An interesting Altium video about right angle vs 45 degree vs arc traces and signal integrity. The findings are pretty clear, no significant differences up to mm wave frequencies.
An interesting Altium video about right angle vs 45 degree vs arc traces and signal integrity. The findings are pretty clear, no significant differences up to mm wave frequencies.
I didn’t watched it, but want to say something about the problem.
Many people (electronics also) confuse the speed of the signal with the speed of the electrons and imagine that electrons in track travel with close to light speed and because of this have a problems with cornering (I took it from Google translator, hope correctly - I wanted to say like a driver who enters a curve at too high a speed).
But that’s not true and electrons don’t fall out on the bends hitting the neighboring tracks and pads.
It’s worth a watch as it is concise and to the point, a rarity on Youtube these days. The summary is that our standard 45 degree tracks are the best by a small margin - simply by being shorter.
Curved tracks are for aesthetics, not for performance.
In addition: Don’t forget to ignore the acid traps!
Thomas_Gundersen (Below) post a link to a youtube video about an interview with Sierra Circuits and apparently acid traps are still a valid concern
I’m sure its due to induction being more of a problem at 90 than 45 degrees and changing the impedance.
Also the electrons move very slowly, without looking it up, I think its less than walking speed. The electromagnetic wave travels at the speed of light.
Last time I was using curved tracks was when as teenager (and little later) I was drawing my tracks (using ballpoint pen refill with ballpoint removed) directly at PCB.
I don’t need to see anything to just use 45° tracks and not plan to change it. I sometimes use rounded tracks when I am constructing coils only.
Watch the video. The simulations and EM solver found no significant impedance irregularities
Yeah, does look like it makes no real difference unless you have a poor manufacturer, though that would still make 45 degree best practice. (Probably still best for differential pairs due to extra skew in 90 degree turns)
Looks like hanging Vias cause more issues than 90 degree traces.
Which is fine, and useful for copying vintage hand drawn boards
And it feels nice, if you imagine it is a water slide and you are on it. ![]()
“45 degree tracks” are partly esthetic, party a habit for me. But there is a more important reason. It saves board area that can be used for other items. On some boards this is neglible, but on other PCB’s it may be significant. Especially when for example a QFP is also rotated 45 degrees to make fanout more compact.
The first PCB design program I used (Ultiboard, around late '80-ies) had a function to chamfer all tracks, and as I still remember that, I must have liked that function.
For a point of reference: Some years ago, I did TDR (Time Domain Reflectometer) measurements on a board with USB (480 Mbits/sec) and Wifi antenna (2.5GHz) traces. The USB lines were routed with 45 degree turns and the Wifi lines were routed using smooth curves.
With the TDR, You could see reflections at each 45 degree bend (USB lines). For the curved WiFi lines, you could not see any reflections along the routing.
So, there is some validity to using curves on high speed signals. How significant that is on the end result, will depend on the details of the design.
For a long time, I thought that sharp 90 degree bends were more subject to peeling away from the substrate. I am surprised to not see anyone mentioning that.
In the texts / links above it has also been mentioned that you can see differences, but they are negligible up to 50GHz or thereabouts.
The video focus on the SI part of 45 vs 90 deg.
The other part is manufacturing/etching.
As some of you have already have mentioned; acid traps is more of a concern with 90 deg, especially if you are pushing the manufacturing capabilities; trace width and spacing.
Philip from Phil’s Lab has a video: “PCB Design for Manufacturing (DFM) Q&A with Sierra Circuits” where he talks with Amit Bahl. They talk a little bit of acid traps and (>)90 deg.
Amit: “..acid traps are real and if you have the ability to teardrop or again remove as many 90 deg angles as possible I would do it.”