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.

TRF3722: TRF3722 and TRF3705 third, fifth order harmonics

Part Number: TRF3722
Other Parts Discussed in Thread: DAC38J84, , TRF3705


I am measuring fairly high odd order harmonics on the output of both TRF3705 and TRF3722 on the TSW3XJ8X development board. The peak power of the harmonics is around 5 dB less than the power of the main harmonic. I connected the TRF3705 LO input to an external sine generator and the results are the same as using the LO generated on the TRF3722. I'm mixing a constant DAC output (I selected a constant output on the DAC38J84) and the problem is consistent on the whole spectrum.

I'm trying to figure out whether the article (*1)  could explain this phenomenon, however I'm not confident since the LO on the TRF3722 is generated internally. Please advise what course of action to take, or whether this is expected behavior.


  • Re-examining the datasheets I realize that the 3rd and 5th order harmonics are expected at lowe frequecies with both modulators (if I'm reading the datasheets correctly). Does TI have any recomended filtering solutions?

    - Figure 107 
    Figure 50

  • Blaz:

    I assume based on the graphs that you reference that the harmonics in question are the LO harmonics.  As you have seen in the datasheet plots the odd order harmonics are high, particularly the 3rd order harmonics.  The harmonics are generated in the LO limiter/polyphase circuit that generate the quadrature LO tones to the mixers.  From the graphs you see the harmonics trail off at high frequency; this is due to output "loosing steam" at the high frequencies of the high order harmonics.  Both the TRF3705 and the TRF3722 use similar limiter/polyphase circuits.

    If these harmonics are problematic then they can indeed be filtered.  I do not have a specific filter recommendation.  Typically you would use a band pass filter that is centered at your band of interest.  This would pass your signal and reject all the unwanted higher frequency components.  Depending on the frequency of operation and signal bandwidth, there are a variety of LC (LTCC style), RF SAW, or ceramic filters that can serve this purpose.


  • Hi RJ,

    thank you for the explanation. It indeed is related to the polyphase circuit, the thing that confused me is that I was expecting the polyphase circuit to be outside the chips, but of course, the circuit is the part of the modulators. Similar circuits - hence similar characteristics.

    With the question regarding the filters, I was more wondering whether Texas Instruments' portfolio pre-made filters, or whether I should look to alternative suppliers. 

    Best regards


  • No, TI does not make or have any filters.  It is best to seek out alternative suppliers.  --RJH

  • Hi,

    I would just like one clarification regarding the Figure 109 in the datasheet for the TRF3722.

    Is the X-axis in the figure frequency of the LO or the frequency of the harmonic? For instance,  is the 5th harmonic power  -15 dBc at  at LO=1500 MHz  or LO=1500Mhz / 5 = 300MHz.  I understand is as the former, but I'd like to be sure. I'm also unsure how to square understand the performance measurements in section 6.15 in conjunction with figure 109.

    Best Regards


  • Blaz:

    Correct, the x-axis is the frequency of the LO itself.

    Section 6.15 relates the LO harmonic performance at the modulator RF output port.  Section 6.18 which includes figure 109 relates to the PLL/VCO output.


  • Hello RJ,

    The data in figure 6.15 doesn't agree with my measurements at all, while the figure 109 seems to be something observe - that is the the 3rd harmonic for instance will be - 7dbc at 500 MHz. I'm thinking that the reason perhaps is in the fact I'm mixing a constant signal (frequency = 0), and by measurement is exactly the LO from figure 109. The performance measurements in 6.15 use 4.5 and 5.5 MHz BB signal.

    However, I have checked LO output on the tsw3xj8x evaluation board (at one of the resistors next to the trf3705) and the signal seems devoid of harmonics (measured at 300 MHz). It's as if the data was exactly the reverse: LO numbers atat RF out and vice versa.

  • Blaz:

    I agree that with a 0-Hz BB signal it will be difficult to distinguish between LO harmonics and RF harmonics.  LO, image, harmonics will all line up together.

    I do not completely follow the 300 MHz reference as to whether that is the LO frequency or the BB frequency.  In any event, with respect to the odd order RF harmonics, there is an inversion that will "flip" the spectrum relative the LO harmonic.  In other words, a BB signal at fBB is mixed to LO + fBB.  The third order product will be at 3xLO - fBB.  There is a signal on the positive side, but it is not as strong as on the negative side.