Update oversampling example with working code

[jamwaffles]

See #368 (comment)

Update examples/oversampling.rs with new SYNC settings, and verify with a couple of oversampling devices.

[kraemr]

So i tried to do oversampling with the EL4732, very similiarly to the example in here: #368 (comment) .It seems like the sync0_period is used for calculation of next_cycle in tx_rx_dc. If i understand correctly In Dc-Sync Sync0 is not always equal to Cycle Time and especially when using Oversampling output terminals Sync0 < Cycle time because the Sync0 determines the next "pulse" of the oversampling terminal. Also for ref Beckhoff shows here, how the sync0 and sync1 compare to cycle time etc: https://infosys.beckhoff.com/content/1031/ethercatsystem/Images/gif/2469765643__Web.gif I was just wondering if you are already aware of this issue with oversampling? Should I open another issue?

[jamwaffles]

Thanks for commenting, I appreciate it! I'm aware that the current oversampling setup is pretty broken, but I need to dig into it to figure out why. I think I misunderstood fundamentally how it works when I was implementing it the first time.

[plomtb]

Since this is an interesting/important feature for us, I could try to help. I'm not sure if I'm able to contribute code yet, but I can test the EL4732 if you have a version of it available. If you have any other ideas for how I could support the effort, please PM me, @jamwaffles.

[kraemr]

I think this could be related: #308 The EL4732 only has 32bit dc and every like 4ish seconds the sine-wave seems to collapse even though my round trip times are fine.

[kraemr]

Ok so i got it working. I had to change some stuff in subdevice_group/mod.rs specifically the first_pulse_delay where the sync1_period is used instead of sync0_period for both sync types. And the timing for the frame is now just cycle_time +- some offset to continually align the dc clock. though the timing is not yet in the my fork im also not 1000% sure its correct yet. need to verify with DC Time difference register. Im going to be adding the timings to the fork soon in tx_rx_dc, for this i will add a cycle_time config for groups or the maindevice, with which the next_cycle_wait is calculated. Here is the current state of my fork if you are interested: https://github.com/kraemr/ethercrab