Autopilot overshoots turns with TDS GTNXi

[Pixel of Life]

When using the GTN 750 as the autopilot nav source, it overshoots even shallow turns and has to correct to get back on course. As far as I can remember, the UNS-1 doesn't do this.

Not sure if this is an issue with the plane or with the GTN, so I'm asking here first.

[Cameron]

This is accurate behavior, as the GTN 750 simply feeds heading info to the autopilot when reaching a WPT to change direction in the configuration setup the 525 is in. The UNS calculates an actual turn.

[Pixel of Life]

Thanks. How would one handle this in real life when piloting a C525 with this setup? Is it acceptable to use the GTN to drive the autopilot even though it doesn't follow the route exactly, or should it be used for monitoring only while the UNS steers the plane? Presumably the GTN isn't certified for certain procedures because it's not precise enough?

Edited July 29 by Pixel of Life

[Pils]

44 minutes ago, Pixel of Life said:

Thanks. How would one handle this in real life when piloting a C525 with this setup? Is it acceptable to use the GTN to drive the autopilot even though it doesn't follow the route exactly, or should it be used for monitoring only while the UNS steers the plane? Presumably the GTN isn't certified for certain procedures because it's not precise enough?

IRL you'd have the GTN replace the UNS as the "FMS" input on the digital autopilot. Depending on the airspace, the required navigation precision is probably going to be +/- 1nm or 5nm, so a little deviation/navigator imprecision is accounted for, see: https://skybrary.aero/articles/performance-based-navigation-pbn