Continue to Site

Welcome to EDAboard.com

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.

Register Log in

Width of trace and distance between the traces

Status
Not open for further replies.
C

cdz

Junior Member level 2
Joined
Mar 21, 2008
Messages
20
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,434
Hi All,

I have a question about the layout of IC. For high speed differential signals (few GHz to 10GHz), I know the width depends on the current of the signals, but what is distance between these trace? does this related to the width of the trace?

Thanks in advance

CDZ
 

hi cdz - you're right that the current depends on the Width of the wire...if it's high current then you need wider signal wire you can compute the current handling of different metals.

The distance of the metal from point A to point B of the signal concern. e.g. you have 1micron width of metal with 20micron long(distance).

Width is for current handling and distance for IR drop or resistance of the metal.
 

Hi Thanks for reply,

Maybe my question is not clear. Sorry about that. [/img] I upload the figure to help me explain what's my question. the current will decide the width(W) of the trace A and B. My question is: for the speed of few GHz, what's the distance between the trace A and B (d in the figure) and why? Thanks.

CDZ
 

I usually choose the distance between the traces so that it does not add too much capacitive load to the driver but still keeps the traces close enough to experience the same noise (so that it is common mode noise).
 

JoannesPaulus said:
I usually choose the distance between the traces so that it does not add too much capacitive load to the driver but still keeps the traces close enough to experience the same noise (so that it is common mode noise).
Click to expand...

Thanks for reply. That is true, if it is too close, then the capacitance between the traces is large and the crosstalk will increase. But if it too far, the CMRR performance will decrease and also the area will increase, therefore the cost may increase. Does the distance has relation with the width of the traces. in one of my interview, they asked my how to choose the distance in number with relation of the width of the traces, for example, 1d, 1.5d, or any other number.

Thanks.

CDZ
 

If it is a differential signal there is no crosstalk.

Where you talking about transmission lines in your interview?
 

JoannesPaulus said:
If it is a differential signal there is no crosstalk.
Click to expand...

Yes, for differential signal there is no crosstalk. But if there are several pairs of differential signals, then there is crosstalk between the differential signals. You mentioned that the trace is close enough for CMMR. Is there any typical value for the signals at a few GHz?

Thanks.

CDZ
 

You could put a GND shield between each pair of differential signals, this might allow you to reduce the distance
 

I do not know a formula for the distance between different differential pairs. What I usually do is post-layout simulations to verify that the isolation between the different pairs is within specs.

Adding a GND in between will increase the capacitance but will certainly decrease the area required.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top