• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 2 answers
  • Subscribers 18 subscribers
  • Views 2927 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Manchester encoding necessary for reliable CC1101 reception?

Payton
Prodigy 190 points
Other Parts Discussed in Thread: CC1101, CC1190, CC1120

How necessary is Manchester encoding needed to achieve the BER and sensitivity stated? 

Which would yield better link margin:  Manchester encode data (as typical) or use NRZ+data whiten+FEC+interleave, while maintaining the same chip rate?

With raw data, there was certainly need to recover the clock throughout the packet and maintain a good moving average for the data slicer.  But, with the well synchronized newer receivers now, is the advantage of Manchester diminished?  Datasheets often now refer to sensitivity for raw one-chip-per-bit.

  • 0
    TI__Mastermind 36385 points

    Payton, 

    Manchester is generally used only in combination with ASK/OOK modulation as it helps the AGC circuit by ensuring there is always the same amount of power. On the other hand data whitening is commonly used for FSK type applications to ensure that you do not have long sequences of all "0"s or all "1"'s.

    We use pseudo random data sequences in our labs to measure the performance of the system.

    Regards,
    Thomas 

  • 0 in reply to TA12012
    Prodigy 190 points

    Hi Thomas,

    Thanks for your response!

    So for FSK modulation, there's really no/little need for Manchester, even if there are potentially long sequences of 0's and 1's.  Otherwise, any difference expected in the receiver (e.g. CC1101) FSK sensitivity if there are long sequences?

    In my application, I will likely use data whitening of some sort, but it's helpful to better understand the limitations.  Being able to realize better sensitivity and reduced bandwidth by going to lower chip rates is very enticing.  Hopefully my questions aren't too general - was hoping to glean better insight into the CC1101 advantage with its packet handling capability and accurate timing.

    Best regards,

    Payton

  • 0 in reply to Payton
    TI__Mastermind 36385 points

    Payton, 

    I recommend using the "data whitening" option in the CC1101. It has no overhead and removed long sequences of 1 and 0's. The lower the data rate the better sensitivity you will get, however at the lowest settings and lowest RX bandwidth you will need to make sure you RF crystal reference are good enough to support the narrow band operation. Perform a quick calculation of worst case drift using 40ppm crystals versus 20ppm and 15ppm, you will find that to operated at 58kHz RX bandwidth (the most narrow for the CC1101) you need good crystals in the order of 10ppm.

    If you want more range consider using the CC1190 (its a range extender) it will amplify you output signal and also amplify the received signal.

    PS: I also recommend staying with settings from SmartRF Studio or at least start with them and use some trial and error to find you own preferred settings.

    Regards,
    Thomas 

  • 0 in reply to TA12012
    Prodigy 190 points

    Hi Thomas,

    Thanks for your recommendation.  

    Understood regarding the lower limit on Rx BW effect of crystal tolerance.

    Data-whitening sounds like the very clear choice over Manchester when it comes a CC1101 driven system.  Reading up on the datasheet, it says that for Tx packet, the data whitening is performed before optional FEC/interleaving.  Doesn't this potentially undo some data white-ness?

    Is there a recommended range extender for the lower frequency bands (i.e. < 450MHz)?  The see the CC1190 was meant for the 800-900MHz range.  I think I saw a similar question on another post recently.  Familiar with other methods to put an LNA on the Rx end, but haven't come across any suitable Tx output power amps yet.  

    Best regards,

    Payton

  • 0 in reply to Payton
    TI__Mastermind 36385 points

    Payton, 

    Data whitening is just XOR of a pseudo random sequence with the incoming data, this XOR is also performed on the RX side. So it becomes transparent the MCU subsystem. There is no significant overhead.

    For Front end modules we do not have Front end modules operating in any other range than 868-930MHz. We have partnered with both Skyworks and RFMD to supply our customers with solutions where we do not provide solutions.

    Skyworks: http://www.skyworksinc.com/uploads/documents/201060E.pdf

    RFMD:       http://www.rfmd.com/CS/Documents/RF6504DS.pdf

    Regards
    /TA 

  • 0 in reply to TA12012
    Intellectual 845 points

    Thank you so much for this thread. Just this morning, I am having some intermittent communications with my board (lots of data). after implementing data whitening, the communication is much more reliable.

  • 0 in reply to TA12012
    Prodigy 100 points

    Hi Thomas,

    I've just read this topic and I would like to ask you if this argument should be considered as valid also speaking about the CC1120 Transceiver.

    I'm considering if use or not the Manchester encoding in an example application .The modulation that I would like to use is the 2-GFSK and the data rate around 38.4 ksps (with conseguent RX BW filter of 100kHz)... in your experience, do not use Manchester encoding but using data whitening in these conditions can guarantee a reliable communication? Or it is better to use Manchester encoding (with conseguent increasing of bandwith or diminshing of data rate)?

    In this way the crystal should be excellent?

    Many Thanks for your reply and best regards.

  • 0 in reply to Mat
    TI__Mastermind 36385 points

    Mat,

    I suggest whitening all types of communication where the payload can have long sequences of 0's or 1's. This helps the AGC for ASK modulation and helps the automatic frequency tracking for FSK modulation.

    Most people use Manchester for ASK, but for FSK using data whitening is sufficient. Manchester doubles the data over the air or effectively reduces the bit rate capability of the transceiver. 

    Regards,
    /TA