see, i was talking to some people on my placement, and they said curved tracks were good for RF stuff? like an onboard antenna connection for GPS stuff?I'd stick my neck out on the line and say that, in most cases, curved tracks are nonsense. There just isn't a need for them.
i agree 45 is neater.. dunno why it just looks awesome when you get your pcb back and its got that sheen in the light as it bounces of the 10way (or however many) bus tracking across the pcb..I don't like 90° turns, so I use 45° turns instead. It makes the board neater, and helps defeat choke points.
You are asking the wrong question. Who cares what opinions people hold? Facts are all that matter. So what do we actually know?ok.. had differing opinions of the past several months about using right angled tracks, curved tracks and 45degree tracks..
what are the opinions on which is best?
Keep in mind that technology changes fast and 12 years is a long time in "tech-time", if you consider Moores Law and similar trends. The faster you go, the more it matters. In the latest high speed designs, a right angle bend can typically create a few ohms of impedance discontinuity on a 50 ohm impedance trace connection. That implies several percent of reflection. Of course, digital systems are pretty robust and so the system might easily tolerate the degradation of the digital signal, but that doesn't change the fact that a curved trace pretty effectively eliminates the sudden discontinuity and minimizes the chances of problems due to reflections.
Well to be just "technical" then 90 degree bends are far inferior to 45' bends, the tracks are longer with greater resistance (up to 41% longer in the worst cases) and PCB density suffers greatly with only 90' bends allowed....
OK, a board can be “pretty” or “professional-looking” if it has 45° angles perhaps, and I'm all for aesthetically pleasing designs, but let's be honest about why the choices are made: personal preference vs. technical reasons.
Not even at kilovolts. Best practice is to "chamfer" a 90 degree bend so that it looks like a short 45 degree trace connecting two perpendicular traces.Could sharp corners have significant effects in circuits operating at higher voltages (kilovolts)? I have seen rounded pads specified to reduce electrostatic stress, so perhaps swept bends would also be required.
Thanks for the real information in the midst of the unsupported opinions that dominate this thread.http://www.montrosecompliance.org/technical_papers/corners-Japan.pdf
This is a quite good approach to the problem. For hobbyist DIY PCBs. I would say it does not matter anything. But for a good looking PCB most often the creator have had some plan behind the work, and showed some care about component placement. And that is pluss for the board. Many of the design "rules" or do's and don'ts for PCBs have its origen in industrial production. And for a one of a kind board they are not so important. Like if you can simplify routing your hobbyist board by not place all ICs in the same direction just do it.
Well, was going to say goodbye to this thread, but this is a good rule of thumb in terms of making efficient use of board space for trace routing. When the traces all run in the same direction on a layer, space is much more efficiently used.For the "straightness" of traces, try to keep them in the same direction on one layer, and in the opposite direction on the other layer (for two layer+ boards), so if it is horizontally aligned on the bottom, try for vertically or diagonally aligned on the top.
|Thread starter||Similar threads||Forum||Replies||Date|
|RF Performance Issues - Board Layout?||Wireless & RF Design||6|
|W||Review PCB layout ZIGBEE/ADC||PCB Layout , EDA & Simulations||7|
|A||PCB Layout Advice||PCB Layout , EDA & Simulations||5|
|P||PCB Design: Footprint creation for Connector Molex Part 5024301430||PCB Layout , EDA & Simulations||0|
|E||Board layout opinions (redesign of existing board)||General Electronics Chat||8|
by Luke James
by Steve Arar