Independent control of QTR delays

[taganrog]

Hello

Can I increment NQTR delay independently, without incrementing PQTR delay?

[sandrao]

Physically they are two separate lines with separate RIU registers so I think its possible though I’m not sure I understand the use case. Why do you want to move the P & N QTR delays away from each other?

[taganrog]

Data valid windows can be different. And clock edges can has different positions in the windows. So one QTR delay can stop during calibration, when the other QTR delay can continue to move.

PG150: "The data captured in the PQTR clock domain is transferred to the NQTR clock domain before being sent to the read FIFO and to the general interconnect clock domain. Due to this transfer of clock domains, the PQTR and NQTR clocks must be roughly 180° out of phase. This relationship between the PQTR/NQTR clock paths is set up as part of the BISC start-up routine, and thus calibration needs to maintain this relationship as part of the training (BISC_ALIGN_PQTR, BISC_ALIGN_NQTR, BISC_PQTR, BISC_NQTR)."

Thus, how far I can move apart P & N QTR delays from each other?

[sandrao · Accepted solution]

I doubled checked and they can be independently controlled. They’re two different registers with different controls, the delays are 512 taps so you could move them apart by as much as the delay line allows.

[taganrog]

Do I need to maintain any relationship between the PQTR/NQTR clock paths as pointed in PG150?