Ethercat random jitter fix

Hello everyone. The machine has only one Ethernet port, dedicated to EtherCAT.

I wanted clarification regarding Hakan's statement:
If any of the slaves have an assignActivate bit 8 set (DC sync), switch on loop synchronization.

The drive's XML file contains these entries:

<OpMpde>
<Name>DC</Name>
<Desc>Dc-Synchron</Desc>
<AssignActive>#300</AssignActive>
<CycletimeSync0 Factor="1">0</CycleTimeSync0>
<ShiftTimeSync0Z0>/ShiftTimeSync0>
<OpModie>

Does this entry mean bit 8? Does the manufacturer declare that the drive supports DC?

So, to conclude.
If I write in the XML file:
refClockSyncCycles=-1 and syncToRefClock="false", DC is enabled, so the slaves synchronize with the master managed by LinuxCNC.

If I write only:
refClockSyncCycles=1 in the XML file, DC is enabled, so the slaves synchronize with the master managed by LinuxCNC.
But I can choose the synchronization frequency from 1-5-1000.

If I write only:
refClockSyncCycles=-1 in the XML file, DC is not enabled, but free mode is enabled, and the drives do not synchronize with the LinuxCNC master.

Correct?

I imagine that the best feature for a CNC machine is DC mode, or am I wrong?

I verified that the lcec.0.dc-phased pin is always false.

refClockSyncCycles=-1 and syncToRefClock="false"

Is identical to 

refClockSyncCycles=1

What synctorefclock false clock does is make -1 become positive, if it was -5 it would be 5

refClockSyncCycles=-1 DC is enabled but master follows the reference slave, master will change its own time
This is identical to syncToRefClock true

About having only one Ethernet, USB to Ethernet adapter could work to give it internet, but I won't insist, but if possible I'd really love to test for myself, as some things get lost in communication 

DC has one start time, that's set on start usually by the first Ethernet DC capable device, from there you have two choices, either the slaves follow the time that the master declares every x cycles with refClockSyncCycles=x

Or everyone follows what the first slaves says the time is, master included. To do that master nudges its own cycle time to match what the first slaves report, the problem with your setup I think, is what the first master is reporting, so this system seems to not work correctly.
This mode is also DC and is set with 
refClockSyncCycles=-1
Or 
syncToRefClock="true"
 

I tried writing in the XML file:
refClockSyncCycles=5, without syncToRefClock, and the machine makes noise.

I also tried:
refClockSyncCycles=1000, without syncToRefClock, and the machine makes noise.

So in my case, the only way to get it to work is:
refClockSyncCycles=1, without syncToRefClock
or
refClockSyncCycles=-1, without syncToRefClock=false

So I'm not using DC mode?

refClockSyncCycles=-1, without syncToRefClock=false

Are you sure this works? I think you stated differently before.

Yes excuse the correct question Is :

So in my case, the only way to get it to work is:
refClockSyncCycles=1, without syncToRefClock
or
refClockSyncCycles=-1, with syncToRefClock=false

So I'm not using DC mode?