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

Differential line pair swip

Status
Not open for further replies.
ctzof

ctzof

Full Member level 3
Joined
Mar 1, 2012
Messages
157
Helped
12
Reputation
24
Reaction score
11
Trophy points
1,298
Location
Munich
Activity points
2,516
Hello,

I want to design a USB connection from a chip to a pcie header. The problem is the differential pins of header is in the opposite direction of that of chip. So I have to create a via and change the direction of one line to another layer in order to connect the two pins. The topology of the components cannot be changed. So I have to do something with what I have. I know that passing one of the differential lines through a via to another plane isn't a good thing, and also I have to be careful to make the lines with the same length which is difficult this way. Is there another solution to this problem. The attached photo can show you what I mean. Unbenannt.png
 

What is up with the 90 degree jog with a "stub" route, which will result in a impedance change on the differential pair.

What trace width are you using? USB requires a 90 ohm differential impedance +/- 15% from what I've read. I can't see any way around routing one of the lines through two vias (big differential impedance change). Given the spacing you can't even route both traces between the corner of the IC.

As it stands the layout was done incorrectly as the designer of this circuit should have looked carefully at the differential pair and made sure either the connector pins were placed correctly or the part was oriented correctly (in this case on the opposite side of the board).

Regards
 

The lines are just for demonstration I know that USB is 90 ohm differential the lines are going to be thinner. The layout of the chip is correct the problem is I have to swap the D+ and D- signals and the only solution is as it seems to use a via and change one of the line to a different layer to achieve that.
 

well the reason I asked about the widths was can you run a trace between the pads? or both traces between the corner pads?

I also wasn't saying the pad layout of the chip was bad, just that it should have been examined along with the connector to determine the correct placement of both the chip and the connector so the runs would not require crossing over each other, or at least you would know before hand that you would need a mitigating strategy (instead of finding out when the board is getting routed).

Regards
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top