Post Go back to editing

LO Bleed-through with ADL5375


I have a single ended signals transmitted from an AC coupled DAC. The DAC has 4 outputs and I am using 2 of the outputs as I and Q (90 degrees offset).  In order to drive the ADL5375-15 modulator with a DC biased differential input I am using  2 x DC1147 (LTC6404) differential amplifiers. I trimmed the DC bias as follows:

  • QBBP - 1.499 V
  • QBBN - 1.501 V
  • IBBN - 1.502 V
  • IBBP - 1.495 V

DC1147: I have also replaced all of the differential gain resistors with 0.1% resistors to attempt to balance the outputs without individual trimming on the DC1147. The baluns have been been removed from both the input and output and replaced with zero ohm resistors, along with the output resistors.  The negative input is open circuit and positive input resistor R4 was replaced with a 33 ohm resistor.

While this has reduced the bleed-though by approximately 20-25 dB, I still get a substantial amount of bleed-through (>-20 dBm)

I am using the  EV-ADF4355 to drive the LO (AC coupled @ -1 dBm).

Any thoughts?

Kind Regards,


  • Hello David,

    Are you driving the LO differential? How do the LO harmonics look from the ADF4355?

    The LO input polyphase filter on the ADL5375 is sensitive to harmonics. I would try to apply an external differential LO signal with good harmonics to rule that out.

    Beyond that, any block diagram, test data, and frequency configuration you can provide would help troubleshoot further.

    Best Regards,


  • Hi David,

    Sorry about the delayed response.

    I have attached a basic block diagram. 

    I performed a number of tests.  

    • Tested the LO bleed-through before the addition of DC Biased differential amplifiers (single ended 0 Vdc bias) (Result ≈ 4 dBm). I believe that the bleed-though is worse than this as it causes non-linear (saturation) effects on the signal.

    • Tested the LO bleed-through after the addition of DC Biased differential amplifiers (Result ≈ -16 dBm)

    • Tested the LO bleed-through with unfiltered internal single ended LO (Result ≈ -16 dBm)

    • Tested the LO bleed-through with filtered internal single ended LO (Result ≈ -16 dBm)

    • Tested the LO bleed-through with filtered internal differential LO (Result ≈ -16 dBm)

    • Tested the LO bleed-through with external differential LO (Stanford Instruments) (Result ≈ -16 dBm)

    Note: for the internal differential LO, the balun had to be removed and feedthrough (0 Ohm) resistors needed to be re positioned on the ADL5375-15 demo board to allow both inputs.

    See the figures below.

    Let me know if you require more information.

    Kind Regards,


    Block Diagram - Note that the LNA has a gain of approximately 20 dB.

    Unfiltered LO

    Filtered LO

    LO Bleed-through before addition of DC Biased differential amplifiers 

    LO Bleed-through after addition of DC Biased differential amplifiers