• TI Thinks Resolved

TRF372017: DC coupling

Expert 4000 points

Replies: 7

Views: 124

Part Number: TRF372017

Hello team,

As DC component is needed in my application, I tried to change coupling caps at BBI and BBQ pins to 0Ohm resistor for DC coupling.

However, it didn't work because RFOUT level decreased by approx. 18dB.

Could you please suggest how to use TRF372017 with DC-coupled configuration?

Please find detail and spectrum from the internal link here.

Regards,

Itoh

  • Hi Itoh-san,

    The DC coupling mechanism will depend on your DAC output driver (or driver to the TRF372017), common mode voltages, swings, etc. Please see below for reference materials on example DC coupling interface with the DAC3484. You cannot simply change 0ohm resistors as the two devices may have different common mode shift.

    -Kang

  • In reply to Kang Hsia:

    Hi Kang-san,

    Thank you for suggesting the app note.

    My customer tried to compare ac-couple (1uF) and dc-couple (0Ohm) network with same input level.
    In DC couple, the RFOUT level was decreased by approx.18dB compared to ac-coupled network.

    My customer is checking the app note and trying to redesign the network.

    Do you think the designing proper DC-coupled network will resolve this output level decrease issue?

    Regards,

    Itoh

  • In reply to Kazuki Itoh:

    Hello Itoh-san,

    I suspect the DC coupling method actually attenuates the baseband input to the TRF372017. Could you please ask the customer to measure the baseband input with AC coupling and with DC coupling? I suspect somehow the DAC or baseband signal is attenuated, not RF output of the TRF372017. 

    -Kang

  • In reply to Kang Hsia:

    Hello Kang-san,

    My customer has confirmed that the input level is the same in both cases.

    FYI, the DAC at the previous stage is AD9787. My customer assumes that the attenuation is happening inside TRF372017.

    Regards,

    Itoh

  • In reply to Kazuki Itoh:

    Itoh-san:

    If you are DC coupling, then the common mode voltage between the DAC output and the modulator must match.  It looks like the AD9787 Vcm is somewhere between 0.8V and 1.2V depending on its configuration.  The Vcm of the modulator is 1.7V.  A translation network is needed to translate the Vcm of the DAC to the Vcm of the modulator,  The App note provided previously outlines the calculations to design the network.  Note, there is a bit of insertion loss due to the translation network.  I suspect that the DAC load and/or matching Vcm is the cause for the low output gain.

    FYI...the TRF3722 modulator with PLL/VCO operates with a 0.25V or 0.5V common mode voltage.  That could be a more beneficial interface.  Further, a DAC like the DAC3482 interfaces seamlessly with the TRF3722 modulator.

    --RJH

  • In reply to RJ Hopper:

    Hello RJ-san,

    Thank you so much for your suggestion.

    My customer is working on the network design referencing Figure 86 on the DS and the app note.

    Could you please let me know the recommended register setting. As far as I know:

    1)“VCM Setting”(VREF_SEL_n、Reg6 Bit23,22,21)=[000]

    2)“VCM Enable”(PWD_BB_VCM、Reg4 Bit15)=[1](OFF)

    3)“Bias select”(IB_VCM_SEL、Reg7 Bit25)=[0]

     As the it is set to OFF at #2, #1 and #2 is don't care, right?

    Regards,

    Itoh

  • In reply to Kazuki Itoh:

    Itoh-san,

    Figure 86 on the D/S will *not* work with AD9787. AD9787 has compliance voltage of -1V to 1V, and is not a current sinking DAC like the DAC3283 in the diagram.

    If the VCM is externally biased at 1.7V, then the use of VCM generator of the TRF372017 is a don't care. The VCM is biased through Hi Z 5kohm resistor. If VCM external is biased at 1.7V, then there will be *no* current flow between internal VCM generator and external VCM generator. Therefore, it is a don't care.

    However, the customer still cannot use the Figure 86 on the D/S. They will need to figure out how to meet the compliance voltage of AD9787 while maintain VCM of 1.7V for the TRF372017.

    -Kang