• State Resolved
  • Locked Locked
  • Replies 11 replies
  • Subscribers 17 subscribers
  • Views 1199 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.

Error in Design note DN005 of Texas Instruments

Ricardo Medina
Prodigy 185 points
Other Parts Discussed in Thread: CC1101

Dear Sirs of Texas Instruments,

I have detected  errors in table 1 of Design note DN005, page 5 of 11  (revision 2009.03.13)

There are some configurations with Rx filter bandwidth smaller than total frequency offset. This configuration produces communication errors

Please, correct DN005 as soon as possible, and complete table with GSFK as well.

Best wishes,

Ricardo

  • 0
    TI__Mastermind 39860 points

    DN005 is correct and table 1 refers to plots of real measurements. 

    What you should note is the text under Table 1, which states that FOCCFG.FOC_LIMIT[1:0] = 11b. That is, the CC1101 built-in frequency compensation algorithm is enabled. With FOCCFG.FOC_LIMIT[1:0] = 11b  the RX filter BW is effectively programmed RX BW +/- RX BW/2 (noise BW is still RX BW though).

    Refer to Section 14.1 in the CC1101 data sheet and the FOCCFG regsiter description.

  • 0 in reply to Sverre
    Prodigy 185 points

    Thanks for clarify this.

    I have problems of communications with CC1110 between different lot productions.

    CM020G 1149 CC1110 Mark cannot communicate with boards with 28M2579 1231 mark, with default configuration: Data rate: 38.4 kbps, GFSK, 20 kHz deviation, no DC filter. Optimized for sensitivity, RX filter bandwidth: 101Khz. The others components are the same:

    C1=C2=15pF 5%

    X1: NX3225SA-26.000000MHZ-B1, tolerance: 10ppm, Cload: 10pF

    The frequency that microcontroller generates is the same:

     

    Sample: 1 (1CM020G 1149)

    Sample: 2 (28M2579 1231)

    Sample: 3 (1CM020G 1149)

    868[Mhz]

     868,299866

    868.28

    868.28

    868.28

    Difference f [KHz]

    19.86

    19.86

    19.86

    915[Mhz]

    914,999969

    914.992

    914.992

    914.992

    Difference f [KHz]

    7

    7

    7

    To solve it I have to increase RX filter bandwidth from 101Khz to 135Khz, I do not know the reason.

    According to the following configuration “38.4 kbps, GFSK, 20 kHz deviation, no DC filter. Optimized for sensitivity”, and considering crystal tolerance, data rate and that channel spacing is 199.951172Khz. What is the recommended value to RX bandwidth filter?

  • 0 in reply to Sverre
    Prodigy 185 points

    I have test again changing from  FOCCFG.FOC_LIMIT[1:0] = 10b to  FOCCFG.FOC_LIMIT[1:0] = 11b, and the result is the same. Modules cannot communicate each other.

    I have to change RX filter from 101Khz to 135Khz. Please, why it happen? It is crytical for us

  • 0 in reply to Ricardo Medina
    TI__Guru**** 317180 points

    - You state that you can't communicate with chips from different batches. Does that mean that sample 1 and sample 3 does communicate?

    - Difference f in the table is that the measured delta from expected RF frequency?

    - If the dleta is nearly 20kHz it looks like the parasitics are a bit from what's normally expected. Have you tried to adjust the xtal loadcap to get smaller delta?

  • 0 in reply to TER
    Prodigy 185 points

    Yes, devices with different batches cannot communicate.

    Yes, We used Smart RF Studio configuring TX continuous mode and measure RF frequency and compare with expected.

    I would like to know the signal BW for my configuration in order to select correct filter bandwidth: 

    Data rate: 38.4 kbps, GFSK, 20 kHz deviation, no DC filter. Optimized for sensitivity

    Could you help us?

    Best wishes

  • 0 in reply to Ricardo Medina
    TI__Guru**** 317180 points

    In your table I notice that sample 1, 2 and 3 has the same delta from expected frequency. You say that sample 1 and 2 does not communicate, but does sample 1 and 3? As far as I can see from the lot codes they are from the same lot.

  • 0 in reply to Ricardo Medina
    TI__Mastermind 39860 points

    SmartRF Studio recommends a 100 kHz RX BW when using 38.4 kbps, GFSK, +/-20 kHz deviation. All the "typical settings" recommended by Studio have been thoroughly tested when characterizing CC1101 so we know 100 kHz RX BW is ok to use. Apparently, 100 kHz RX BW does not work in your application and you need to increase the RX BW to 135 kHz to set up an RF link (theoretical degradation in sensitivity of 10*log(135/100) = 1.3 dB).  Assuming you use the register settings recommended by Studio and FSCTRL0.FREQOFF = 0, the only logical explanation  I can think of is that there is a significant frequency offset between transmitter and receiver, but according to your post this is not the case. 

  • 0 in reply to TER
    Prodigy 185 points

    28M2579 1231 with 28M2579 1231 cannot communicate each other using default configuration of Smart RF Studio; 

    Data rate: 38.4 kbps, GFSK, 20 kHz deviation, no DC filter. Optimized for sensitivity

    To solve it I have to increase RX filter from 101Khz to 270Khz, and I do not know the reason.

    I would like to know RX filter bandwidth properly for my configuration. Can you inform about this?

  • 0 in reply to Sverre
    Prodigy 185 points

    I can see in TX continuous mode that frequency is the same in all devices, so I do not understand why devices cannot communicate each other.

    Can I have probles if we increase RX filter?? Channel spacing is close to 200Khz and if RX filter bandwidth is 270Khz, I think that we might have problems with channels. Can you advice us?

  • 0 in reply to Ricardo Medina
    TI__Mastermind 39860 points

    Increasing RX BW degrades sensitivity and selectivity. Changing RX BW from BW1 to BW2 degrades sensitivity by 10*log(BW2/BW1).

    For 200 kHz channel spacing you should use an RX BW of 200 kHz maximum.

  • 0 in reply to Sverre
    TI__Mastermind 39860 points

    If you have to change the RX BW from 101 kHz to 270 kHz to set up an RF link there is something fundamentally wrong with your set-up. Crystal inaccuracies is not the root cause if you need to increase the BW by approx 170 kHz.

    If you have a channelized system make sure you use the same CHANNR for the transmitter and receiver.