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.
Mentor Power PCB is quiet useful
Its simple to use and quiet powerful.
A supplier came to me to introduce the new version of Zuken's cadstar software. Cadstar 6.0. The new version seems quiet good but i've yet to test it.
Free use n download but certain functions are disable for the freewares.
"The new improved CADSTAR Express Software allows you to use all the functionality of CADSTAR 6, with an increased design limit of 50 library components and 300 pins"
Mostly angular routing of 45 degrees is possible in all EDA software. But circular routing is not supported in many eda platforms which is required in some ciritcal boards.I do the circular routing in autocad and import into protel and pads.
For extremely high-speed (microwave range) designs, the sharp 45 and 90 degree angles of PCB routing traces present a variation in the observed impedance of the line, as far as I know mostly due to skin effect. Designers that are working on such boards need to pay close attantion to the way they change direction of their traces, doing a gentle (circular) curve is a lot better on the signal than using a sharp angle...
will this create a considerable effect in the design.
upto what extend the tolerence difference can be noticed. when a design with any angle and digonal routing is tested.
For extremely high-speed (microwave range) designs, the sharp 45 and 90 degree angles of PCB routing traces present a variation in the observed impedance of the line, as far as I know mostly due to skin effect. Designers that are working on such boards need to pay close attantion to the way they change direction of their traces, doing a gentle (circular) curve is a lot better on the signal than using a sharp angle...
The ultimate acute angle connections occur where the non-linear user connections are made to the PCB, and where the IC is soldered to the board, and where the bonding wires inside the package transistion from the package lead, and again where the bonding wires transistion onto the die - if the "rule of thumb" is true, why not worry about those acute angles as well?.
The myth of right angle pcb trace effects is a frequently repeated "rule of thumb" that just doesn't hold up under analysis. I supose that some day, with frequencies in the 5-10GHz range, there may be some board somewhere on which pcb trace corners will make a difference. I haven't seen one yet in the 4 decades I've been in the business.
Below are some references to look at - there are many, many more:
I've been working with 2.5 Gb/s serial for a couple of years and 45's don't seem to be any problem - I remember seeing in a Motorola ECL design guide where arcs or 45's with the points shaved off were preferred to 45's but many thousands of ECL designs are around that don't follow this guidline - an article I believe on Ultracad.com even says that 90's should not be a problem, could be the third article in the last post.
For all HSD, orthogonal routing is good, without any diagonal chamfers / miters at every corners , and if we use chamfers at the trace corners, we can calculate the length how much could be chamfer length for HSD-ECL . This can be referred in the book, " EMC complaince..... " by Montrose .
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.
