PCB question

On3s&Z3r0s

I have sort of a general question about circuit boards used for speaker crossovers and other audio applications. I'm doing an upgrade of my 2B's so had the XO's out and took a close look at them for the first time. It looks like all of the interior traces on the SDA circuit boards have their corners chopped off. By this I mean instead of the corner making a hard right angle at 90 degrees, the corner is cut so that there is a little extra non-conductive space at the corner. The corner ends up being about 45 degrees instead of 90. (I could take a picture if that would help.)

Anyway, this reminded me of a hobby electronics kit I put together (maybe an Elenco radio kit or something like that) that mentioned that somehow this truncated corner design of the PCB made a difference in how the signal was carried. If I remember right, the materials included with the kit didn't explain how or why this made a difference, it was just an explanation of why the traces weren't square.

Has anyone else heard of this before, and if so do you know what difference it makes in the signal that passes through the board or why this might be done otherwise?

gimpod

I know exactly what your talking about. If you look again you'll notice that the chamfer is only on the outside corners of the traces and the inside corners are all 90 deg's also notice how the spacing between the traces is not very consistent at all. I believe the chamfering is a function of the design software that Polk used (if any these may have been hand drawn and photo etched).

What effect this has on the signal, I believe none because of the width of the traces and in this application (Crossovers) the overall shape of the trace does not really come into play. Where you run into problems with 90 deg bends in traces is in digital circuits where the trace width and spacing is 0.012 or less this is a big no no. Think of electrons flowing along a trace like water flowing though a hose, if you put a sharp 90 deg bend in the hose the water flow will be slowed down due to the fact that it has to change direction in order to make it through the bend. In digital circuits this can cause all kinds of problems like added resistance, noise, cross talk, shunt capacitance, etc...

Now I don't know how much you know about computers but back in the late 80's when IBM brought out there PS2 line of computers they had a problem with there 20-Mhz 80386 cpu model 80s. You could put a 80387 math coprocessor in them but when you did and used any software that used the math coprocessor the computer would emit a high pitched beeping that was loud enough to be disturbed by it over 25ft away. It turned out to be a problem in the way IBM laid out the traces for the 80387.

Hope this answers you question.

On3s&Z3r0s

Um... that's definitely a good answer... it went a little bit over my head. But I do get the direction change analogy. I wish I could remember where I originally saw the comment about the chamfer. I have seen a number of audio circuit boards (including other Polk XOs) and never noticed it until looking at the SDA XOs. Although I just looked back at pictures I took of the Monitor 5Jr XOs I modified and those were actually chamfered on the exterior and interior corners. I guess that didn't stand out to me as much since the gaps between the traces were parallel. It made the chamfering a little less obvious.

It did seem weird considering they are big, fat "traces" (if you still call them that when they are a large conductive area) that just the corners were cut off. I also see what you're saying about it only being the exterior corners. When I said interior I meant on the outer edges of the PCB they seem to be left at 90 degree angles. But on every trace not on the outside edge of the board the exterior edge of the trace is chamfered. It seems too regular to just be a quirk, but I could see where it might have something to do with the design software. Anyway, thanks for the explanation.