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.

LMX2582: No RF Output

Part Number: LMX2582

I am using LMX2582 to generate a single frequency. I have followed the schematic as given in the Page 36 (Sec. 8.2) of the datasheet and I am programming LMX2582 using microcontroller. But there is no RF output (even a default output ) is not coming (free running frequency) . Does LMX2582 need to be programmed even for a default output? I have programmed it using the configuration settings as given in the forum (Previously I used settings that I generated from TICS Pro):


1) Write R0 = 0x002216; (Soft reset the device, R0 [1] = 1, other bits set to default)

2) write other registers;

3) Write R0 = 0x00221C; (Frequency calibrate, R0 [1] = 0, R0 [3] = 1, other bits are set to default)

e2e.ti.com/.../588825

I am unable to know if my programming is wrong or is the LMX2582 not soldered properly. I want to know if LMX2582 without programming and after power reset, should generate a default frequency (VCO free running frequency) or not.


Thanks in Advance,


Regards

Alok

  • Alok,

    Before worrying about the output, the first thing to see if the device is powered up at all. You should be able to see this in the current consumption. Also, the OSCin pin has a bias level of about 1.5V when powered up, 0V when powered down. If the device is not powered up, no point in worrying about the output frequency.

    Next step is to see if response to SPI commands. Power it up and down with the POWERDOWN bit.

    If you can program the device, but it is not locking, then there are many causes. Make sure the CE pin is high. Make sure you are programming ALL the registers, even the ones with 1's and 0's only in the datasheet. Ensure you have a valid input reference when you do step 3). Make sure that the output is powered up and you have a pull-up component.

    Regards,
    Dean
  • Hi Dean,

    Thank you for your valuable advice. I was able to send SPI commands. To test this I made RF output ON and OFF using R0[0]. Further, I was able to change output frequency by changing PLL_R value.

    0x0B0F28 - 1.15 GHz

    0x0B0028 - 2.43 GHz

    0x0B0328 - 1.38 GHz

    But when Iam changing PLL_N value output is not changing.

    In the Loop Filter as mentioned in the datasheet (Page 36), Iam using C1 = 3.7 nF, C2=100 nF and R5=51 Ohm. Ia m not using R3_LF, R4_LF, C3_LF and C4_LF.

    Is it that PLL is unable to track frequncy with the above loop filter settings ?

    Thanks and Regards

    Alok

  • Alok,

    OK, so now it sounds like you can program the device, but you can not get it to lock to the right frequency. Here's a few more things:

    1. Realize that you need to program all the registers to get the device to work.
    Also, after the N divider value is changed, you need to program the R0 register with FCAL_EN=1 to activate the VCO recalibration.

    2. I would test the charge pump. For this, look at tuning voltage at the Vtune pin. Now program the PLL_N value to something really low, like PLL_N=3. Then program it to something really big, like PLL_N = 30000. Hopefully you can make this jump. If not, it suggests the charge pump is not connected to Vtune or perhaps the charge pump current is tri-stated.

    3. For more robust programming, set register R0 with RESET=1, then regsiters R0 with RESET=0, and then load all registers in reverse order with the last being register R0 with FCAL_EN=1. Then write register R0 with FCAL_EN=1 again about 10 ms later.

    Regards,
    Dean
  • Hi Dean,

    1) I programmed all the registers

    2) I measured Vtune by changing PLL_N = 3 to PLL_N = 4095 (max. value with 12 bits) , still Vtune was not changing. I checked the continuity between CPout and Vtune Pin. They are connected. I changed the R14 register to change Charge Pump Currents . I had the following results

    Command, 0x0E0000 => CP_IDN = 0, CP_IUP = 0, Vtune = 0.3V

    Command, 0x0E018C => CP_IDN = 3, CP_IUP = 3, Vtune = 2.4V

    Command, 0x0E0084 => CP_IDN = 1, CP_IUP = 1, Vtune = 2.4V

    3) I have followed the sequence of resetting first then loading the registers and programming Higher Address Registers first . I also calibrated after writing PLL_N value, but still frequency is not changing.

  • Alok,

    From 2), it seems you are programming the device and the charge pump is connected the the VCO input.   However, realize two things:

    1.  The max charge pump output is 2.5 V as it is supplied by a 2.5 V regulator.  You are measuring 2.4 V, which is very close to the rail, maybe we could consider this to be on the rail of the charge pump.

    2.   The VCO has a negative tuning coefficient.  So it looks like it is trying to force the frequency lower.

    My first thought is that if the device perceived that there was no input reference at the Fosc pin, then it would try to force a lower output frequency and never get there, showing similar symptoms to what you are seeing.  Ist this possible?

    Regards,

    Dean

  • Hi Dean,

    I changed PLL_R value and the output frequency changes. This wouldn't take place if device doesn't perceive Reference signal. That's what I think. Kindly correct me if I am missing anything.

    Thanks and Regards

    Alok

  • Hi Dean,

    Thank you Dean for giving insight into the Charge Pump working and also the Analog wire post e2e.ti.com/.../what-to-do-when-your-pll-does-not-lock.

    "My first thought is that if the device perceived that there was no input reference at the Fosc pin, then it would try to force a lower output frequency and never get there, showing similar symptoms to what you are seeing.  Ist this possible?"

    The above comment helped me a lot in examining my register configurations in TICSPro. I found that with the settings I had I was generating frequency in order of hundreds of GHz (VCO freq was highlighted as red in TICSPro). Then I made changes to N value properly to bring VCO frequency to its operating range. Then I made little changes in PLL_N value and the output was moving a bit...I think. I will confirm that tomorrow. But the frequency I was generating was not as per TICSPro. All the details , I will post tomorrow .

    And thanks Dean for all the prompt responses you have made.

    Thanks and Regards

    Alok

  • Alok,

    OK, so it sounds that perhaps the N divider ratio was set wrong and hopefully the correct value will fix the issue.


    Regards,

    Dean

  • Hi Dean,

    Thank you for all your valuable advices. My problem is resolved now. My output is matching with that calculated from TICSPro, I had reading error.

    Thanks and Regards,

    Alok

  • Alok,

    Great!  I am glad we were able to resolve this.


    Regards,

    Dean