• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 21 subscribers
  • Views 754 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.

LMX2592: VCO tuning sensitivity?

Carl-Friedrich Braun
Prodigy 10 points
Part Number: LMX2592
Other Parts Discussed in Thread: LMX2491

Tool/software:

In our design we control VTUNE ourselves in order to generate a frequency sweep. In contrast to the values shown in table 7-3 of the datasheet we only observe a sensitivity of about 30MHz/V at 6GHz. What could be the reason for this?

  • 0
    TI__Guru** 112380 points

    Hi Carl,

    To measure Kvco accurately, we need to first of all, lock it at the desired VCO frequency without any VCO calibration assist.

    After it is locked, Vtune should be around 1.3V. 

    Then we can change the reference clock frequency slightly, say +/-100Hz, and measure the change in Vtune voltage. 

    Please note, the structure of the VCO is different from a regular standalone VCO. Do not expect VCO frequency will change linearly according to the applied Vtune. 

    For more information on the VCO structure, read https://www.ti.com/lit/pdf/snaa336

  • 0 in reply to Noel Fung
    Prodigy 10 points

    Hi Noel,

    thanks for your answer. In our design we did not connect the charge pump to VTUNE. Instead we use a PWM based triangle wave generator on the VTUNE pin. We disabled the tuning state machine in the LMX2592 and use the override bits to program VCO_SEL, VCO_CAPCTRL etc. When we apply a 0 to 2.5V triangle wave to VTUNE we expected the generated frequency to vary about 70MHz at 3.7GHz (as table 7-3 mentiones a kVCO of 28MHz/V at this frequency). We only see about 40MHz variation, though.

    As we do not use the tuning algorithm of the chip at all we also set VCO_IDAC_OVR = 1 and kept VCO_IDAC at its default value. From the documentation I do not fully understand exactly what VCO_IDAC does. Could you explain further?

  • 0 in reply to Carl-Friedrich Braun
    TI__Guru** 112380 points

    Hi Carl,

    Again, because of the structure of the VCO, Kvco is not linearly proportional to Vtune. Kvco is usually higher at around Vtune = 1.3V and then getting smaller and smaller when Vtune is away from 1.3V. In short, even though we specify Kvco as yMHz/V, frequency will not change 2*yMHz when Vtune changes 2V. 

    If your system requirement is 70MHz linear sweep range at 3.7GHz carrier, I don't think any integrated VCO synthesizer can support. You may need to consider discrete PLL + VCO solution. For example, LMX2491 + external VCO. LMX2491 supports frequency ramp application, we can program it to do linear frequency sweep without the PWM triangle wave generator.