• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 8 subscribers
  • Views 3707 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.

Pros/Cons of preamble and sync word lengths.

Dave Delage
Intellectual 480 points
Other Parts Discussed in Thread: CC2500

There are many possible combinations of preamble and sync word lengths but little information about why you would adjust them up or down.

For example, I want to send:

4 bytes preamble, 2 bytes sync, 1 byte address, 6 bytes data for a total of 13 bytes or 104 bits at a data rate of 500kbits/sec. BUT, I'd really like to do it 5 times per second which would require getting rid of one byte. Which byte?

My guess is that given the high duty cycle (nearly 100%) I could eliminate one of the preamble bytes since the CC2500 receive circuit would already be preset from the previous transmission. But that's a guess. Does anyone know?

BTW, this would operate in the ISM band in an area with little other radio and outside both WiFi and Zigbee channels for a few hours a day so the duty cycle is not an issue. And, ideally, the data can not be delayed such as sending 12 bytes every other time.

  • 0
    TI__Guru**** 317180 points

    As written in the datasheet, 8 byte preamble and 4 byte sync is recommended for this datarate (page 28 in http://www.ti.com/lit/ds/symlink/cc2500.pdf). With 2 byte sync you will get a lot of false sync do to noise.

  • 0 in reply to TER
    Intellectual 480 points

    The above answer certainly corrects my error in misinterpreting the register value. Indeed SmartRF sets the register to 8 preamble bytes. But "you will get a lot of false sync due to noise." does not answer the fundamental question of why are the number of preamble bytes and sync bytes sent adjustable? Why would you use more of one than the other. In most cases there is a large amount of time from one transmission to another so I can see the need for lots of preamble bytes but how many? In my case the time between the end of one transmission and the beginning of the next is vanishingly small. Does the system really still need 8 preambles? Are there no charts/formulas or general guidance beyond SmartRF says so?

  • 0 in reply to Dave Delage
    TI__Guru**** 317180 points

    - Sync word: This depends if the default sync word is used or something else and the system requirements. The default sync word has a lot of transitions and the auto correlation is coog. Basically the sync word is good if plotting the auto correlation does not give any side lobes. It is up to the user if a 16 or 32 bit sync word is used but with a short sync word the probability for noise to be equal to the sync word is larger. For a 16 bit sync on CC2500 a false sync will be found within minutes or even less generating a lot of extra work for the MCU.

    - Preamble: The amount of preamble is set from sending one packet. If you send multiple packets back to back (no break between them) you don't need that long preamble since all the regulation loops have settled. But if it's a break between the packets you will need a long preamble to settle the receiver (bit and byte slicer, frequency track) since between the packets the receiver will start tracking noise.

    The recommended preamble length is found partly through lab work since the needed amout will vary some dependent on the datarate, deviation.

  • 0 in reply to TER
    Intellectual 480 points

    Thank you.

    From this "But if it's a break between the packets you will need a long preamble to settle the receiver (bit and byte slicer, frequency track) since between the packets the receiver will start tracking noise." can I take it that there is no equivalent of "squelch" that would cause the receiver to "hold" the current values in the tracking circuits and so in a matter of microseconds the previous values are lost?