The TI E2E™ design support forums will undergo maintenance from Sept. 28 to Oct. 2. If you need design support during this time, contact your TI representative or open a new support request with our customer support center.

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.

DAC5687 TRF3702 on separate PCBs?

Other Parts Discussed in Thread: DAC5687, ADS5500, TSW3003, TRF370333


I am starting on a design for a digital radio using the DAC5687 + TRF3702 for the transmit chain and the ADS5500 + a demodulator for the receive chain. The baseband processing will be done in an FPGA. We have built a working prototype using evaluation boards already and now we are starting to work on our own hardware design. We plan to implement the design on two separate PCBs - a baseband board and an RF board that will be stacked to make the design more compact. I am writing to ask for advice on how to partition the parts - is it typical, advisable, not advisable, etc., to put the TRF3702 on a separate board from the DAC5687 with connectors between them? Or should the DAC and ADC go on the RF board with the digital signals going over connectors? Are there any comments on how others have done this?

Best regards,



  • Hello Russ,

    We have seen our customers do the design either way, and the trade-offs are as follow:

    1. FPGA + ADC + DAC on one board, and modulator on another.

    FPGA and data converters have many I/Os to route, and this approach eliminates the layout constraint of routing I/Os to the connectors. The draw-back for this is that you will need to be very careful when routing the DAC analog IF signals to the modulator through connectors. The vias and connectors introduce impedance mismatches, which ultimately affects signal integrity of your IF and RF signal. However, there are only 4 lines of I/Q data to worry about instead of the many I/Os to the FPGA.

    2. FPGA on one board, and ADC + DAC + modulator on another.

    This is a solution we would prefer and is the solution we are currently implementing. By keeping all the analog traces on one board maintains the impedance matching, and it would be easier to maintain differential matching since you can control the gap between the differential lines easier (it will be harder to do to a connector since you may need to separate a differential signal due to space/size constraints). This is pretty important for the I/Q IF signal and RF signal. Also, depending the data rate you are running at, the digital signals can be more immune to the impedance mismatches of the connectors than the analog IF signal. 

    One good example is our TSW3100 + TSW3003 solution evaluation boards. The TSW3100 is a FPGA board to provide the data for our data converter EVMs. The TSW3003 is a evaluation board containing our DAC5687 and our TRF3703-33 modulator. The TSW3100 has connectors for the CMOS interface, and it also has the SAMTEC connectors for the LVDS interface. The TSW3003 can mount to the TSW3100 through the connectors to receive the data necessary for the DAC5687.

    We would recommend you follow good layout technique for either type of routing since good routing is important to achieve good performance of your system. 

    Also, may we recommend you the TRF3703-33 modulator? It is designed with 3.3V common mode voltage, and this allows an ease of interface with our 3.3V DAC5687. Also, the TRF3703 has higher performance and linearity than the TRF3702.

    Hope this helps!

    Kang Hsia

  • Hi Kang,

    Thank you. That is good information for me.

    Regarding the TRF3703 - We built our first prototype system (using eval boards) with the TRF3701 because it operated at 915MHz but now we want to use 2.4GHz so I automatically went to the TRF3702 because it looked identical except that it works at 2.4GHz. I did not look at the TRF3703 at all. I see now that it covers the whole frequency range of both the TRF3701 and TRF3702 plus some.

    Can you highlight the major differences between them? Are there any disadvantages to the 3703 as compared to the 3702 or 3701 or is it just a newer, better part?

    Best regards,



  • The TRF3703 provides improved performance compared to the TRF3701/2, specifically in the linearity parameter.  There is another advantage that the TRF370333 device operates at a 3.3V common mode that mates seamlessly with the DAC568x family.  There is a lot of history, support material, application notes available for designing with the DAC to TRF3703 interface.  Further, the wide bandwidth capability of the TRF3703 family allows multiple design applications with just one part.  There is also a soon to be released TRF3704 that covers 50 MHz to 6 GHz.  If you are in a new design phase the TRF3703 is the recommended modulator.


  • That's great, thanks for the info. One thing I noticed is that the TRF3703 is showing no stock and a 20 week lead time. I will need about 20pc for prototypes in the next few weeks. Do you know if samples are available?


  • Response taken off-forum.  --RJH