• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 53 subscribers
  • Views 977 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.

CC1350: CC1350 BER-PER

Niels Harthoj
Intellectual 310 points
Part Number: CC1350
Other Parts Discussed in Thread: CC1310, CC2510

Hello

In datasheet for CC1350 sensitivity and other specifications are typically specified with a BER (bit error rate?)  of 10^-2 ~0,01.

For CC1110 similar specs are with a PER (Packet error rate) of 10^-2.

In datasheet for CC1110 PER and BER is related with the following expression:  PER = 1- (1- BER) ^packet _ length. (page 200 in SWRS033G)

How should we compare data for CC1110 and CC1350? Is a sheet with sensitivity specified with PER at a given level , a graph with level vs. PER or other similar information?

Regards

Niels Harthøj

  • 0
    TI__Guru**** 317180 points
    If we recalculated what 20 byte packet/ 1% PER on CC1110 would correspond to on CC1350 using 3 byte it looks like the sensitivity would have been ~-106 dBm if measured with 20 byte / 1 % (50 kbps) => 3- 4 dB difference.
  • 0 in reply to TER
    Intellectual 310 points

    Comparing CC1110 and CC1350:

    CC1110 (38,4kbit GSK, 1% packet error rate, 20 bytes packet length, 20 kHz deviation, 100 kHz digital channel filter bandwidth): sensitivity: -101dBm

    CC1350 (50 kbps, GFSK, 25-kHz deviation, 100-kHz RX bandwidth (same modulation format as IEEE 802.15.4g mandatory mode), BER = 10^-2.(0,01) 868 MHz and 915 MHz) Sensitivity: -110dBm

    Which radio is better? I need some figures for same scenarios to compare performance.

    20 bytes packages, 1% package loss =>8*20*100=16000bits, one bit error. One lost package is around 100ppm BER, depended on spread of errors

  • 0 in reply to Niels Harthoj
    TI__Guru**** 317180 points
    We did a quick test on 6 CC1310 samples for 38,4kbit GSK, 1% packet error rate, 20 bytes packet length, 20 kHz deviation, 100 kHz digital channel filter bandwidth and got -107 dBm
  • 0 in reply to TER
    Intellectual 310 points

    Thanks this looks better than expected.

    To complete specification, for our estimated application, can you give similar data for:

    250 kBaud data rate, GSK, 1% packet error rate, 20 bytes packet length, 127 kHz deviation, 540 kHz digital channel filter bandwidth and

    500 kBaud data rate, MSK, 1% packet error rate, 20 bytes packet length, 812 kHz digital channel filter bandwidth (for 2,4GHz applications).

    This is because we have been working with these data rates until now.

  • 0 in reply to Niels Harthoj
    TI__Guru**** 317180 points
    To do these measurements will require some time (we already had a setup for 38.4 kbps but not the rates you ask about now), what is the potential volume for this project?
  • 0 in reply to TER
    Intellectual 310 points

    I don't know. We are looking at CC1350 for new design and want to have an idea of specification compared to CC1110 and CC2510 which we are familiar with.

    For the 38,4kbaud setup the sensitivity is ~6dB better than CC1110 - according to your latest tests. The other faster bitrates is expected to use less power, but we also want to compare sensitivity to estimate range.

    A good guess, I think, is that these 6dB improvement will be found in the other bit ranges as well, but I find datasheet is not informative about this?

  • 0 in reply to Niels Harthoj
    TI__Guru**** 317180 points
    38.4 kbps was easy to do an extra measurement with since we already had the test case set up but the two other datarates we have not set up in the test system and hence a short (in time) test is not possible. I would assume the sensitivity delta is close to the same for all datarates.