About latency if we do not use each internal blocks (PREMOD, HB1, HB2, HB3, NCO and IQ Modulation)

Hello,

Latency specs are shown below for the different modes of operation.  Note that 1x interpolation is bypass mode.

Hello PMH2023-san,

Thank you for your reply.

Please tell me the latency of all functions are same latency?

For example, 

PREMOD bypass: 64 cycles

HB1, 2 and 3 bypass: 64 cycles x 3 = 192 cycles 

NCO bypass: 64 cycles

IQ Modulation bypass: 64 cycles

So, total latency is 64 x 6 = 384 cycles

Is that correct?

Best Regards,

Kakehi

Hello,

With regard to latency specification, it would appear that latency is per function.  Note, the inclusion of course modulation does NOT add to latency due to the manner in which it is implemented (i.e. modify filter coefficient polarity vs actual NCO).  The fine NCO is implemented  with multipliers hence the added 12 cycles.   Also, the pre-mod function does not add to latency since this does not require digital multipliers since data polarity is only being affected (i.e.multiply by +1 or -1).  

With regard to your example which pertains only to 1x (bypass) mode...........latency is 64 cycles.  If you were to operate in 8x interpolation mode (no inv Sinc or fine mod).............the latency would be 951 cycles (=64+125+254+508) regardless if course modulation via filter was enabled or not.

Hello PMH2023-san,

Thank you for your answer.

Best Regards,

Kakehi