4
\$\begingroup\$

I read many experts here recommend the caps (usually 100nF ones) as close to the power and ground pins as possible. Why do they have to be soldered very near to the pins?

And I used the term decoupling capacitors. Is decoupling and bypass caps same thing in this context.

\$\endgroup\$
  • 7
    \$\begingroup\$ To minimize inductance. \$\endgroup\$ – Ignacio Vazquez-Abrams May 10 '16 at 7:56
  • \$\begingroup\$ inductance? but there is no coil like wiring. \$\endgroup\$ – user16307 May 10 '16 at 8:35
  • 4
    \$\begingroup\$ At those frequencies the PCB tracks serve as coil like wiring. \$\endgroup\$ – Brian Drummond May 10 '16 at 8:42
  • 8
    \$\begingroup\$ Actually any wire is an inductor. You get about 1 nH inductance per millimeter of wire. That doesn't sound like much but it adds up and at high (RF) frequencies these nano Henrys can be a nuisance. \$\endgroup\$ – Bimpelrekkie May 10 '16 at 8:45
  • 2
    \$\begingroup\$ @user16307 Correct. Inductors looks like shorts at DC. As frequency goes up, the reactance of inductors increases. Capacitors are the opposite. Keep in mind that any physical conductor has resistance, capacitance, and inductance that need to be taken into consideration when designing high speed or sensitive circuits. \$\endgroup\$ – CHendrix May 10 '16 at 11:48
11
\$\begingroup\$

The ICs contain fast switching transistors generating RF signals. As you know all electrical signals travel in loops. For these ICs the loop is through the power supply pins.

The decoupling caps form a short circuit for these RF signals so the closer you mount the decoupling caps to the power pins of the ICs the smaller the loop will be. This is desirable as it increases the effectiveness of the decoupling because any distance increases parasitic inductance of the wires (about 1 nH per mm). Also large loops emit more RF signals so you have more chance of violating EMI (Electromagnetic Interference) specifications.

Also, a longer distance to the decoupling cap means that the supply voltage inside the IC will be more noisy and polluted with spikes. Worst case the IC stops working because of all the ripple on the supply !

Decoupling caps and bypass caps are indeed the same thing.

\$\endgroup\$
  • \$\begingroup\$ you mean it would be futile if the cap legs are lets say 500cm far away from the pins? No use? \$\endgroup\$ – user16307 May 10 '16 at 8:25
  • 2
    \$\begingroup\$ Yes, at that distance the decoupling cap would do almost nothing. I would consider 2 centimeters or so the maximum distance that would be OK-ish if there was no way to place the caps closer. Note how on allmodern PC motherboards and graphic cards there are many capacitors directly under (I mean, on the other side of the PCB) the large Chips (the CPU and GPU). These are all decoupling caps ! \$\endgroup\$ – Bimpelrekkie May 10 '16 at 8:29
  • \$\begingroup\$ For power rails of opamps they always recommend 100nF. Is that really because the switching transistors cause a related EMI freq? \$\endgroup\$ – user16307 May 10 '16 at 8:33
  • 1
    \$\begingroup\$ For opamps it is not because of switching as opamps do not contain fast switching digital circuits. For opamps (and other linear ICs like regulators) supply decoupling is needed also to keep the RF loops short but that has more to do with preventing oscillations and stabilizing the power supply voltages. Also, large capacitors like 100nF cannot be integrated on the ICs itself so they have to be added externally. \$\endgroup\$ – Bimpelrekkie May 10 '16 at 8:40
  • 1
    \$\begingroup\$ @NickMiller Its is all about the impedance (resistance) of the decoupling, this should be as low as possible to be effective. If the impedance of the decoupling is too high, it does nothing. The distrurbing signals will go elsewhere and that must be prevented. For the high frequencies we want to short out with decoupling, any wire is an inductance and the more inductance, the higher the impedance. So placing a decoupling capacitor "far away" will make it do nothing as the impedance for the high frequency signals will be too high. So it needs to be close to the chip to be effective. \$\endgroup\$ – Bimpelrekkie Jun 28 '16 at 20:03
5
\$\begingroup\$

Decoupling Caps are used to avoid the noise/glitch on the power supply line. Basically when you say noise, it can be of many types. For the decoupling caps, one of the primary advantage is to remove the ground bounce(from Ground plane) & voltage sag (from voltage rail).

Inside an IC, using NMOS & PMOS circuit, the switching happens. In the figure, enter image description herewhen the signal is pulled to VSS, there will a voltage drop across parasitic inductor as shown which will lead to sagging of voltage & when the signal is pulled down to ground, there will be a voltage drop on the ground side parasitic inductance which leads to ground bounce.

As a board designer, you can not do anything about it. So, to remove this issue on the board level, which is caused by parasitic inductance of trace & plane, we add a decoupling capacitor to provide a local path of voltage & ground. During fast switching , the capacitor acts as a decoupling element to reduce the drop across parasitic inductance. The board level figure with decoupling capacitor is given below :-

enter image description here

The farther the capcitor is , the more is the trace length & the more is parasictic inductance. So, it is advised to place it as close to the voltage or ground pin as possible.

It is a trade off or vendor recommendation to put it near to voltage pin or ground pin.

\$\endgroup\$
  • 4
    \$\begingroup\$ Inside an IC, with the CMOS circuit And then your drawing shows two bipolar NPN transistors. Replace those with an NMOS and a PMOS and your story will make more sense. \$\endgroup\$ – Bimpelrekkie May 10 '16 at 11:52
0
\$\begingroup\$

Yes "decoupling" and "bypass" capacitors are the same thing.

Ideally the power supply to a chip would have a zero impedance at all frequencies. If the power supply has a finite impedance it will act as an unwanted coupling path. The higher the impedance the stronger this unwanted coupling path.

The unwanted coupling path can have various effects. In an amplifier it can cause feedback and hence oscilation in analog circuits. In a multi-channel analog circuit it can cause crosstalk between channels. In a digital circuit current spikes from gates switching can potentially cause glitches in other gates.

It is important to realise that the frequencies that cause a circuit to misbehave can be much higher than the operating frequency of the circuit. An amplifier can potentially oscilate at any frequency where it has gain. If you look at a fast edge in the frequency domain you will find very high frequency components.

All electrical connections have inductance and the longer the connection the higher the inductance. So to keep the impedance at high frequencies down we place a capacitor as close as possible to the device. This bypasses the power supply providing a low impedance path between power and ground at high frequencies and hence reduces the coupling between circuits fed from that power supply.

\$\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.