• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 22 subscribers
  • Views 1139 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.

Maximum FM Deviation of LMX2571 at VHF Band?

ufukt
Prodigy 220 points

Hi,


I am using this chip in SPI FAST mode to implement FM modulation at VHF (30-300MHz) band with 70-100kHz max frequency deviation. As I have studied in the data sheet, I found out that the maximum frequency deviation can be calculated as,

FDEVmax = fPD*maxStep*Prescaler/(DEN*CHDIV_1*CHDIV_2)

here maxStep is 16 bit signed so it is 2^15-1.  I chosed Prescaler = 4 and fPD = 120MHz (maximum is 130MHz given in the data sheet) to get maximum FDEVmax value. CHDIV_1*CHDIV_2 = 32 for my application. DEN = 2^24 should be set since I am using FSK mode. This will give the maximum frequency deviation about 30kHz. This is the maximum value that this chip supports in the given parameters. Am I right with this calculation? How can provide 70-100kHz maximum frequency deviation in my application ?


Best regards,

  • 0
    TI__Expert 4085 points
    Hi Ufuk,

    Thank you for your interest in our products. We received your question and will be answering it as soon as possible. We appreciate your patience.

    Regards,
    Julian
  • 0 in reply to Julian Di Matteo
    TI__Guru** 107450 points
    Hi Ufuk,
    Your calculation is correct. The suggested DEN is the bottle neck. To implement a higher frequency deviation, a smaller DEN is required. We suggest DEN=2^24 because in some cases, you will not get the desired negative deviation if DEN is not equal to 2^24. In short, you can have DEN not equal to 2^24 but you have to check to ensure that you get the right deviation for every target output frequency.
  • 0 in reply to Noel Fung
    Prodigy 220 points

    Hi Noel,


    Thank you for the reply. I have tried your suggestion in my operation frequency range; but I could't get the desired negative deviation like you said. However, A new idea came into my mind. It seems that I can run the device in PLL mode with external VCO working at lower frequencies about 500MHz. Hence,  the equation becomes,

    FDEVmax = fPD*maxStep*Prescaler/(DEN*CHDIV_3)

    Since my new VCO operates at lower frequencies I will set CHDIV_3 value as 3 or 4 so that FDEVmax will be 8-10 times higher than its previous value when CHDIV_1*CHDIV_2  was 32.

    Am I right?

    Best regards,

  • 0 in reply to ufukt
    TI__Guru** 107450 points
    Hi Ufuk,
    In reality this would not work because you cannot put fpd too high with external VCO. For example, if VCO=500MHz and fpd=120MHz, then N-counter becomes 4.16, which is too small to support 2nd or higher order delta sigma modulator.
    maybe you can try again with the internal VCO, use DEN not equal to 2^24, try a fractional channel (N = a fractional number), it should works. If your application falls into an integer channel, change the fpd to a fraction or non-integer multiple of VCO freq.
  • 0 in reply to Noel Fung
    Prodigy 220 points

    Hi Noel,

    I had used an integer channel that's why I couldn't get negative deviation. As you said, now I have used a fractional channel and got desired negative deviation. I figured out that if I used DEN not equal to 2^24 (for example 2^23) and use fractional channel with very small NUM value like 1000 negative swing does not work properly. That is, It swings negatively but only some point very close to the center frequency. However, if I set NUM to bigger values it works. For DEN = 2^24, it works even at integer channels. So I think, I have to be careful about choosing NUM values and setting a fractional channel in order to get the desired frequency deviation at the desired channel. But I don't know how to choose the fractional values for Ndivider. Since for smaller NUM values, it fails. Can I find a minimum NUM value for a spesific DEN value? For example, when I use DEN = 2^23 what should be the minimum NUM value to be used? Is there another mathematical relation to play with to choose desired fractional value other than the ones in the datasheet ?

    Thanks for your support.


    My best,

  • 0 in reply to ufukt
    TI__Guru** 107450 points
    Hi Ufuk,
    I don't have an answer for this, let me check and get back to you later.
  • 0 in reply to Noel Fung
    Prodigy 220 points
    Hi Noel,

    You were going to get back to me about my last question on this topic.

    By the way, I have found out that, If I choose NUM values bigger than the maximum step value to get the maximum frequency deviation, it works. So I think the mathematical relation about choosing the right NUM value is that you have to make sure that NUM > maxStep so that it can swing to the negative frequencies up to the maximum negative deviation. Also make sure that NUM < maxNUm - maxStep so that it can swing to the positive frequencies up to the maximum positive deviation.

    Am I right ?

    Thanks,

    Best,