5
\$\begingroup\$

The ground bounce is effect that ground rises due to inductance in the ground lead. https://www.eetimes.com/document.asp?doc_id=1255019

It is said that ground plane has low impedance, for that reason there is less electrical noise (so less ground bounce!)

But It was inductance that leads to ground bounce, and not impedance? Can any one explain me this contradiction?

Thanks!

\$\endgroup\$
  • \$\begingroup\$ Inductance is a kind of impedance (more-or-less) \$\endgroup\$ – immibis May 24 '18 at 22:31
7
\$\begingroup\$

In a ground plane, the most prominent contributors to the impedance are the inductance and the capacitance. In most cases the sheet resistance is low enough to be not mentioned.

Inductance is the only one with a contribution that increases the impedance when it increases, therefore these terms are cited interchangeably pretty often.

I think when one is using ground planes, albeit inaccurate, it is not a dire mistake to call it an inductance of a plane.

A conducting plane can be modeled as inductors that are woven together on a 2D lattice, and capacitors that connect the nodes formed by pair of inductors to a reference node. (Being not pedantic, lets call this the "real ground")

What creates a bounce is the EMF generated by the current that is trying to flow through this lattice of inductors. Higher the inductance, higher the bounce.

\$\endgroup\$
6
\$\begingroup\$

But It was inductance that leads to ground bounce, and not impedance?

Inductors have impedance i.e. they impede the flow of a changing current by creating a back-emf: -

$$V_{emf} = -L\dfrac{di}{dt}$$

The impedance of an inductor is also \$2\pi f L\$ (from the perspective of an AC signal) and that is the impedance they are referring to when they talk about the impedance of a ground plane.

\$\endgroup\$
  • \$\begingroup\$ So whats the inductance? \$\endgroup\$ – Kono May 24 '18 at 11:24
  • \$\begingroup\$ But It's not the impedance that causes ground bounce, but inductance? \$\endgroup\$ – Kono May 24 '18 at 11:25
  • 3
    \$\begingroup\$ And, an inductor exhibits that phenomonom of impedance. \$\endgroup\$ – Andy aka May 24 '18 at 11:26
4
\$\begingroup\$

Something that took me some hard lessons to learn:

A ground plane in itself is not low impedance. If it was, we could market copper foil as ultra-low-impedance wire.

A closed circuit formed from a conductor close to a ground plane, AND the ground plane, can be very low impedance, though: it is equivalent to a very closely spaced pair of wires that perfectly follow each other's path, which is the lowest impedance circuit possible compared to anything with more spacing (loop area, which is a key ingredient in building coils!) between the conductors. The path of an AC return current in the ground plane will actually mostly follow the path of the conductor above it, since it is the path of least reactance.

Ground bounce is what happens when an RF (or fast rise pulse) signal causes a voltage drop across the reactance of the ground path, so we will minimize it by a ground plane design.

Since this tends to keep the return paths out of each other's way (compared to discrete ground wires, were it will be often the smallest loop and not the assigned ground wire carrying the current!) even though they are galvanically interconnected, you will also lessen the effect residual ground bounce will have, since the reactances the voltage drops occur across are not shared - especially not shared between circuit sections handling amplifyable and amplified signals.

Additionally, if a ground plane AND a power plane are used, they form capacitor plates, acting as one of the best RF filter capacitors available.

Additionally, having currents flowing in shortest loops possible lessens problems with parasitic transformers and also EMI.

Additionally, ground planes make good shields.

\$\endgroup\$
2
\$\begingroup\$

It would appear at a glance they were talking about the leads / traces leading to the Vcc and to the ground plance from the components. Normally these traces could pose problems at high frequencies, and certainly the vias to the ground plane below the top layer will add some inductance to ground. The size of the vias will matter vis a vis their inductances. For this I suggest looking at page 11 of this article https: https://www.keysight.com/upload/cmc_upload/All/2_Demystifying_Vias_in_High-Speed_PCB_Designs.pdf The impedance increases significantly for frequencies > 5GHz though so this may not be an issue in your case, but is reduces overall when vias are added in multiples Other inductances added from battery leads etc. probably more significant than the actual distributed inductance of the ground plance itself.

Regards,

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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