• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 17 subscribers
  • Views 590 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Microstrip width sanity check

Amer Abufadel
Intellectual 470 points
Other Parts Discussed in Thread: CC1101

Hello all,

  I am trying to layout a microstrip line on a 2 layer FR4 board for CC1101.  The ISM frequency that I will be using is 915MHz

According to  several (online) calculations, the strip width should be around 54 mil wide!  

Board overall thickness is 31 mil with 1oz copper.

Copper thickness: 1.37 mil

Core thickness:  28.26 mil

The CC1100DK trace is much less than that and even the EZ-RF2500 is also much less even though it is at 2.4GHz.  

What am I missing?

Thank you,

A

  • 0
    Guru 22270 points

    Hello Amir,

    The reference design found in http://www.ti.com/tool/cc1101-cc1190em915rd should take precedence over any online calculator.

    If you follow the design accurately, you should expect the same performance as found in the evaluation modules you would purchase from TI.

    As for why the online calculators should recommend that width, I could only guess that it is for the worst case scenario for a hobbiest making their own boards at home.  A design house would not be using such a calculator and they can produce thinner boards with better RF response.

    Thanks,

  • 0 in reply to Greenja
    TI__Guru**** 317180 points

    It is not clear to me which line you are refering to. Note that if a line is shorter than about 1/10 of a wavelength the line will not behave like a mikrostrip line. The layout is simulated in ADS and then some component values are adjusted to take into account non ideal effects that even a good model does not include. A calculator will only give you a approximation where to start the iteration. 

  • 0 in reply to Greenja
    Intellectual 470 points

    Hello Greenja,

      Thank you for the reference design, I had already seen that.  The board thickness in the design is twice what I can accommodate due to physical restrictions on the enclosure.

      Obviously, RF is not my strong suit so when you say that board houses do not use such calculators, what do they use instead?

    Thanks again,

    A

  • 0 in reply to TER
    Intellectual 470 points

    Ter,

      I am referring to the trace between the balun and the antenna, and I am trying to make the trace have a 50 ohm impedance at 915MHz.

      My first design iteration is a proof of concept and I don't have much resources to delve deep into a perfect design.  However, I would like to understand it so when the time comes, I can move forward on the correct path.

      In my case, the trace I refer to is about 20mm which is less than 1/10 wavelength.  

    What would be neat if there is a list of good references (text books or articles) somewhere that are recommended by the more experiences so people like me can refer to.  

    Thank you 

  • 0 in reply to Amer Abufadel
    TI__Guru**** 317180 points

    It is the distance from the routing layer to the first ground layer that is important. For the CC1101 standalone ref design this distance is 0.8mm. If you look at the CC1101-CC1190 ref design which uses 4 layers the distance is 0.25mm down to the first ground layer. Based on this you should be able to reduce your PCB thickness to 0.4mm without issues. When the thickness goes down to 0.2mm the parasitic caps will start to influence the performance.

    A text book will only give you the theoretically equations that will give you an idea of what will happen if you go up or down on a parameter but for development (not using the ref design etc) a electromagnetic simulator like ADS should be used.