The output of the SPDIF iPurifier is direct-coupled for the precise reasons the customer states. If no isolation is needed this is best.
Further, however, in any competent design one cannot simply bypass or exchange pulse transformers and expect correct or improved performance. In a fully optimised design (such as we employ) additional complex conjugate pole and/or zero circuits are added to maximise transformer bandwidth and return loss across the widest possible frequency range and must critically tuned to the actual transformer used.
As a further note, pulse transformers that perform best for SPDIF generally use bifilar (aka. high capacitance) construction on high-frequency ferrite cores of small size. Using amorphous iron or permalloy cores (or other “exotic” metals) in SPDIF transformers (as some manufacturers promote) or a shield between windings (as some other manufacturers do) is not conductive to good performance but makes for good marketing spiel that allows high prices to be charged for very limited performance and those transformers are best avoided.
In AMR & iFi products with isolated SPDIF inputs, we use a specific wideband pulse transformer fully tuned with complex conjugate pole and/or zero circuits for a very wide bandwidth and maximum return loss for the widest bandwidth while providing isolation on the input. This is followed by a linear mode, zero-feedback active circuit (not logic circuitry) using either tubes (AMD DP-777) or Fets (iFi) with a designed-in “clipper” circuit that regenerates a near-perfect Logic waveform even from badly degraded SPDIF signals.
It is this regenerated signal that is fed to the rest of the circuitry. So the presence of galvanic isolation is not an issue with AMR or iFi products.
More on this fundamental technology here:
https://amr-audio.co.uk/tech-notes/#1539092351973-29777ab8-bf2c
