0
\$\begingroup\$

Analog  input  for  AFE channels,  AC-couple  to  device  input  with  a  10-nF  capacitor.

Hello community!

I'm currently working on the design of an ultrasound scanning system, and I'm seeking guidance, particularly in the realm of high frequency PCB design and transmission lines.

My primary concern revolves around the return signals from the transducer that need to be routed to an Analog Front End (AFE) for filtering and further processing. Considering that ultrasound frequencies typically range from 2 MHz to around 15 MHz, with occasional use of even higher frequencies, Additionally, the data transfer cables back to the PCB can be up to 2 meters, mainly 50-ohm coaxial cables.

Given the low amplitude of the return signals (maxing out at around 2.2V), I'm curious about whether reflections are a common concern at these frequencies. Moreover, I'm contemplating whether matching the impedance of the lines is necessary. If impedance matching is advisable, would placing termination resistors on low-amplitude and low-power lines be a suitable approach?

I appreciate any advice or insights you can provide on these aspects of design. Feel free to share your experiences or suggest resources that might aid me in navigating this aspect of my project.

Thank you!

\$\endgroup\$
3
  • 2
    \$\begingroup\$ As far as I know, to avoid reflections in a cable, it is always necessary to do impedance matching. \$\endgroup\$ Commented Dec 2, 2023 at 14:50
  • \$\begingroup\$ Are you concerned with phase-matching of those multiple ultrasound sources? If so, then those 2 meter coax cables should be length-matched. \$\endgroup\$
    – glen_geek
    Commented Dec 2, 2023 at 15:26
  • \$\begingroup\$ the data transfer cables back to the PCB <-- I don't see them in your diagram nor do I see any data rates specified. \$\endgroup\$
    – Andy aka
    Commented Dec 2, 2023 at 16:31

1 Answer 1

1
\$\begingroup\$

15 MHz corresponds to 20 m (in vacuum). Maybe 13-17 m in coax (with 65-85% velocity factor). With 2 m cables you're on the edge of the 1/10 wavelength rule here.

With that in mind, it's probably a good idea to terminate the line, unless a moderate amount of reflection artifacts are acceptable in your application. However you probably don't need to terminate both ends --- if the source is already matched to the line then a reflection at the receiving end will only increase the signal amplitude without causing much frequency-dependent gain issue, because the source termination prevents further reflections from appearing at the receiver.

Whether terminations at both ends are required depends on the quality of the terminations and the level of errors you can accept in your application.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.