8
\$\begingroup\$

Where does this number come from?

For single ended it must be 50 ohm and for differential pairs 100 ohm. Why?

For PCB with controlled-impedance these are that common numbers. Outside of the PCB you can find others numbers for characteristic impedance. But what is the reason of use these numbers for the PCB tracks?

\$\endgroup\$
  • \$\begingroup\$ While I can guess where this question is coming from, you might want to expand upon your presupposition, providing basis and context. 50 Ohms for which type of situation? You'll find if you search a little, that characteristic impedance or nominal impedance values of 600, 450, 300, and 75 ohms, or 600, 75 or 50 are very common depending on the transmission line context. This question, as it stands, is not representative of actual facts. \$\endgroup\$ – Anindo Ghosh May 10 '13 at 8:32
  • 2
    \$\begingroup\$ See also this article, which covers the math for 50 Ohm and 75 Ohm characteristic impedance, and the justification for standardizing on these. Basically, 50 Ohms is a compromise between optimal power handling and lowest loss for air-dielectric coaxial lines. \$\endgroup\$ – Anindo Ghosh May 10 '13 at 8:37
  • \$\begingroup\$ @Sorry. I talked about PCB Tracks. I'll take a look that article. \$\endgroup\$ – Jesus Castane May 10 '13 at 8:42
  • \$\begingroup\$ Yup, hence I gave the question an upvote after your clarifying edit. \$\endgroup\$ – Anindo Ghosh May 10 '13 at 8:53
  • 2
    \$\begingroup\$ You can design PCB tracks to be any impedance you want within reason \$\endgroup\$ – Andy aka May 10 '13 at 10:35
8
\$\begingroup\$

The 50Ω standard is basically just convention. There are various stories about how 50Ω came to be chosen. The article Anindo linked is good. There is also The History of 50 Ω or There’s Nothing Magic About 50 Ohms. But the long and short of it is that it is a compromise between low attenuation and power handling.

But it became the standard impedance when designing for transmission line applications way back when. When an IC datasheet says you need to design your PCB traces with a controlled impedance, then you're designing to compensate for transmission line effects. If the impedance of the trace is matched to the output impedance of the IC or source, you reduce the possibility of reflections which would lead to standing waves on the trace and cause all sorts of headaches. Since the designers of the IC are designing with transmission line effects in mind, and since 50Ω is commonly used by convention, the 50Ω standard proliferates.

But 50Ω is by no means special. From this paper on controlled solutions by Advanced Layout Solutions:

Within reason, the absolute impedance value chosen is not normally important, providing it is controlled along the entire length of the line. Other constraints in a design often dictate the impedance for us; it may be chosen based on a design specification (e.g. 65 Ohms for PCI) or chosen to reduce current (a high impedance). It will generally be between 45 and 80 Ohms due to typical material geometries, and if the signal changes layer then the trace geometry should be adjusted as necessary to maintain a consistent Zo.

\$\endgroup\$
0
\$\begingroup\$

50 ohms is used because it is the impedance of the coaxial cable, except audiovisual applications is 75ohms, and so we avoid having to use an impedance matching circuit to connect the PCB to the outside.

\$\endgroup\$
  • \$\begingroup\$ But when you make a connection between ICs you also use this impedance. Is there any reason? \$\endgroup\$ – Jesus Castane May 10 '13 at 12:16
  • \$\begingroup\$ Any time a signal passes between media whose impedance differs, some of that signal will be passed through and some will be reflected. If a short section of a 50 ohm cable has a 75 ohm impedance, then 33% of the voltage signal will be reflected at each end of the 75-ohm section. About 11% of the signal will survive one round trip, 1.2% will survive two, and 0.15% survive three. If the round-trip time is short enough, reflections may die down quickly enough to not pose a problem. Whether one needs to design for impedance matching depends what sorts of reflections are tolerable. \$\endgroup\$ – supercat Nov 22 '13 at 19:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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