Paying attention to the "polarity" during schematic edits can save the PCB designer from having to "flip it" if you've altered something that otherwise does not matter.
I use resistor, cap, and inductor symbols with "polarity" to avoid this "flipping," and the marker is less clutter than two pin numbers.
Conical inductors must be oriented properly, of course.
I remember having a board made with a right angle 0.1" header. The assembl y house called me up to point out I numbered the connector backwards. Huh, what??? I can plug the mating connector on either way. It's not a polari zed connector. It can only go into the holes one way because of parts inte rfering. WTF??? They believe the manufacturer's data sheet no matter what .
I guess I should be glad they didn't install it from the bottom of the boar d!
Sometimes surface mount inductors have colored dots to indicate values
All traces and wires have an inductance from 5 to 10 nH/cm and air coil mut ual coupling with full bandwidth may be small but orientation matters great ly in VHF to microwave unless there is a shield or cavity.
RF Capacitors are not polarized but the layout of caps must be examined so the similar problem of stray pF coupling from transient E fields does not c ause a spurious responses.
Low impedance traces become potential targets to high H field transients (d I/dt) and High impedance traces become potential targets to high E field tr ansients (dV/dt) The impedance ratios and path loss determines attenuation while polarity determines phase.
utual coupling with full bandwidth may be small but orientation matters gre atly in VHF to microwave unless there is a shield or cavity.
o the similar problem of stray pF coupling from transient E fields does not cause a spurious responses.
(dI/dt) and High impedance traces become potential targets to high E field transients (dV/dt) The impedance ratios and path loss determines attenuatio n while polarity determines phase.
Running traces over ground plane does limit the spread of the electromagnet ic coupling - it becomes a dipole-to-dipole coupling which falls off as the fourth to sixth power of the spacing (depending on orientation).
Putting grounded fingers between traces adds a bit more shielding.
Adding inner ground planes to what had started off as double-side boards ki lled off a great deal of previously troublesome electromagnetic interferen ce. The double-sided boards had needed a ground metal plate next to the boa rd to stop them interacting with adjacent boards. The four layer replacemen t boards didn't need the grounded screen, and performed rather better, even without it.
The nice thing about ECL is that you can use balanced traces (with equal an d opposite voltage swings) which are even quieter and even more insensitive to radiated noise.
Quite usefull if you use several of them to build filters or delay lines.
Have something to add? Share your thoughts — no account required.
Ask the community — no account required