# A bypass-capacitor dialogue peels back the layers, Part 1

Editor's introduction: Bypass capacitors, which are relatively low-visibility, low-glamour components and generally not the subject of many feature stories, are vital to a successful, reliable, error-free design. Our multipart series on the subject (links at the end, below author biographies) were the most popular articles by far we presented over the past 12 months.

Authors David Ritter and Tamara Schmitz of Intersil engaged in a further dialogue on the subject. Here's Part 1 of the conversation. Dave and Tamara believe in the value of arguing, the value of education, and getting to the heart of the problem without ego; in short, ripping apart a problem for the sake of knowledge. “Listen in” and learn:

David : There is a notion that when we want to bypass, we want to use big caps (microfarads) for low frequencies and little caps (nanofarads or picofarads) for the high frequencies.

Tamara : What's wrong with that? I agree.

David : That sounds really good and makes sense but the problem is that when we tried to verify that in the lab that's not what we got! I want to challenge you about this, Dr. T.

Tamara : Go ahead. I have no fear.

David : Let's see, you have a voltage regulator and it needs a supply. The supply line has some series impedance (usually inductance and resistance) so for short terms it can't supply instantaneously large changes in currents. It needs to be supplied by a local capacitor, Figure 1.

Figure 1: Function of bypass capacitor
(Click on image to enlarge)

Tamara : I'm with you so far. That's the definition of bypassing. Go, Dave, Go!

David : Someone might, for example, bypass with a 0.1 µF capacitor. They might also put 1000 puff (picofarads) right next to that to handle the higher frequency. If we've used an ideal 0.1 µF capacitor, it wouldn't make sense to add 1000 pF next to it. It would add 1% to the value. Who cares?

Tamara : But there's more to the story than capacitor value. Neither capacitor is ideal.

David : We have to look at the actual circuit of the 0.1 µF; it's got some effective resistances in series called ESR and it's got some effective inductance in series called ESL.

Tamara : Sometimes you also consider a loss term for the dielectric as a parallel resistance, Figure 2 .

Figure 2: Model of bypass capacitor
(Click on image to enlarge)

David : Now when we hit this with a transient we're assuming that the ESL of the 0.1 µF is much larger than it is for the 1000 puff. We need something to supply current in the short term because the ESL prevents the 0.1 µF from doing so. The presumption is that the 1000 puff has a lower ESL and therefore can supply the current better.

Tamara : The ESL is related by the type of capacitor you are getting and the packaging. That can be completely separate from the size and value of the capacitor itself, Figure 3 .

David : (shows shock at younger colleague's knowledge)

Tamara : I've seen some people who step 100 nF, 10 nF, and 1 nF in parallel. They might use all the same package, say 0402, because that's what they are used to. But each of the 0402 packages has the same inductance (ESL). Since they have the same inductance, they will have the same high frequency response, so there is little improvement in there.

Figure 3: Impedance of bypass capacitor
(Click on image to enlarge)

David : The issue that we discovered in the lab was with packages that are similar. Most of the ceramics we used were 0805 or 0603 capacitor in area. I was testing an 0603 0.1 µF cap next to an 0603 100 puff cap That was not as good as just using two 0603 caps of 0.1 µF together.

Tamara : That makes perfect sense. I'd guess you were in the range of frequencies where 0603 capacitors were optimal at 0.1µF.

Figure 4: Comparison of impedance with capacitors of same size and different size
(Click on image to enlarge)

David : Yeah, the ESR and ESL was half the original value and that worked fine. In these applications I was working with switching regulators operating at around a megahertz.

Tamara : Scaling the capacitor values and packages in your case improved the bypass network at frequencies of NO INTEREST to you. By the way, Figure 4 assumes we are talking about the same type of capacitor (i.e. ceramic). There are other types, like tantalum, have a much higher ESR which bumps the entire graph up. On the other hand, sometimes a tantalum may be all that is available.

David : We're speaking about history now. In the old days people used whatever they could get their hands on. You couldn't get 100 µF capacitor in a small package. You had to improve bypass networks by shortening leads on bypass capacitors. Today the size on the large caps is getting down to where they are similar to the smaller caps. When you get down to a 0.1 µF, you can certainly get them in 0603 and eventually 0402s (and since I can't see an 0402, I have a tendency not to use those.)

Tamara : The stepped capacitors values in stepped packages came from a discussion with a colleague at Xilinx. Their FPGAs are used in a wide variety of applications and they try to test over all conditions. Therefore, they need a low-impedance supply bypassed over a wide bandwidth of frequencies up to 5 Gsps. On the other hand, you needed a smaller bandwidth solution.

David :My comments all came out of power applications of relatively low speed switchers compared to Xilinx. Your argument is very intelligent because you refer to package sizes while, other people don't go that deep. They usually say that high frequencies need small caps and the low frequencies need big caps.

Tamara : Aww, shucks. I'm blushing.

David : My bypass career has been pretty boring because most of the time, the rule was to bypass every chip with 0.1 µF and it would work.

Tamara : It's not just package, though, it's also placement.

David : Absolutely! I follow where the currents are going on the board and see what inductance is on a board. Inductance in any current path is proportional to the enclosed area of the path. So, when you wrap things around an area, you need to keep things close and tight. That's why you keep them close and tight—to keep the inductance down. Then choose capacitors with good ESL and ESR. I wish there was some more magic to it, but it's really a few simple rules applied correctly.

Tamara : Of course, you can buy caps with lower ESL and ESR, but they are usually more expensive than standard ceramic ones.

David : In most cases, though, 0.1 µF bypassing as close as possible to every chip still works just fine.

Author biographies
Dave Ritter grew up outside of Philadelphia in a house that was constantly being embellished with various antennas and random wiring. By the age of 12 his parents refused to enter the basement anymore, for fear of lethal electric shock. He attended Drexel University back when programming required intimate knowledge of keypunch machines. His checkered career wandered through NASA where he developed video -effects machines and real-time disk drives. Finally seeing the light, he entered the semiconductor industry in the early 90's. Dave has about 20 patents, some of which are actually useful. He has found a home at Intersil Corporation as a principal applications engineer. Eternally youthful and bright of spirit, Dave feels privileged to commit his ideas to paper for the entertainment and education of his soon to be massive readership.

Tamara Schmitz grew up in the Midwest, finding her way west with an acceptance letter to Stanford University. After collecting three EE degrees (BS, MS, and PhD), she taught analog circuits and test development engineering as an assistant professor at San Jose State University. With 8 years of part-time experience in applications engineering, she joined industry full-time this past August at Intersil Corporation as a principal applications engineer. In twenty years, she hopes to be as eternally youthful as Dave.

Previous articles on this topic: