We’re testing an LTC5594 eval board (DC2645A) where a signal generator feeds both f_RF and f_LO.
We converted the differential IF outputs (IFIP/IFIM, IFQP/IFQM) to single‑ended with baluns and observed with an oscilloscope using probes and equalized cable lengths, yet we consistently measure ~80° instead of the expected 90°. The datasheet’s quadrature accuracy spec implies only a few degrees of error, so why are we ~10° off?
Are we just seeing the stack-up of that spec plus balun/cabling effects, or does this indicate a problem in our measurement chain? What’s the best, most reliable way to measure and quantify I/Q gain/phase error? Finally, how much drift with RF/baseband frequency, temperature, or time should we expect on this part, and is our result typical or unusually high?
