STM32G4 HRTIM operations of push-pull mode

[iy]

I have two questions regarding RM0440 Rev.8, Figure 195:

1. In Figure 195, it shows that the roll-over events invert the push-pull logic. A roll-over event is defined as "a counter roll-over event is generated when the counter goes back to 0 after reaching the period value set in the HRTIM_PERxR register in continuous mode," which does not include the reset event.

   However, after verifying, I found that both roll-over and reset events toggle the CPPSTAT bit.

   • Does the reset event also invert the push-pull logic?
   • What is the relationship between the push-pull logic output and the CPPSTAT bit?

2. How should Figure 195 be interpreted?

   • When the output of the push-pull logic is 0, the top-left mux selects Out 1 (from the crossbar) due to the inverted push-pull logic output, and the bottom-left mux selects Out 2 (from the crossbar). This setup seems inconsistent.
   • When the output of the push-pull logic is 1, both left-side muxes select "inactive," which also appears incorrect.

Could anyone help clarify the correct behavior of the push-pull logic and the output selection in Figure 195?

[liaifat85]

The CPPSTAT bit reflects the current state of the push-pull logic, essentially toggling with each roll-over (and, as you've observed, each reset) event. In Figure 195, this bit is part of the mechanism controlling the output selection logic. When CPPSTAT toggles, it switches between outputs connected to the push-pull logic, thus alternating which output channel is active.

In practical terms:

• When CPPSTAT is 0, one set of output pins or channels is activated.
• When CPPSTAT is 1, the other set is activated.

[iy]

Thanks for your reply, could you explain more about the Fig. 195?

As my understanding from this figure, when the output of push-pull logic is 0, M1 output "Out 1 (from crossbar)" signal due to its select signal in inverted, and M2's is "Out 2 (from crossbar) signal". Both signals are selected! 
Vice versa, both outputs of M1 and M2 are inactive when the output of push-pull logic is 1.

Is something wrong in the figure or I lost anything?