0
\$\begingroup\$

I was placing some components on PCB recently, and, being relatively new to the subject, there are still things I miss out in terms of knowledge even though I try to google every question that comes to my mind in terms of how to do things right and why.

While I found a pair of really helpful posts around here about bypassing and grounding, I still have a question to which I found no answer. At least, not in a way I tried to word it.

The bypassing rule "cap right to the pin" is simple enough, but I had situation when trace would approach the chip from under the chip, meaning I couldn't place bypass cap between power source and pin itself, I kind of have to make a fork from the pin (there is a known good option of placing bypass cap on the other side of the board right against the pin, but let's say I can't.) Bypassing guidelines gave me a lot of information, but didn't really cover my question.

Now, there is a post here, that's asking, as I see, exactly my thing, and while a pair of answers discuss it, I kind of don't see a definitive and confident answer that will say "It's totally OK" or "It's not recommended". At least nothing struck me as a confident and definitive answer for me to understand it once and for all. OK or not OK. And why, of course.

Question is: is this kind of placement OK or bad? I have a feeling it's not recommended purely intuitively (intuition feels trace inductance and extra path the noise have to make to go to the ground), but I only trust some hard proof and not my guts when it comes to scientific stuff. So I would want to know for sure and listen to the opinions of more experienced and professional people.

Edit 1: I'm talking about common low frequency circuit, where most stuff is kHz or a few MHz, maybe 100MHz as an absolute maximum for MCU, nothing more

enter image description here

\$\endgroup\$
3
  • 1
    \$\begingroup\$ The answer would depend if the chip is a simple logic chip with slow signal edges where the placement is rather irrelevant as it is good enough, or if the chip is a 100 GHz RF chip where the bypass placement is absolutely critical. Which chip would that be? \$\endgroup\$
    – Justme
    Jun 29, 2020 at 9:43
  • \$\begingroup\$ I'm talking about basic circuits and basic components, up to 100MHz in case of MCU or some 10MHz down to kHz for other chips \$\endgroup\$
    – Ilya
    Jun 29, 2020 at 9:44
  • 1
    \$\begingroup\$ It's not only the signal frequency, but the speed of signal edges that matters. 100 MHz requires fast edges. You can have 1 Hz signal with just as fast edges than 100 MHz, and it needs equally good bypassing as the 100 MHz signal \$\endgroup\$
    – Justme
    Jun 29, 2020 at 9:47

1 Answer 1

1
\$\begingroup\$

For the functioning of that IC, it's always best to have the high frequency decoupling capacitor as close to the pin as possible. It's less important whether the power supply trace comes in to the IC pin or the decoupling capacitor.

In some circumstances, we are concerned with the coupling of noise from that IC to other devices that share the power supply trace, and in those cases it is slightly preferable to have the decoupling capacitor on the power supply side of the IC. However, if we are concerned about such noise coupling, then we would likely be using additional means, like series resistors, ferrite beads or inductors in the power supply traces, in addition to decoupling capacitors. The effect of these will be orders of magnitude more significant than the order of IC and capacitor on the trace.

\$\endgroup\$
1
  • \$\begingroup\$ So it's not really crucial for a basic circuit without superaccuracy or high frequencies? Thank you \$\endgroup\$
    – Ilya
    Jun 29, 2020 at 13:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.