It requires mixing traces on different layers and determining the coupling capacitance/inductance between traces on different elevations, so it is a more complex calculation by hand and from a numerical approach. These are also not the best trace geometry and routing option due to the crosstalk you can create if you route multiples of these differential pairs on two layers, which will happen if you reverse the polarity of the traces on different layers.
@@Zachariah-Peterson Thank you for your response. I agree that these may not be the best option, and that they may give the software fits. On the other hand they do seem to work well enough for the slower signals (
I prefer to ask my fabricator for these calculations as they are more aware of material availability, process considerations such as pressing and etch factor which will impact on dielectric thickness and trace widths. These are 'real world' implications that software doesn't consider.
Why doesn't Altium support broadside-coupled differential pairs?
It requires mixing traces on different layers and determining the coupling capacitance/inductance between traces on different elevations, so it is a more complex calculation by hand and from a numerical approach. These are also not the best trace geometry and routing option due to the crosstalk you can create if you route multiples of these differential pairs on two layers, which will happen if you reverse the polarity of the traces on different layers.
@@Zachariah-Peterson Thank you for your response. I agree that these may not be the best option, and that they may give the software fits. On the other hand they do seem to work well enough for the slower signals (
"broadside-coupled differential pairs" , never heard of that.
I prefer to ask my fabricator for these calculations as they are more aware of material availability, process considerations such as pressing and etch factor which will impact on dielectric thickness and trace widths. These are 'real world' implications that software doesn't consider.