This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

evm 5515 with c2500 transceiver - fifo threshold

I've configured  the GDO register to assert when  sync byte is received/transmit (0x06) .

My packets length are  variable (max 61 bytes)

what should be the value of the fifo threshold  ? does it have any affect on the GDO signal (with the above configuration)?

BR

Talmor

  • The fifo threshold has no effect on the GDO signal when configured for sync byte received/transmitted.

    The fifo threshold value you need will depend on some parameters of you system in a deployed environment.  If you can guarantee the fifo will never overflow (because of an extra packet being sent) then it's value really doesn't matter.  If you are in a noisy environment where your packets are tossed due to invalid crc checks then you would probably want to set the fifo threshold to the length of your smallest packet to minimize being interrupted unnecessarily.  If you are in a high traffic environment, you may want to set the value low so you can keep the fifo as empty as possible.  Also the value chosen must take into account the latency of responding to the signal which is dependent upon the code architecture, low power modes, etc.

    In essence, this is an input to your design parameters which will drive your overall performance.  In many cases it is not very easy to quantify and therefore is set to some intermediary value for design and development.  Once the project is fairly mature, start modifying the value while measuring some performance characteristic to optimize it.  As long as the environment used during optimization reflects the deployed environment this should work well.

    Jim Noxon

     

  • Thanks.

    so I assume I can leave the fifo threshold as default. (0x07) right ? even if I have packets smaller than 33  

  • For a maximum packet size of 61 this seems acceptable as long as this value includes any automatically appended LQI metrics as well.  I believe there are issues with the read fifo if it fills to a full 64 bytes thus 63 is really the maximum.  Check the erratum for the device to be sure.

    Jim Noxon