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.

AFE7950EVM: Matching Network

Part Number: AFE7950EVM
Other Parts Discussed in Thread: AFE7950


I was curious as to whether or not the matching network on both the ADC input and TX output of the AFE7950EVM is optimized across it's entire frequency band that it specifies, namely (600MHz - 12GHz for TX) and (600MHz-12GHz for RX)?

I'm trying to make the circuitry around the AFE7950 as broadband as possible. Ideally, I would like to use the values used on the evaluation board on my board as to transmit and/or receive at whichever band across 600MHz - 12GHz.

The matching network in question is to the left of the baluns T7, T8, T9, and T10 for the TX and the matching network to the right of the baluns T1,T2,T3,T4,T5, and T6 for the RX on the AFE7950EVM. 

I also noticed that RXA_IN has a slightly different matching network compared to RXB_IN, why is this?

I look forward to hearing from you.



  • For some additional information, the balun we are using is a very broadband balun. It has a frequency range of 10MHz - 12GHz. This balun is used on both the TX and RX path.

  • Hi Brandon,

    The matching circuits on the AFE7950EVM are not optimized for the entire frequency band, 600MHz - 12GHz. The matching circuits were designed to be matched for X-band, specifically around 9.5GHz. 

    The reason that the matching circuits are different across channels is because on the AFE7950 substrate the routing for the channels is slightly different. So, to get the best match for 9.5GHz we had to tune the circuits for each channel.


    David Chaparro

  • What does TI think is possible for a broadband match using the wideband 10MHz-12GHz balun from Marki assuming less then 4dB or so of loss is acceptable and perhaps a few dB of slope across the band?  Would just always having a few dB of DSA loss broaden the match easily?

  • Hi Jeff/Brandon,

    Can you please advise on which Marki balun you plan to use? In these cases, wanting a broadband match, the suggestion is to use a passive attenuation circuit, 3dB T pad on each analog input/output, in order to keep the match "stiff". This is what the balun would want to see in terms of impedance in order to get a widest BW out of the device.



  • Rob,

    I believe it is the BAL-0012SSG.  Yes, the passive loss will deQ things and hopefully allow for a broader/flatter match.  The key is to know how to minimize the loss of the Tee pad such that it will broaden the match to some degree of overall flatness.  We don't want a pad larger such that we hurt the distortion of the required driver.  I think the application is planning to use the HCM8411.


    Across the 600MHz to 12GHz can you find out what the widest instantaneous bandwidth is required and the required flatness? E.g. 1dB in any 500MHz region of the 600MHz to 12GHz range.   You should assume we can put in a simple analog or digital slope equalizer, we just can't easily correct complex ripple.easily.



  • Thank you for your responses Rob and David. The Marki balun we plan on using is the BAL-0012SSG Would increasing the attenuation via the internal DSA also broaden the response of the AFE, making it more broadband as well?

  • Hi Brandon,

    yes, you can use the internal DSA on the AFE, but that probably won't help broaden the match itself with the external balun/frontend. I would definitely put the T attenuator pad there to help. At least put the component placement there for it on the pcb design, that way you can "zero" is out if you find don't need it. But you will....:)