss is avoided. The chip has an internal antenna switch for this purpose.
Note that, for full-duplex operation, two ICs will be required. They're cheap, so no problem there. This is what is being done for the GOLF program's RT-IHU design.
--- Zach N0ZGO