Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
A differential pair doesn't need a return path and not necessarily a ground plane below it. But the differential impedance will be lower respectively the required trace width smaller with a ground. In addition, a differential pair has a common mode impedance referred to ground. It will be strongly affected by the existence of a ground plane. For USB, a common mode impedance of 30 ohm +/- 30% is required. It can be only achieved with a ground plane.
Kotsche:mid layers will definitely solve your problem(considering your copper thickness+tin plating thickness,Er etc).......
but you can try reducing the dielectric thicknes and increase the copper thickness to achieve the mentioned 90 ohm differential impedance(like 47mils from 62mils(1.6mm)) and achieve the same in 2 layers.....i have done the same but for a different impedance value....
unless it affetcs any mechanical or other parameters....
i just want to know that waht is the use of gnd plane for differential signal as i know that in differential pair both the signals gives refrence to each other.please expalin.
i just want to know that waht is the use of gnd plane for differential signal as i know that in differential pair both the signals gives refrence to each other.please expalin.
in a differential plane the inverted/opposite line acts as the return current path.......but as FVM mentioned in the earlier path common mode impedance and the capacitance felt on the differential lines is a factor affected by the gnd plane ......the importance of a gnd plane in a differential signal scenario can come into picture if there is a length mismatch between the differential lines.....gnd plane can act as the return path to avoid the noise/radiation generated due to the timing mismatch.....
Differential pairs can be closely coupled or loosely coupled, have a look at this: **broken link removed**
When a ground plane is present they will couple to it, same with a power plane if its near enough, they can also couple to each other without a ground plane, thats how ethernet works.
If there is a length mismatch you will get noise, whether the ground is there or not.
But what if for the case, if I take away the ground underneath it in order to reduce or eliminate the induced current loops under the differential pair, but replace it with other signal traces, I guess this will couple noise from the differential pair which has length mismatch to the signal traces below it.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
