• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 51 subscribers
  • Views 1398 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

CC2640: GFSK Demodulator Performance

Antonio Escobar
Prodigy 30 points
Part Number: CC2640
Other Parts Discussed in Thread: CC2420, CC2650

Hello,

I am using the CC2640 for a localization project, in which I need to have an accurate idea of the modulation/demodulation process; since I am interfacing it with BLE transceivers implemented in SDR/FPGAs and I would like to clarify some areas.

My question is  mostly theoretical.

As the CC2640 uses GFSK, and this modulation has different BER vs. EB/N0 curves depending on the demodulation principle, I would like to know if the CC2640 can be considered as a coherent receiver (accurate frequency and phase recovery) or non-coherent (coarse frequency adjustment).

Is there some mechanism for frequency offset compensation, or the devices just "works fine" as long as the transmitter stays within the ppm tolerance specified by the BLE standard?

Thank you,

Antonio

  • 0
    TI__Mastermind 36395 points

    Antonio,

    Our receiver architecture is a non-coherent demodulator. We perform coarse frequency estimation during the preamble and then use a non coherent FSK demodulator to demodulate the burst. We use the preamble to estimate the frequency error as the begin of the burst and we use this information during the decode of the burst.

    We are actually able to demodulate bits from a signal that is several hundreds of PPM offset (>100ppm).

    Regards,
    /TA

  • 0
    TI__Mastermind 36395 points
    Antonio,

    We use the same non coherent demodulator to demodulate MSK also, if you program your 2FSK modulator/demodulator to a modulation index of 0.5 it becomes capable of transmitting/receiving a MSK signal. The catch is that you do not get the SNR advantage of using a coherent signal like MSK, but you able to use it.

    The main reason we chose non-coherent demodulation is ease of use ( greater than 100ppm frequency tolerance) and low power. We make compromises like this to keep low power a key differentiation in our products.

    Regards,
    /TA
  • 0 in reply to TA12012
    Prodigy 30 points

    This was my original follow-up question (sorry, I was editing it while you answered):

    Thank you for your prompt answer and for the PPM offset tolerance data.

    Do you know if that also applies to the classic CC2420 (or newer ZigBee transceivers, like the CC2650) in 2.4GHz 802.15.4 mode?

    I know the OQPSK with half-sine pulse shaping resembles MSK, and can be non-coherently detected, but the CC2420 features a frequency offset compensation module, and the DSP is more complex (DSSS, soft decision demodulator, etc.).

    Is in this case also non-coherent, or that only applies to the FSK-based demodulators?

    ---------------------------------------

    Thank you again, it makes sense!

    I suppose with modern receivers, which feature extensive digital processing, the classical distinction between coherent/non-coherent is not so clear anymore; since even when it is a non-coherent demodulation, we can also say that extra information is used to aid the demodulation process (frequency offset detection during preamble, error feedback, etc.).