• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 47 subscribers
  • Views 572 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LO phase difference on TSW30SH84EVM RF ouputs

Martin Mataln
Intellectual 295 points
Other Parts Discussed in Thread: TSW30H84EVM, DAC34H84, TRF3705

Hello!

I recognized a frequency depending phase shift between the two LO signals on the output of the two RF paths of the TSW30H84EVM (see figure).

I used a constant on the DAC34H84 output to see only the behavior of the two TRF3705. I have used equal length very short lines for two LO input and RF output connections to measure the phase difference. I have used my own splitter (not that one installed on the TSW30H84EVM). The dynamic range and every other parameter of the device is very excellent and meets my specifications.

For measuring my Doherty PA the best would be that the LO phase difference between the two channels is nearly zero! I will compensate this in my FPGA design but the phase difference is very large and also frequency depending. Is this a normal variance between two TRF3705 modulator implementations or is this coming from the output matching of the TSW30H84EVM?

Greetings!

  • 0
    TI__Guru 71679 points

    Martin,

    Good to hear that the EVM performance meets your requirement. Due to the space contraint of the EVM, we cannot match the two LO traces to the TRF3705s. The upper channel AB LO trace is about 1720mils, while the lower channel CD LO trace is about 2120mils. There is about 400mils of trace difference. I suspect this is the main cause of the RF output phase difference.

    The phase difference that you are seeing is very linear. Assuming constant trace delay across frequency, the delay of the trace is about 280ps/inch, or 112ps per 400mil. (i.e. phase delay = 360*delay*frequency). I will ask our board shop to verify this delay on our stack-up and board material. I will also ask our RF team to comment on the potential phase difference among the TRF3705 modulators. 

    -KH

  • 0 in reply to Kang Hsia
    TI__Guru 71679 points

    Martin,

    The dielectric material for the TSW30H84 EVM is 370HR with permittivity (DK) of 4.4. The estimated propagation delay is about 180ps/inch. There is definitely phase difference caused by the LO trace length difference of channel A and B. The additional delay could be introduced by your external circuit as well (i.e. phase difference between the splitter output, cable length, etc.)

    -KH

  • 0
    Intellectual 570 points

    Hi Martin,

    Your post had been very insightful for my system too. As you already measured it, I am just wondering, if this phase shift constant, which I might compensate before sending modulated signals.

    Please correct me, if I am wrong, if LOs have this phase difference, then modulated signals will suffer from same phase difference.

    Looking forward to your comment,

    Regards,

    Meenakshi

  • 0 in reply to Meenakshi Rawat
    Intellectual 295 points

    Thats right! 

    The phase difference within the LO's is acting like a phase shift of the original un-modulated baseband signal. it can be easily corrected by applying a phase offset to one of the two baseband signals.

    greetings!