June 7, 2024 • 46 mins
In this episode, Parker and Stephen dive into the intricacies of capacitor selection, discussing various types of capacitors, their properties, and how different factors affect their performance. They also touch on the new custom bare PCB service offered by MacroFab. The episode is rich with technical details, tips for selecting capacitors, and practical advice for engineers.
Key Discussion Points:
- Introduction to the new custom bare PCB service by MacroFab
- Detailed explanation of capacitor types: Class 1 and Class 2
- Importance of voltage coefficient of capacitance (VCC)
- Challenges in selecting capacitors for different applications
- Comparison between different capacitor sizes and their effects
- Tips for avoiding edge cases in capacitor selection
- Explanation of how barium titanate affects capacitor performance
- Murata's Sim Surfing tool for capacitor data
- Impact of capacitor choice on manufacturing and performance
- Best practices for selecting capacitors in design projects
Relevant Links:
- MacroFab Custom PCBs
- Episode 141 - It Depends - An In Depth Look at MLCCs with James Lewis
- Murata Sim Surfing
Community Questions:
- What are your thoughts on the voltage coefficient of capacitance (VCC) and its impact on capacitor selection?
- How do you handle the challenges of selecting capacitors for your projects?
- Have you used Murata's Sim Surfing tool for capacitor data? If so, how has it helped you?
MacroFab:
This show is brought to you by MacroFab, hardware development, designing and prototyping for individuals, startups, and businesses. Key MacroFab services include PCB (Printed Circuit Board) fabrication, assembly, and testing. Customers can use MacroFab's platform to upload their PCB designs, select components, and specify manufacturing requirements.
We Want to Hear From You!
Subscribe to Circuit Break wherever you get your podcasts! And join our online discussion hub at forum.macrofab.com to keep the conversation going with electrical engineering experts and experimenters! You can also email us at podcast@macrofab.com.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Parker Dillmann (00:10):
Welcome to circuit break from MacroFab, a
weekly podcast about all thingsengineering, DIY projects,
manufacturing, industry news,and choosing capacitors. We're
your hosts, electricalengineers, Parker Dillmann.
Stephen Kraig (00:24):
And Stephen Kraig.
Parker Dillmann (00:25):
This is episode 434. And before we jump into the
podcast, we have a macro fabannouncement or an update so to
speak. We now offer bare slashunpopulated PCBs. So forever
we've always done PCB assemblyand we've kind of shied away
(00:46):
from just supplying bare boardsto customers but we've had
enough customers essentiallyrequest for them to able to buy
bare PCBs and so to explain moreabout that I have Kyle from
product to talk about that.
Kyle McLeod (01:01):
We're really excited to offer bare PCBs to
our customers. This is somethingour customers have been asking
for. And we found that we have areally, good position to offer,
high quality boards, boards witha lot of custom specs,
engineering support on gettingyour boards manufactured and
really, our goal is to give youthe opportunity to, you know,
(01:23):
order a board that you areconfident in when you're doing
your prototyping, that the boardyou get is gonna be, is gonna
work, is gonna be usable whenyou go to add your components to
it and then you're not gonnahave to start over or spend a
lot of time questioning and didyou make a a mistake or is there
something wrong with the boardthat the manufacturer didn't
(01:45):
capture. So we think we havereally good, high quality
boards. We have lots of customspecs.
And really, I think the mostexciting thing is is that when
you go to scale to production orif you're in production and you
wanna iterate on your design, wehave made it really easy in the
platform to switch between abare PCB and your PCB with
(02:06):
components. So, if you'veuploaded your PCB and you've
added your bill of materials,you don't have to go and create
a different PCB to order thebare boards. You can just turn
off the components for yourorder and we'll let you order
that bare board. When you wantto view the components again,
you turn them back on and allyour bill bill of materials, all
your placement data is stillthere. And so you can work, on
(02:29):
your on your designs, keep keepyour designs intact without
having to maintain a separatebare PCB file and a, assembled
PCB file.
And we think that's gonna makeit really easy for our
customers. Another thing, byhaving all of the same offerings
from our specs and from ourmanufacturers. We allow our
(02:53):
customers to easy transitioneasily transition from prototype
to production because you don'thave to upload new files, you
don't have to wonder, you know,if if you were to do a prototype
with 1 shop and then shift tosomebody else, you know, they're
gonna have differentrequirements. And so, by doing
everything with Macrofab, youcan keep all of your your data,
with the same person. We make itreally easy to add those
(03:14):
components on and order yourassemble boards, once you're
ready for production.
So, yeah. Bare PCBs is somethingwe're we've been very excited
about. We think it's a reallynice, smooth feature for
prototyping and for foriterating on your your
production products as well.
Parker Dillmann (03:29):
So, yeah. Check out our new custom PCB service
on our websiteatmacrofab.com/customhyphen PCBs
and take your projects to thenext level. Whether you're
prototyping or ready forproduction, MacReb has you
covered. Okay. Let's jump intothis capacitor selection topic,
Steven.
(03:50):
What what is this about?
Stephen Kraig (03:52):
You know, it's at work. We've been talking about
this recently. Just the idea ofhow do you pick a capacitor.
Like you go to digi key, you goto mouser, you go anywhere And
as a designer, you know you needa particular value. You know
some of the other parameters youneed.
But how do you go about actuallyselecting a an appropriate
(04:16):
capacitor? And it ends up beingquite a bigger thing than you
think. There's every decisionyou make flows downstream to all
the people that have to dealwith your decision. And I like
framing it that way because, youknow, we as electrical
engineers, a lot of times canget pigeonholed in our thinking
(04:37):
of I found this perfect thingthat works for me, and this is
it. That's it.
It's done. I as long as I justput it on my bill of materials,
put on my schematic, my work isdone, wash my hands. But there's
so much more to it, you know,when it comes to purchasing,
when it comes to actuallygetting it, when it comes to
putting it on the board, when itcomes to validating it. There's
(05:00):
so much more to that. So Iwanted to touch on a handful of
topics with that and maybe bringup some stuff that that we
haven't really spoken aboutbefore.
Way back when I don't actually,I don't remember the date, but
but on episode 141, so manyepisodes ago, we had James
Lewis, otherwise known as theBald Engineer, come on. The
(05:21):
episode was called It Depends,an in-depth look at MLCCs with
James Lewis, where we weretalking about some similar
concepts of what we'rementioning today with selecting
capacitors. James Lewis has someexperience working with KEMET in
the past, and and he reallybrought some really cool
information about how capacitorsare made and a lot of other good
(05:44):
information, and I wanna kindajump off that today since we
haven't talked about this in awhile. So first of all,
capacitors kind of come in infact, let let me stop by saying
capacitors come in a ton ofdifferent varieties. Today, I
just wanna talk about MLCCs,multilayer ceramic capacitors.
And that's your, like, bogstandard tan looking capacitor
(06:08):
that you see on a bazilliondifferent boards. There are so
many other variants of howcapacitors are made that we're
not going to be covering today.Really wanna just look at MLCCs.
Now MLCCs come in 2 differentclasses, class 1 and class 2.
Class 1 is your COGNPOdielectric capacitor.
(06:32):
So that should be TEMCO thatthey, that is part of the
material that they're made of,which is calcium zirconium
oxide. The pros of class 1capacitors is that they have
really good temperaturestability. They have really good
aging effects, and theybasically are very stable
(06:53):
capacitors. The cons to these isthat they typically have a
limited capacitance range.You're not really gonna get high
capacitance values in a classone capacitor, and they tend to
also be in the smaller sizerange.
So, you know, if you're lookingfor a 100 microfarad 1210
(07:14):
capacitor, just pretty muchforget class 1 and forget the
benefits you get from having, agood dielectric and a stable
capacitor. You you're prettymuch relegated to class 2. Class
2 caps are made
Parker Dillmann (07:29):
of They are also more expensive typically.
Stephen Kraig (07:33):
Typically. Correct. Typically. Class 2
capacitors are made of bariumtitanate, and that's the bulk of
your capacitors. The x seven r,the y five v's, the just the
generic capacitor that you findeverywhere is almost guaranteed
to be a class 2 capacitor.
(07:54):
So the one the main topic that Iactually wanted to kind of break
into today was something thatwas mentioned back in episode
141 but not really expandedupon, and that is the voltage
coefficient of capacitance. Soit may not it's interesting
because this topic is somethingthat isn't called out on
(08:16):
datasheets and it doesn't evenhave a standard way of
determining it. But class 2 MLCCcapacitors have a voltage
dependence to their capacitance,which I find somewhat
interesting that it's that thischaracteristic of capacitors are
(08:39):
is not on data sheets. It's notcalled out. It's not defined.
But if you know it exists, youknow it exists. This this kind
of stuff that actually gets tome when it comes to new people
getting into electronics becauseI remember when I was first
learning electronics,capacitors, like, after I
finally got a grip on what aresistor actually is, I'm like,
(09:01):
okay. Let's look at this nextcategory. And it was capacitors,
and it's like, oh my god. Like,you you 10 x the parameters that
I have to care about now, and Ihave no idea what any of them
are.
And without even being told,like, the data sheet doesn't
even mention it, the actualparameter that you care the most
about, the capacitance, is afunction of the voltage that you
(09:25):
apply across them, which is mindboggling, but it's something
that you have to keep in mind.So in in research for this, I
actually was not aware of this,but there is a body called the
Electronics Industries Alliance,the EIA. And this is one of the
bodies that sets the names ofthe parameters of everything
(09:47):
that goes into a datasheet. Ishouldn't say everything, but, a
a good chunk of them. So whenyou say, like, capacitance or
whatnot, it's defined, like,this is one of the bodies that
says, here's what capacitance isand capacitor should or a
capacitor datasheet shouldinclude that.
And it's interesting becausethis voltage coefficient of
(10:09):
capacitance just is not part ofthe EIA whatsoever. So this can
really catch newcomers. And infact, I I think there's plenty
of people who are well pastbeing a newcomer who probably
are not aware of this justbecause you don't know about it
unless you know about it.
Parker Dillmann (10:26):
Yeah. And the thing that this would affect the
most is, like, on feedbackcircuits.
Stephen Kraig (10:33):
This is yes. This this can affect virtually
everything.
Parker Dillmann (10:37):
It affects everything that you
Stephen Kraig (10:39):
Timing circuits. If you are trying to couple an
AC signal from one part of yourcircuit to another part of the
circuit, this can adddistortion. There's so many
different things that you know,this one if you weren't aware of
this and your circuit startedacting strange, it could take
you forever to figure out whatthis is. So Mhmm. Let's let's
(11:00):
let's sort of define it.
So voltage coefficient ofcapacitance. Basically, if you
apply a DC bias to a capacitor,its capacitance value actually
changes. And changes as indecreases, not always. There are
some situations where itactually increases, but for the
most part, it is a decrease, andit can be a significant
(11:23):
decrease. We're not talkingabout, you know, 1, 2%.
In some cases, it can be 85%,90% across the voltage range of
a capacitor. So that lovelylittle 1 microfarad you got and
you that you think is 1microfarad in your circuit, if
you apply, you know, aparticular voltage across it, it
(11:43):
could be acting like somethingsignificantly less if you're not
aware of this. It has differenteffects at DC and AC as well.
And so even if you havecombination of DC and AC, it has
different effects. For the mostpart, though, any data you find
will show, what happens across,with a DC bias on there.
(12:05):
Something that's alsointeresting to note is when you
look up a data sheet for acapacitor, the you will be
provided with a capacitancevalue. You should also be
provided that datasheet shouldinclude the test conditions at
which they determined thatcapacitance. So because if you
(12:26):
are trying to replicate that orif you are trying to test a
capacitor yourself, if you testat different parameters, like a
like testing at a differentfrequency or testing with a
different DC bias with whateverLCR meter you're using, you very
likely will get a differentvalue. So that's important to
keep in mind if you ever do needto test capacitors. If you're
(12:48):
comparing what you're monitoringor measuring, you or if if
you're looking at it to validatesomething on a datasheet, make
sure you're using the exact sametest conditions exactly because
of this voltage coefficient ofcapacitance.
Parker Dillmann (13:00):
Yeah. My experience with those testers is
they use a low voltage, and thatmight be because or maybe not
because of this, but it helpsthem more accurately represent
the capacitance that because ofthe the that that it's almost
(13:20):
like a negative coefficient,basically.
Stephen Kraig (13:23):
Yeah. Effectively, it is. If you zoom
in on the the charts at DCvalues that are really close to
0, it can actually increase incapacitance. And it's important
to note that, you know, we weretalking about class 1 and class
2 capacitors earlier. Thisvoltage coefficient of
capacitance only applies toclass 2 capacitors.
So your x seven r's, your x fiver's, all of those type of
(13:47):
capacitors. It only applies tothose. It does not apply to
class one caps. And here's thething, it applies to them
because this is a materialprocess. This or not a material.
This is due to the material thatthat they're made of. So, I
mentioned earlier that class 2capacitors are made of barium
tight, titanate. In my research,I found some really cool stuff
(14:09):
about this. So, when capacitorsare made, I think James Lewis
even went through this. In fact,yeah.
He he did back in episode 141,because he was asking us
questions about how long does ittake to make a capacitor, which
if I remember right, it was 2weeks.
Parker Dillmann (14:26):
Something like that.
Stephen Kraig (14:27):
I think it was 14 days, something like that.
Regardless, there's a kilmingoperation where they have to
heat and cool the capacitors,where they're basically drying
out the ceramic material andsolidifying it.
Parker Dillmann (14:40):
Yeah. They're making capacitor brownies.
Stephen Kraig (14:42):
Exactly. So when the capacitors are hot, the
crystal and structure of thebarium titanate forms what's
called a cubic crystal. Sobasically, like a diamond shape.
And barium titanate has barium,oxygen, and titanium in it.
Above the curie temperature ofthe material the the titanium
(15:05):
atom actually sits at the deadcenter of that diamond lattice
shape.
And because of that actually,let me let me let me move on to
the what happens when it whenthe temperature goes below the
curie temperature. The crystaland structure actually morphs
slightly. So as the capacitorcools, the crystal forms a
(15:26):
different shape which isactually called a tetragonal
crystal. So, think of takingthat diamond shape and
stretching it out a little bit.And the interesting thing about
it is during that process whenit's going from the cubic
crystal to a tetragonal crystal,that titanium atom that's at the
center of the diamond displacesand moves ever so slightly.
(15:49):
And that actually causes thecrystal lattice in the capacitor
to become unbalanced, and itforms an electric field within
the the crystal and structure.Now as it's cooling and
annealing effectively, you get abunch of grain boundaries and
and areas with that all canceleach other out. So, yes, you
(16:12):
have individual e fields, butthey all cancel each other out.
But perhaps you can see whereI'm going with this. That all
occurs or happens when thecapacitor is effectively at 0
volts.
But now you have all of thesedipoles created inside of this
crystalline lattice. So once youstart applying an external e
(16:35):
field, which it's not externalbecause we're talking about the
actual you're applying a voltageto the plates, you're adding an
e field inside the crystal andlattice. Now the individual
dipoles of every crystal withinthere start fighting against the
DC bias that you're applying toit, and that actually reduces
(16:56):
the capacitance, because nowyour electric field is being
modified by every single dipolewithin there. So it's
fascinating that just the slightmovement of a titanium atom as
the crystal cools is what is thecause of this voltage
coefficient of capacitance.Because you're literally setting
(17:18):
up a whole bunch of tiny efields that resist when you're
applying an e field.
Parker Dillmann (17:25):
Didn't a couple podcast episodes ago, we talked
about barium titanite.
Stephen Kraig (17:32):
I I would not be surprised if that's used in the
production of many othercomponents.
Parker Dillmann (17:38):
Well, it was with capacitors. It was when was
this?
Stephen Kraig (17:44):
Oh, was this that
Parker Dillmann (17:45):
It's like the new capacitors That
Stephen Kraig (17:47):
was that that battery. Right?
Parker Dillmann (17:49):
Yeah. It's it's a new capacitor to help offload
charging cycles off batteries.Mhmm. Yeah. And that was they
were using barium basically,it's the same stuff, but I think
they fixed they were able to,like, fix something with the
structure.
I wonder if that's what it was.
Stephen Kraig (18:09):
Okay. So I have done a bit of research on it
seems like most people or mostmanufacturers of MLCCs use
barium titanate as the the thefundamental ceramic. But
everyone has their own 11 herbsand spices that they that they
dope it with basically in orderto adjust the parameters based
(18:31):
off of your capacitor, whateveryou're looking for. And so maybe
there's a slightly differentformula that they used for
whatever that charging capacitorwas.
Parker Dillmann (18:43):
Yeah. They're saying that it's they're
basically modifying the 3 dstructure of it, and then
they're sandwiching it between 2others materials, which might be
your the 11 herbs and spices
Stephen Kraig (18:56):
Exactly. And And maybe those extra little dopants
reduce this effect or change itin whatever way.
Parker Dillmann (19:04):
But that now it makes more sense when we're
about that article. That wasepisode 431, so it was Oh,
Stephen Kraig (19:11):
that was just a few weeks ago. Yeah.
Parker Dillmann (19:12):
It was a month ago at this point. That's why I
was, like, I remember titanatebecause I couldn't pronounce it.
Yeah. And so, yeah, that that'sso I was wondering why because I
saw a bit of titanate. I'm like,oh, that was I know this is in
capacitors.
We talked about this a monthago. And so they must have those
researchers must have been ableto fix that problem, or they're
(19:36):
using it in a different way.
Stephen Kraig (19:39):
Here's the here's the thing. Because because
capacitors are so inexpensiveand they're so prolific, they're
everywhere, that just makes methink that barium titanate is
probably not that expensive.Because for the most part with
manufacturing, everyone's goingto standardize on the cheap
thing that just works. Right?Mhmm.
(20:00):
So so I wouldn't be surprised ifthere are methods to fix or or
correct or or maybe modify thisvoltage coefficient of
capacitance, but it may beexpensive. The additive that you
add in there may, you know, blowthe blow the lid off of how
expensive capacitors are. Idon't know. We'll we'll see.
(20:22):
Because because if that's thecase, then start throwing this
into everything and now yourclass 2 capacitors all become
class ones and Mhmm.
I mean, not necessarily.Regardless. So so yeah. To kinda
get back to where, you know,we're we're talking about
selecting capacitors and nowwe've gone off under this
voltage coefficient ofcapacitance thing. The whole
(20:43):
point of bringing this up isthat this is something that
doesn't necessarily isn'tnecessarily in your head when
you go to Digi Key and clickpoint 1 capacitor 50 volt 0 805.
Right? And you're like, oh,cool. I I get it.
Parker Dillmann (20:57):
Sort by cheapest. Right.
Stephen Kraig (21:00):
If it if it's COVID, sort by most available.
Right?
Parker Dillmann (21:03):
Yeah. Yeah.
Stephen Kraig (21:06):
2018 was dark times. It was rough. It was real
rough.
Parker Dillmann (21:10):
That was the capacitor shortage era. Yeah.
That COVID was the activecomponents shortage. Yeah.
Stephen Kraig (21:18):
So I wanna bring up I wanna bring up something
from Murata that is pretty cool.If somebody hasn't seen this
before, it's called sim surfing,s I m surfing. Murata has a
database of all of theircomponents or all of I guess,
all of their components. I'veI've mostly only looked at
capacitors. But in thisdatabase, it provides you with
(21:42):
much more information than whatyou get out of just the data
sheets.
So on sim surfing, you canactually get the curves that
show this voltage coefficient ofcapacitance. And so I I had some
fun earlier today. I went andgrabbed 2 Murata capacitors. 2
with all exactly the sameparameters except one was an
(22:06):
0805 size and one was an 0402size. And so if you're just an
engineer going out to pick a capand just you you you select all
the parameters the same, What'swhat's the difference between an
0805 and an 0402 other than thefact that one's bigger than the
other.
Right?
Parker Dillmann (22:24):
Yeah. One's twice as big.
Stephen Kraig (22:27):
So, you know, obviously, this is a podcast.
Our listeners won't be able tosee this, but I actually in in
our show notes, I posted 2graphs that show the voltage
coefficient of capacitance forboth this 0805 and this 0402.
And across 0 to 50 volts, the0805 has a reduction in in
(22:48):
capacitance of about 40%. And incomparison, across a voltage of
0 to 50 volts on the 0402, ithas a change of capacitance
close to 90%. So all parameters,exactly the same, dielectric.
They're even made of the samematerial, just different size
(23:10):
capacitor has a huge impact onthe voltage coefficient of
capacitance. And that's notsomething like, I keep saying
this. You're not gonna find thatin the datasheet. You just have
to know it, it, or you have tolisten to circuit break and
you'll you'll learn.
Parker Dillmann (23:22):
Right? Yeah. You have to be a breaker.
Stephen Kraig (23:24):
You have to be a breaker. That's so so and a lot
of this is what's got methinking more recently about
talking with my team at workabout how do you pick a
capacitor? How do you pick howdo you know when to say 0 805
versus 0402 versus 0603? Becauseright now, I could go to Digi
Key, and I could find capacitorswith the exact same parameters
(23:47):
in 10 different sizes. Which onedo I pick?
You know, I've been designingboards for over a decade now. I
like 0 4 0 twos. They tend towork with the ICs that I play
with. They're small enough thatthey fit the pitches of a lot of
the ICs I I go with. But sizeisn't just everything.
(24:07):
Right? And this is one reallygreat example of that where,
yeah, you could be picking apass a capacitor and you could
be completely screwing yourselfover by doing that.
Parker Dillmann (24:16):
Yeah. It's I don't know if you're gonna get
to this, but, well, I Itypically do too because I use,
you know, 0 42s for bypasscapacitors a lot.
Stephen Kraig (24:26):
Mhmm.
Parker Dillmann (24:26):
0.1 microfarad bypass capacitors. What I'll do
is I'll pick a much highervoltage so my, like, my rail
will be 3.3 volts and I won'tpick a 6 volt capacitor. I'll
pick, like, a 16 volts because Iknow that that curve, I'm gonna
be in the better part of thecoefficient curve, so to speak.
Stephen Kraig (24:47):
That's that's that's a good way of handling
it. Yeah.
Parker Dillmann (24:49):
And I I actually I learned that on the
episode with James Lewis,episode of 41, 141. So that
sounds like,
Stephen Kraig (24:57):
oh, you you basically prederate your design.
Parker Dillmann (25:01):
Yes. You have to prederate just because of the
voltage ratings on thesecapacitors or if you can swing
the price and like, 0.1microfarads is a very small
value, and so you can probablyget a class 1 that's in an 042
that fits that.
Stephen Kraig (25:19):
Yeah. Maybe. Maybe. Class class ones, they
they I bet you could. It's beena while since I've looked for a
class 1.1 microfarad, but that'sstarting to get to the point the
upper limit of what class 1 capscan handle.
So this voltage coefficient ofcapacitance, it is obviously
(25:40):
material dependent. Right? It'sbarium titanate, but it is also
dependent on your capacitiveplate spacing. So let's take a
second and just think aboutwhat's inside of a capacitor.
Yeah.
I mean, it is a it's a brick ofceramic, right, with a bunch of
(26:00):
plates inside. In fact, if yousee, like, a cross section of a
capacitor, it's surprisinglysimilar to the drawings you had
in your textbooks. Right? It'sliterally just plates next to
each other with ceramicdielectric in between them.
Nothing particularly special.
So let's think about the knobsthat you can turn to adjust
(26:22):
parameters on a capacitor. Let'sjust play with an 0402 in our
mind real quick. Let's say I'm acapacitor manufacturer and I've
got an 0402 that I wanna designand it's a, let's say, it's a
point 1 microfarad. And so Idesign it up and blah blah blah.
And then marketing comes downand was like, oh, no.
That was supposed to be a 1microfarad capacitor. What are
(26:45):
the knobs that I have to turn inorder to make my design go from
1 to sorry, 0.1 to 1 farad, Ican add more plates into the
capacitor. Right? Or I can getthe plates closer to each other.
Parker Dillmann (27:01):
I'm gonna derail you a little bit. Yeah.
But you are right. So I'm why amI why am I on Digi Key what's
dotbe Belgium? I don't know.
Why am I on Digi Key's Belgium?Yeah. It is Belgium. Why am I on
Digi Key Anyways, I'm on theBelgium Digi website.
Stephen Kraig (27:19):
Yeah. Finding a class 1.1.
Parker Dillmann (27:21):
A class 1.1 microfarad, the smallest size is
apparently o seven 4, which I'venever heard of that, but 1206.
Stephen Kraig (27:30):
Yeah. Right. So if you want What's o
Parker Dillmann (27:33):
seven zero four?
Stephen Kraig (27:35):
Isn't that like a military designation? I think it
might be.
Parker Dillmann (27:39):
I'm looking. I've never heard of it before.
Stephen Kraig (27:42):
Yeah.
Parker Dillmann (27:44):
Like, keep going. Sorry.
Stephen Kraig (27:45):
But but yeah. So that's a good point. A quick
tangent real quick. Class 1capacitors with NPO, COG,
dielectric, or tempco is youyou're limited to the high
picofarad range. You're you'renot gonna get much past the
nanofarads range withoutgetting, you know, a 1206.
(28:07):
And then at that point, youknow, if if your circuit can
handle it, great.
Parker Dillmann (28:11):
But They're expensive too.
Stephen Kraig (28:13):
They're they're expensive. Class 1 is usually
your low value stuff.
Parker Dillmann (28:18):
Yeah. When I'm building a board or designing a
circuit, I'll use a class onefor, like, my loading capacitors
on my oscillators, like, mycrystal oscillators.
Stephen Kraig (28:30):
Oh, yeah.
Parker Dillmann (28:30):
Because those are also they're, like, 0.0047
picofarad or something likethat. I can't remember on top of
my head.
Stephen Kraig (28:36):
Well and and what's funny is at a certain
point, it's actually hard to notget a class 1 capacitor. Yes.
Like, at a certain point, theyjust all are class 1. So Yeah. I
shouldn't say all, but
Parker Dillmann (28:48):
it's 27 picofarads or something like
that. Right. Depends on thecapacitor in the circuit, of
course. But
Stephen Kraig (28:53):
Yeah. Okay. So so so back to the back to that
little game we're playing. If Iif I'm making an 0402, if I want
to up the capacitance, so I sayI'm the designer of the actual
capacitor, I can add more platesto a capacitor, or I can get the
plates closer to each other. Ireally can't change the size of
the plates too much because I'mfixed to my 0402 size.
(29:17):
Right? If you think about it,those actions that I'm talking
about, adding more plates orgetting them closer together end
up being the same or having thesame result. And if my voltage
coefficient of capacitance ishighly dependent upon your
capacitor spacing, then just byupping that capacitance via
(29:39):
those options that I've given,I've actually made my voltage
coefficient of capacitance evenworse. So something to keep in
mind when selecting a capacitor,if you have a given package
size, say, 0402, if you'relooking at capacitances on the
high end so, basically, takeyour little you the slider on
(30:00):
Digi Key or the filter slider.Look at the far extreme side
where it has the highest values.
Those are gonna have your worstvoltage coefficient of
capacitance. And then you haveto weigh, do I pick an 0402 or
do I pick whatever size that hasa terrible VCC or do I just bump
up in size? Because if you thinkabout it, like, the next knob,
(30:23):
in terms of me being amanufacturer of the next knob I
would have would be making myplates bigger. If I make my
plates bigger on my capacitor,then I don't have to get them
closer together. I can get abetter VCC by just upping the
size of my capacitor.
So just like those charts Ishowed where I have an 0402 and
(30:45):
0805, it's proof of this showingthat the 0805 has a much lower
VCC than the 0402. Because on inthe 0402 to get the same
capacitance density, they haveto shove more plates closer
together, which leads to a worseVCC.
Parker Dillmann (31:02):
I gonna bet you that's because with the plates
being closer together and moreof those plates probably, you
have more the effect of thattitanium atom shifting around.
Stephen Kraig (31:14):
I think so,
Parker Dillmann (31:15):
like, those interactions between the charge
and those atoms are closertogether.
Stephen Kraig (31:21):
You also have a higher density of the e field in
a more localized area. So theeffect that each individual e
field from the crystallinelattice gets compounded. Mhmm.
So if you want a lower VCC, picka larger capacitor for the a
larger sized capacitor for thesame capacitance value. What's
(31:44):
interesting is kind of rule ofthumb here actually applies to
almost all the parameters on acapacitor.
If you think about it, gettingthe plates closer together will
actually reduce your voltagerating on the capacitor because
now you have less of an of ajump for things to arc across or
for, the dielectric to breakdown. So capacitors with a
(32:08):
higher capacitance value in asmaller package will inherently
have a lower maximum voltagerange unless those 11 herbs and
spices help out. There's alwaysexceptions to all of these rules
that, manufacturers have come upwith. But in we're talking
about, like, generic capacitorshere.
Parker Dillmann (32:29):
Yeah. I'm looking at this this chart
between this 0 805. So the worstcase for that 0805 is negative
40%. If you basically appliedlike let's say negative 40% was
like the worst you could sufferthat actually makes that 042
capacitor actually only reallybe a 16 volt capacitor instead
(32:52):
of 50 volt.
Stephen Kraig (32:54):
Right. Right. Yeah. And in fact, the 0 805 at
at 10 volt DC applied is almoststill the same value. Yeah.
Maybe there is some kind of ruleof thumb where it's like if you
apply 1 fifth of the voltage,then you're within some reason.
I don't know. I don't I'm notsure about that, but it is kind
of interesting that that simsurfing website that's the the
(33:18):
it's made by Murata. It's it'ssort you only find Murata parts
on it, but you can get all ofthis data and you can see all of
this data and you can get,printouts of every single
capacitor you have. So if youreally want to fine tune your
circuit and know every littlething about it, they have all of
their charts on there.
Parker Dillmann (33:36):
Yeah. My rule of thumb is I just I I guess
I've never really used ceramicsand high voltage high voltage.
I'm doing quotes, like, over 24volts. I always pick, like, a 16
volt capacitor and whatever sizeis working on my board for,
like, a 3.3 or 5 volt rail. Thatway I know I'm in the better
part of these curves.
Stephen Kraig (33:57):
Yeah. Yeah. You're in the the closer to 0
part. Yeah. Yeah.
So
Parker Dillmann (34:04):
50 volts seems like a lot for, like, a point 1
microfarad part.
Stephen Kraig (34:08):
It is. It is. But I just picked
Parker Dillmann (34:11):
a voltage fan.
Stephen Kraig (34:13):
Yeah. But but yeah. I mean, you could run this
exercise with basicallyanything.
Parker Dillmann (34:17):
Mhmm.
Stephen Kraig (34:17):
Now this this brings up another point that we
hammered home in episode 141,but I wanna I wanna talk about
it again. And it was do not pickthe extremes or avoid the
extremes. So, say, you you'regoing to find an 0603 extremes.
So, say, you you're going tofind an 0603 cap, and you have
voltages from 6.3 volts all theway up to 500. Don't pick 6.3
(34:40):
and don't pick 500.
What I mean by that is you havethe extremes that the
manufacturers say they canmanufacture all the way across.
If you pick something that's onthe edge of them, first of all,
it becomes more difficult tomanufacture those parts
typically at the edges of theparameters. And if that part
(35:01):
ever goes obsolete, the chancethat you would find a substitute
for that becomes that much moredifficult. That actually kind of
plays into what I was sayingearlier, like, the decisions you
make flow downstream to everyoneelse. So your purchasing
department or your contractmanufacturer is gonna get kinda
grumpy if you pick an 0201 1000volt capacitor and they have to
(35:24):
match those parameters.
I don't even think that exists,but you get where I'm going with
that.
Parker Dillmann (35:29):
Right? Yeah. I I wonder the these those edge
case capacitors. I would love tosee the VCC curve on, like, you
said the 6.3 volt capacitorbecause it might just be, like,
flat. Right?
It could be really
Stephen Kraig (35:43):
good. Yeah.
Parker Dillmann (35:44):
Yeah. It could just be they bend up that part
can be, you know, really good,and we just don't know about it
because we have to go in a simsurfer
Stephen Kraig (35:52):
and check it out. I'm actually on it right now.
Parker Dillmann (35:55):
But I actually only pick parts that I can get
that curve from. Oh. Well, Ijust like I reward the
manufacturer that's puttingforth the effort
Stephen Kraig (36:07):
I Like that that's a good way of putting it
So
Parker Dillmann (36:09):
because I think it's a very important value that
you know, this is also somethingthat should be taught in school
too because this was not taughtto me in school at all.
Stephen Kraig (36:17):
Well, no. I mean, the thing that was taught to me
in school was capacitance value.That's it.
Parker Dillmann (36:23):
Yeah. That's it. And then we had voltage as
well, but that was it.
Stephen Kraig (36:27):
Yeah. I don't even remember if I had voltage.
Okay. I'm looking at okay. Sothis is somewhat of a ridiculous
capacitor.
I'm looking at a 6.3 volt. It'sa 1206, so it's a pretty big
one, but the the cap value is220 microfarads. So it is a
gigantic cap. And across 6.3volts, it drops 80% capacitance.
(36:51):
Oh.
Yeah. But
Parker Dillmann (36:53):
So you could use that
Stephen Kraig (36:55):
at, like, 1 volt? Yeah. Or just be okay with the
fact that you're gonna have aserious
Parker Dillmann (37:02):
But I'll put this way is, you know, you're
getting what? You're at a 220microfarads, you're getting 44
usable microfarads out of thatcapacitor. So if you've
calculated your feedback loop onyour op amp, right, to and
(37:23):
you're like oh, yeah, I need 220microfarads, and oh, let me go
pick this capacitor. It'sactually gonna be 44
microfarads. And so you'rewhatever you're doing with your
feedback loop is gonna be wayout of whack.
Stephen Kraig (37:36):
So I I found another cap. This is an 0 6031
microfarad, and it is a 6.3volt. And across 6 volts, it
only drops slightly less than20%. So it it has so much to do
with capacitance density. Youknow?
If you're shoving to get 220microfarads into a 1206, that
(38:00):
thing is jam packed on theinside, and I guarantee you
those plates are really closetogether.
Parker Dillmann (38:05):
Yeah. I I'd love to everyone out there
that's listening, all youbreakers. What would you a
capacitor like that for? A 6.3volts 220 microfarad capacitor
no, ceramic that loses near itstop end 80% of its capacitance.
(38:25):
I'd love to know what theapplication is for that part.
Stephen Kraig (38:28):
You know what what what comes to mind is maybe
some kind of, like, voltagereference filtering if that
voltage reference was, like, 1.2volts or something like that.
Parker Dillmann (38:38):
Yeah. That could be it.
Stephen Kraig (38:39):
I could see something like that.
Parker Dillmann (38:41):
Yeah. A very low voltage filtering. Yeah. But
let's I'll let them know ifsomeone out there has used these
parts. I'd love to know what theactual application was.
Stephen Kraig (38:51):
And they're like, oh, that's why my circuit wasn't
performing like I thought.
Parker Dillmann (38:54):
I could be.
Stephen Kraig (38:56):
So so yeah. So stay away from the extremes.
Don't pick the lowest voltageand don't pick the highest
voltage of capacitors with whenit comes to values, do the same.
Don't pick the highest valuecapacitor in the smallest
package you can. There's a lotof reasons behind that.
I mean, we just went through allthe VCC problems, but there's
(39:19):
also a lot of difficulty inmanufacturing with those. And
what's interesting, this isthere's difficulty in
qualification of those parts. Soa lot of parts have to go
through life, qualificationtesting, a lot of temperature
testing, and a lot of extrastuff in order to meet whatever
extra ratings they have, likeautomotive and things like that.
(39:42):
And if what's funny is if you goto Digi Key and you just select
aecdashq200. So the this thispassives automotive rating, it
automatically eliminates a wholebunch of capacitors that would
have been available.
In fact, it takes about 2 thirdsof your capacitors out of Digi
(40:04):
Key, and that's not necessarilysaying that those parts are are
bad in any way. But what it doesis it kinda gets rid of the
extreme cases that would notpass the qualifying test for
being automotive. So it sort ofgets rid of the extremes and the
edge cases for you by justselecting that. But James Lewis
(40:26):
had a really interesting pointwhen we talked to him last about
this, about how capacitors startout without a capacitance value,
and then they kiln them and theykind of level out to a
capacitance value after they'vedried out in the in the ovens.
And that's a fully chemicalprocess.
(40:46):
It's not like something that youcan just turn the knob and,
like, move a thing in there.It's done. Right? It's a
chemical process, and so they'retrying to hit hit a target and
if you're trying to hit a targetthat's on the edges, that has
that's the hardest for themanufacturer to actually hit. So
it makes way more sense to tryto pick a value for both for not
(41:07):
both, for capacitance, for yourvoltage, and for your tempco
that fits more within the centerof the bell curve for any
capacitance size.
So in other words, like, a in a402, if you're trying to pick a
10 microfarad or 402, that is byfar an edge case. So that's not
going to be something very easyfor the manufacturer to create.
(41:31):
And so hitting all of theirtargets with that, all the
parameters with it is verydifficult. That would be a
really good indicator to notselect that size of capacitor.
Now if we're talking about a 100picofarads in an 0402, now we're
talking.
Now that's something that'sright in the center of what
makes sense for that size ofcapacitor. Now, at the same
(41:51):
time, if you have, say, find it,like, a 1210, a big cap, and
it's a 1 picofarad cap in there,you have the exact opposite
thing even though the resultsare the same. Like, how do you
make 1 picofarad in a bigcapacitor? You're gonna have 2
plates inside. I'm just I Idon't know exactly the amount,
(42:13):
but you're gonna have, you know,a lot fewer plates and they're
gonna be spaced differently.
It's that something like that isalso hard for the manufacturer
to hit on the low side ofthings. So that would be a great
example of just like the 10microfiber 4 2, the opposite
picking a really low valuecapacitor in a big package is
(42:34):
hard for them to hit. So it allcomes down to get a gut feel for
what the bell curve of everypart looks like in terms of
their spread and try to sticknear the center of that. And
don't be afraid of bumping up tothe next value or the next size.
Yes.
(42:54):
It is nice to have smallcomponents. They fit into your
layout and, you know sure. Iagree. I've I've been there. But
in terms of being able to getparts that consistently work,
and if they go obsolete, you caneasily find substitutes for them
and you don't get bit in thebutt by problems with VCC, try
(43:16):
to stick to the center of thebell curve.
That's sort of the big maintakeaway here.
Parker Dillmann (43:22):
Larger components also enable you to
route between pads.
Stephen Kraig (43:27):
That's true. That's true.
Parker Dillmann (43:29):
I think we we used to call those the sneaky
traces on this podcast a longtime ago.
Stephen Kraig (43:34):
I remember that. I remember being asked, are you
doing sneaky tracing?
Parker Dillmann (43:38):
Yeah. Because at o four zero two level, you
it's I don't think you can fit atrace in there. I need to try,
actually. I wonder if
Stephen Kraig (43:46):
there is With standard manufacturing, yeah, I
don't think
Parker Dillmann (43:49):
you can. With 3 mil, 3 mil, I wonder if you can
sneak 1 in. Maybe not. That's, 9mils total.
Stephen Kraig (43:56):
I don't think you can. Unless you compromise on
your footprint a lot.
Parker Dillmann (44:00):
Yeah. And then then you get more tombstoning,
which is not really acompromise.
Stephen Kraig (44:06):
No. I you lose way worse.
Parker Dillmann (44:09):
Yeah. That's way worse. Yeah. I do like
designing with one size in mind,but capacitors are always a
different game. Usually, I'lltry to pick I do usually do,
like, 042.
Sometimes I'll bump up the sizebecause of the like, let's say a
point 1 microfarad or whateverbecause it makes sense for,
like, the trace it's near. Like,let's say it's on my output of a
(44:31):
LDO or whatever, and so I'llhave, like, a 10 microfarad cap,
which is like a, you know, 1206size. But then next to it, you
know, I need to have, like, asmaller value, like, a 0.1
microfarad or 1 microfarad. AndI'll actually pick probably the
same size if I can just forthermal loading when it reflows.
(44:53):
So that little part doesn't justget overcooked.
Right?
Stephen Kraig (44:57):
Yeah. Yeah. Yeah. That makes sense. So that's
There's also heat shadow.
Right? Yeah. So it might not getcooked enough if it's next to a
big component.
Parker Dillmann (45:08):
Yeah. That that's weird stuff. You just
have to learn by just designinglots of boards. Yep.
Stephen Kraig (45:15):
So Yeah.
Parker Dillmann (45:17):
I got nothing else to add to this.
Stephen Kraig (45:19):
Yeah. I think I think I've vomited enough
information out.
Parker Dillmann (45:22):
Let us know. I really wanna know people who use
those edge case components,especially that, like, 6.3 volt
220 microfarad that loses 80%over over that 6 volt range.
Stephen Kraig (45:38):
Also, did you know that it's losing 80%?
Parker Dillmann (45:41):
Yeah. Yeah. Let us know. What did you use that
in the design for? So so thankyou for listening to circuit
break for macro fab.
We were your hosts Parker Dolan.
Stephen Kraig (45:53):
And Steve and Craig.
Parker Dillmann (45:54):
Later everyone.
Stephen Kraig (45:55):
Take it easy.
Parker Dillmann (45:56):
Thank you. Yes. You breaker for downloading our
podcast Tell your friends and coworkers about circuit break the
podcast from macro fab leave usa review, too If you have a cool
idea project or topic you wantus to discuss, let Steven and I
in the community of breakersknow. Our community where you
can find personal projects,discussions about the podcast,
(46:17):
and engineering topics and newsis located atform.macfabdot com.
Welcome to circuit break from MacroFab, a
weekly podcast about all thingsengineering, DIY projects,
manufacturing, industry news,and choosing capacitors. We're
your hosts, electricalengineers, Parker Dillmann.
Stephen Kraig (00:24):
And Stephen Kraig.
Parker Dillmann (00:25):
This is episode 434. And before we jump into the
podcast, we have a macro fabannouncement or an update so to
speak. We now offer bare slashunpopulated PCBs. So forever
we've always done PCB assemblyand we've kind of shied away
(00:46):
from just supplying bare boardsto customers but we've had
enough customers essentiallyrequest for them to able to buy
bare PCBs and so to explain moreabout that I have Kyle from
product to talk about that.
Kyle McLeod (01:01):
We're really excited to offer bare PCBs to
our customers. This is somethingour customers have been asking
for. And we found that we have areally, good position to offer,
high quality boards, boards witha lot of custom specs,
engineering support on gettingyour boards manufactured and
really, our goal is to give youthe opportunity to, you know,
(01:23):
order a board that you areconfident in when you're doing
your prototyping, that the boardyou get is gonna be, is gonna
work, is gonna be usable whenyou go to add your components to
it and then you're not gonnahave to start over or spend a
lot of time questioning and didyou make a a mistake or is there
something wrong with the boardthat the manufacturer didn't
(01:45):
capture. So we think we havereally good, high quality
boards. We have lots of customspecs.
And really, I think the mostexciting thing is is that when
you go to scale to production orif you're in production and you
wanna iterate on your design, wehave made it really easy in the
platform to switch between abare PCB and your PCB with
(02:06):
components. So, if you'veuploaded your PCB and you've
added your bill of materials,you don't have to go and create
a different PCB to order thebare boards. You can just turn
off the components for yourorder and we'll let you order
that bare board. When you wantto view the components again,
you turn them back on and allyour bill bill of materials, all
your placement data is stillthere. And so you can work, on
(02:29):
your on your designs, keep keepyour designs intact without
having to maintain a separatebare PCB file and a, assembled
PCB file.
And we think that's gonna makeit really easy for our
customers. Another thing, byhaving all of the same offerings
from our specs and from ourmanufacturers. We allow our
(02:53):
customers to easy transitioneasily transition from prototype
to production because you don'thave to upload new files, you
don't have to wonder, you know,if if you were to do a prototype
with 1 shop and then shift tosomebody else, you know, they're
gonna have differentrequirements. And so, by doing
everything with Macrofab, youcan keep all of your your data,
with the same person. We make itreally easy to add those
(03:14):
components on and order yourassemble boards, once you're
ready for production.
So, yeah. Bare PCBs is somethingwe're we've been very excited
about. We think it's a reallynice, smooth feature for
prototyping and for foriterating on your your
production products as well.
Parker Dillmann (03:29):
So, yeah. Check out our new custom PCB service
on our websiteatmacrofab.com/customhyphen PCBs
and take your projects to thenext level. Whether you're
prototyping or ready forproduction, MacReb has you
covered. Okay. Let's jump intothis capacitor selection topic,
Steven.
(03:50):
What what is this about?
Stephen Kraig (03:52):
You know, it's at work. We've been talking about
this recently. Just the idea ofhow do you pick a capacitor.
Like you go to digi key, you goto mouser, you go anywhere And
as a designer, you know you needa particular value. You know
some of the other parameters youneed.
But how do you go about actuallyselecting a an appropriate
(04:16):
capacitor? And it ends up beingquite a bigger thing than you
think. There's every decisionyou make flows downstream to all
the people that have to dealwith your decision. And I like
framing it that way because, youknow, we as electrical
engineers, a lot of times canget pigeonholed in our thinking
(04:37):
of I found this perfect thingthat works for me, and this is
it. That's it.
It's done. I as long as I justput it on my bill of materials,
put on my schematic, my work isdone, wash my hands. But there's
so much more to it, you know,when it comes to purchasing,
when it comes to actuallygetting it, when it comes to
putting it on the board, when itcomes to validating it. There's
(05:00):
so much more to that. So Iwanted to touch on a handful of
topics with that and maybe bringup some stuff that that we
haven't really spoken aboutbefore.
Way back when I don't actually,I don't remember the date, but
but on episode 141, so manyepisodes ago, we had James
Lewis, otherwise known as theBald Engineer, come on. The
(05:21):
episode was called It Depends,an in-depth look at MLCCs with
James Lewis, where we weretalking about some similar
concepts of what we'rementioning today with selecting
capacitors. James Lewis has someexperience working with KEMET in
the past, and and he reallybrought some really cool
information about how capacitorsare made and a lot of other good
(05:44):
information, and I wanna kindajump off that today since we
haven't talked about this in awhile. So first of all,
capacitors kind of come in infact, let let me stop by saying
capacitors come in a ton ofdifferent varieties. Today, I
just wanna talk about MLCCs,multilayer ceramic capacitors.
And that's your, like, bogstandard tan looking capacitor
(06:08):
that you see on a bazilliondifferent boards. There are so
many other variants of howcapacitors are made that we're
not going to be covering today.Really wanna just look at MLCCs.
Now MLCCs come in 2 differentclasses, class 1 and class 2.
Class 1 is your COGNPOdielectric capacitor.
(06:32):
So that should be TEMCO thatthey, that is part of the
material that they're made of,which is calcium zirconium
oxide. The pros of class 1capacitors is that they have
really good temperaturestability. They have really good
aging effects, and theybasically are very stable
(06:53):
capacitors. The cons to these isthat they typically have a
limited capacitance range.You're not really gonna get high
capacitance values in a classone capacitor, and they tend to
also be in the smaller sizerange.
So, you know, if you're lookingfor a 100 microfarad 1210
(07:14):
capacitor, just pretty muchforget class 1 and forget the
benefits you get from having, agood dielectric and a stable
capacitor. You you're prettymuch relegated to class 2. Class
2 caps are made
Parker Dillmann (07:29):
of They are also more expensive typically.
Stephen Kraig (07:33):
Typically. Correct. Typically. Class 2
capacitors are made of bariumtitanate, and that's the bulk of
your capacitors. The x seven r,the y five v's, the just the
generic capacitor that you findeverywhere is almost guaranteed
to be a class 2 capacitor.
(07:54):
So the one the main topic that Iactually wanted to kind of break
into today was something thatwas mentioned back in episode
141 but not really expandedupon, and that is the voltage
coefficient of capacitance. Soit may not it's interesting
because this topic is somethingthat isn't called out on
(08:16):
datasheets and it doesn't evenhave a standard way of
determining it. But class 2 MLCCcapacitors have a voltage
dependence to their capacitance,which I find somewhat
interesting that it's that thischaracteristic of capacitors are
(08:39):
is not on data sheets. It's notcalled out. It's not defined.
But if you know it exists, youknow it exists. This this kind
of stuff that actually gets tome when it comes to new people
getting into electronics becauseI remember when I was first
learning electronics,capacitors, like, after I
finally got a grip on what aresistor actually is, I'm like,
(09:01):
okay. Let's look at this nextcategory. And it was capacitors,
and it's like, oh my god. Like,you you 10 x the parameters that
I have to care about now, and Ihave no idea what any of them
are.
And without even being told,like, the data sheet doesn't
even mention it, the actualparameter that you care the most
about, the capacitance, is afunction of the voltage that you
(09:25):
apply across them, which is mindboggling, but it's something
that you have to keep in mind.So in in research for this, I
actually was not aware of this,but there is a body called the
Electronics Industries Alliance,the EIA. And this is one of the
bodies that sets the names ofthe parameters of everything
(09:47):
that goes into a datasheet. Ishouldn't say everything, but, a
a good chunk of them. So whenyou say, like, capacitance or
whatnot, it's defined, like,this is one of the bodies that
says, here's what capacitance isand capacitor should or a
capacitor datasheet shouldinclude that.
And it's interesting becausethis voltage coefficient of
(10:09):
capacitance just is not part ofthe EIA whatsoever. So this can
really catch newcomers. And infact, I I think there's plenty
of people who are well pastbeing a newcomer who probably
are not aware of this justbecause you don't know about it
unless you know about it.
Parker Dillmann (10:26):
Yeah. And the thing that this would affect the
most is, like, on feedbackcircuits.
Stephen Kraig (10:33):
This is yes. This this can affect virtually
everything.
Parker Dillmann (10:37):
It affects everything that you
Stephen Kraig (10:39):
Timing circuits. If you are trying to couple an
AC signal from one part of yourcircuit to another part of the
circuit, this can adddistortion. There's so many
different things that you know,this one if you weren't aware of
this and your circuit startedacting strange, it could take
you forever to figure out whatthis is. So Mhmm. Let's let's
(11:00):
let's sort of define it.
So voltage coefficient ofcapacitance. Basically, if you
apply a DC bias to a capacitor,its capacitance value actually
changes. And changes as indecreases, not always. There are
some situations where itactually increases, but for the
most part, it is a decrease, andit can be a significant
(11:23):
decrease. We're not talkingabout, you know, 1, 2%.
In some cases, it can be 85%,90% across the voltage range of
a capacitor. So that lovelylittle 1 microfarad you got and
you that you think is 1microfarad in your circuit, if
you apply, you know, aparticular voltage across it, it
(11:43):
could be acting like somethingsignificantly less if you're not
aware of this. It has differenteffects at DC and AC as well.
And so even if you havecombination of DC and AC, it has
different effects. For the mostpart, though, any data you find
will show, what happens across,with a DC bias on there.
(12:05):
Something that's alsointeresting to note is when you
look up a data sheet for acapacitor, the you will be
provided with a capacitancevalue. You should also be
provided that datasheet shouldinclude the test conditions at
which they determined thatcapacitance. So because if you
(12:26):
are trying to replicate that orif you are trying to test a
capacitor yourself, if you testat different parameters, like a
like testing at a differentfrequency or testing with a
different DC bias with whateverLCR meter you're using, you very
likely will get a differentvalue. So that's important to
keep in mind if you ever do needto test capacitors. If you're
(12:48):
comparing what you're monitoringor measuring, you or if if
you're looking at it to validatesomething on a datasheet, make
sure you're using the exact sametest conditions exactly because
of this voltage coefficient ofcapacitance.
Parker Dillmann (13:00):
Yeah. My experience with those testers is
they use a low voltage, and thatmight be because or maybe not
because of this, but it helpsthem more accurately represent
the capacitance that because ofthe the that that it's almost
(13:20):
like a negative coefficient,basically.
Stephen Kraig (13:23):
Yeah. Effectively, it is. If you zoom
in on the the charts at DCvalues that are really close to
0, it can actually increase incapacitance. And it's important
to note that, you know, we weretalking about class 1 and class
2 capacitors earlier. Thisvoltage coefficient of
capacitance only applies toclass 2 capacitors.
So your x seven r's, your x fiver's, all of those type of
(13:47):
capacitors. It only applies tothose. It does not apply to
class one caps. And here's thething, it applies to them
because this is a materialprocess. This or not a material.
This is due to the material thatthat they're made of. So, I
mentioned earlier that class 2capacitors are made of barium
tight, titanate. In my research,I found some really cool stuff
(14:09):
about this. So, when capacitorsare made, I think James Lewis
even went through this. In fact,yeah.
He he did back in episode 141,because he was asking us
questions about how long does ittake to make a capacitor, which
if I remember right, it was 2weeks.
Parker Dillmann (14:26):
Something like that.
Stephen Kraig (14:27):
I think it was 14 days, something like that.
Regardless, there's a kilmingoperation where they have to
heat and cool the capacitors,where they're basically drying
out the ceramic material andsolidifying it.
Parker Dillmann (14:40):
Yeah. They're making capacitor brownies.
Stephen Kraig (14:42):
Exactly. So when the capacitors are hot, the
crystal and structure of thebarium titanate forms what's
called a cubic crystal. Sobasically, like a diamond shape.
And barium titanate has barium,oxygen, and titanium in it.
Above the curie temperature ofthe material the the titanium
(15:05):
atom actually sits at the deadcenter of that diamond lattice
shape.
And because of that actually,let me let me let me move on to
the what happens when it whenthe temperature goes below the
curie temperature. The crystaland structure actually morphs
slightly. So as the capacitorcools, the crystal forms a
(15:26):
different shape which isactually called a tetragonal
crystal. So, think of takingthat diamond shape and
stretching it out a little bit.And the interesting thing about
it is during that process whenit's going from the cubic
crystal to a tetragonal crystal,that titanium atom that's at the
center of the diamond displacesand moves ever so slightly.
(15:49):
And that actually causes thecrystal lattice in the capacitor
to become unbalanced, and itforms an electric field within
the the crystal and structure.Now as it's cooling and
annealing effectively, you get abunch of grain boundaries and
and areas with that all canceleach other out. So, yes, you
(16:12):
have individual e fields, butthey all cancel each other out.
But perhaps you can see whereI'm going with this. That all
occurs or happens when thecapacitor is effectively at 0
volts.
But now you have all of thesedipoles created inside of this
crystalline lattice. So once youstart applying an external e
(16:35):
field, which it's not externalbecause we're talking about the
actual you're applying a voltageto the plates, you're adding an
e field inside the crystal andlattice. Now the individual
dipoles of every crystal withinthere start fighting against the
DC bias that you're applying toit, and that actually reduces
(16:56):
the capacitance, because nowyour electric field is being
modified by every single dipolewithin there. So it's
fascinating that just the slightmovement of a titanium atom as
the crystal cools is what is thecause of this voltage
coefficient of capacitance.Because you're literally setting
(17:18):
up a whole bunch of tiny efields that resist when you're
applying an e field.
Parker Dillmann (17:25):
Didn't a couple podcast episodes ago, we talked
about barium titanite.
Stephen Kraig (17:32):
I I would not be surprised if that's used in the
production of many othercomponents.
Parker Dillmann (17:38):
Well, it was with capacitors. It was when was
this?
Stephen Kraig (17:44):
Oh, was this that
Parker Dillmann (17:45):
It's like the new capacitors That
Stephen Kraig (17:47):
was that that battery. Right?
Parker Dillmann (17:49):
Yeah. It's it's a new capacitor to help offload
charging cycles off batteries.Mhmm. Yeah. And that was they
were using barium basically,it's the same stuff, but I think
they fixed they were able to,like, fix something with the
structure.
I wonder if that's what it was.
Stephen Kraig (18:09):
Okay. So I have done a bit of research on it
seems like most people or mostmanufacturers of MLCCs use
barium titanate as the the thefundamental ceramic. But
everyone has their own 11 herbsand spices that they that they
dope it with basically in orderto adjust the parameters based
(18:31):
off of your capacitor, whateveryou're looking for. And so maybe
there's a slightly differentformula that they used for
whatever that charging capacitorwas.
Parker Dillmann (18:43):
Yeah. They're saying that it's they're
basically modifying the 3 dstructure of it, and then
they're sandwiching it between 2others materials, which might be
your the 11 herbs and spices
Stephen Kraig (18:56):
Exactly. And And maybe those extra little dopants
reduce this effect or change itin whatever way.
Parker Dillmann (19:04):
But that now it makes more sense when we're
about that article. That wasepisode 431, so it was Oh,
Stephen Kraig (19:11):
that was just a few weeks ago. Yeah.
Parker Dillmann (19:12):
It was a month ago at this point. That's why I
was, like, I remember titanatebecause I couldn't pronounce it.
Yeah. And so, yeah, that that'sso I was wondering why because I
saw a bit of titanate. I'm like,oh, that was I know this is in
capacitors.
We talked about this a monthago. And so they must have those
researchers must have been ableto fix that problem, or they're
(19:36):
using it in a different way.
Stephen Kraig (19:39):
Here's the here's the thing. Because because
capacitors are so inexpensiveand they're so prolific, they're
everywhere, that just makes methink that barium titanate is
probably not that expensive.Because for the most part with
manufacturing, everyone's goingto standardize on the cheap
thing that just works. Right?Mhmm.
(20:00):
So so I wouldn't be surprised ifthere are methods to fix or or
correct or or maybe modify thisvoltage coefficient of
capacitance, but it may beexpensive. The additive that you
add in there may, you know, blowthe blow the lid off of how
expensive capacitors are. Idon't know. We'll we'll see.
(20:22):
Because because if that's thecase, then start throwing this
into everything and now yourclass 2 capacitors all become
class ones and Mhmm.
I mean, not necessarily.Regardless. So so yeah. To kinda
get back to where, you know,we're we're talking about
selecting capacitors and nowwe've gone off under this
voltage coefficient ofcapacitance thing. The whole
(20:43):
point of bringing this up isthat this is something that
doesn't necessarily isn'tnecessarily in your head when
you go to Digi Key and clickpoint 1 capacitor 50 volt 0 805.
Right? And you're like, oh,cool. I I get it.
Parker Dillmann (20:57):
Sort by cheapest. Right.
Stephen Kraig (21:00):
If it if it's COVID, sort by most available.
Right?
Parker Dillmann (21:03):
Yeah. Yeah.
Stephen Kraig (21:06):
2018 was dark times. It was rough. It was real
rough.
Parker Dillmann (21:10):
That was the capacitor shortage era. Yeah.
That COVID was the activecomponents shortage. Yeah.
Stephen Kraig (21:18):
So I wanna bring up I wanna bring up something
from Murata that is pretty cool.If somebody hasn't seen this
before, it's called sim surfing,s I m surfing. Murata has a
database of all of theircomponents or all of I guess,
all of their components. I'veI've mostly only looked at
capacitors. But in thisdatabase, it provides you with
(21:42):
much more information than whatyou get out of just the data
sheets.
So on sim surfing, you canactually get the curves that
show this voltage coefficient ofcapacitance. And so I I had some
fun earlier today. I went andgrabbed 2 Murata capacitors. 2
with all exactly the sameparameters except one was an
(22:06):
0805 size and one was an 0402size. And so if you're just an
engineer going out to pick a capand just you you you select all
the parameters the same, What'swhat's the difference between an
0805 and an 0402 other than thefact that one's bigger than the
other.
Right?
Parker Dillmann (22:24):
Yeah. One's twice as big.
Stephen Kraig (22:27):
So, you know, obviously, this is a podcast.
Our listeners won't be able tosee this, but I actually in in
our show notes, I posted 2graphs that show the voltage
coefficient of capacitance forboth this 0805 and this 0402.
And across 0 to 50 volts, the0805 has a reduction in in
(22:48):
capacitance of about 40%. And incomparison, across a voltage of
0 to 50 volts on the 0402, ithas a change of capacitance
close to 90%. So all parameters,exactly the same, dielectric.
They're even made of the samematerial, just different size
(23:10):
capacitor has a huge impact onthe voltage coefficient of
capacitance. And that's notsomething like, I keep saying
this. You're not gonna find thatin the datasheet. You just have
to know it, it, or you have tolisten to circuit break and
you'll you'll learn.
Parker Dillmann (23:22):
Right? Yeah. You have to be a breaker.
Stephen Kraig (23:24):
You have to be a breaker. That's so so and a lot
of this is what's got methinking more recently about
talking with my team at workabout how do you pick a
capacitor? How do you pick howdo you know when to say 0 805
versus 0402 versus 0603? Becauseright now, I could go to Digi
Key, and I could find capacitorswith the exact same parameters
(23:47):
in 10 different sizes. Which onedo I pick?
You know, I've been designingboards for over a decade now. I
like 0 4 0 twos. They tend towork with the ICs that I play
with. They're small enough thatthey fit the pitches of a lot of
the ICs I I go with. But sizeisn't just everything.
(24:07):
Right? And this is one reallygreat example of that where,
yeah, you could be picking apass a capacitor and you could
be completely screwing yourselfover by doing that.
Parker Dillmann (24:16):
Yeah. It's I don't know if you're gonna get
to this, but, well, I Itypically do too because I use,
you know, 0 42s for bypasscapacitors a lot.
Stephen Kraig (24:26):
Mhmm.
Parker Dillmann (24:26):
0.1 microfarad bypass capacitors. What I'll do
is I'll pick a much highervoltage so my, like, my rail
will be 3.3 volts and I won'tpick a 6 volt capacitor. I'll
pick, like, a 16 volts because Iknow that that curve, I'm gonna
be in the better part of thecoefficient curve, so to speak.
Stephen Kraig (24:47):
That's that's that's a good way of handling
it. Yeah.
Parker Dillmann (24:49):
And I I actually I learned that on the
episode with James Lewis,episode of 41, 141. So that
sounds like,
Stephen Kraig (24:57):
oh, you you basically prederate your design.
Parker Dillmann (25:01):
Yes. You have to prederate just because of the
voltage ratings on thesecapacitors or if you can swing
the price and like, 0.1microfarads is a very small
value, and so you can probablyget a class 1 that's in an 042
that fits that.
Stephen Kraig (25:19):
Yeah. Maybe. Maybe. Class class ones, they
they I bet you could. It's beena while since I've looked for a
class 1.1 microfarad, but that'sstarting to get to the point the
upper limit of what class 1 capscan handle.
So this voltage coefficient ofcapacitance, it is obviously
(25:40):
material dependent. Right? It'sbarium titanate, but it is also
dependent on your capacitiveplate spacing. So let's take a
second and just think aboutwhat's inside of a capacitor.
Yeah.
I mean, it is a it's a brick ofceramic, right, with a bunch of
(26:00):
plates inside. In fact, if yousee, like, a cross section of a
capacitor, it's surprisinglysimilar to the drawings you had
in your textbooks. Right? It'sliterally just plates next to
each other with ceramicdielectric in between them.
Nothing particularly special.
So let's think about the knobsthat you can turn to adjust
(26:22):
parameters on a capacitor. Let'sjust play with an 0402 in our
mind real quick. Let's say I'm acapacitor manufacturer and I've
got an 0402 that I wanna designand it's a, let's say, it's a
point 1 microfarad. And so Idesign it up and blah blah blah.
And then marketing comes downand was like, oh, no.
That was supposed to be a 1microfarad capacitor. What are
(26:45):
the knobs that I have to turn inorder to make my design go from
1 to sorry, 0.1 to 1 farad, Ican add more plates into the
capacitor. Right? Or I can getthe plates closer to each other.
Parker Dillmann (27:01):
I'm gonna derail you a little bit. Yeah.
But you are right. So I'm why amI why am I on Digi Key what's
dotbe Belgium? I don't know.
Why am I on Digi Key's Belgium?Yeah. It is Belgium. Why am I on
Digi Key Anyways, I'm on theBelgium Digi website.
Stephen Kraig (27:19):
Yeah. Finding a class 1.1.
Parker Dillmann (27:21):
A class 1.1 microfarad, the smallest size is
apparently o seven 4, which I'venever heard of that, but 1206.
Stephen Kraig (27:30):
Yeah. Right. So if you want What's o
Parker Dillmann (27:33):
seven zero four?
Stephen Kraig (27:35):
Isn't that like a military designation? I think it
might be.
Parker Dillmann (27:39):
I'm looking. I've never heard of it before.
Stephen Kraig (27:42):
Yeah.
Parker Dillmann (27:44):
Like, keep going. Sorry.
Stephen Kraig (27:45):
But but yeah. So that's a good point. A quick
tangent real quick. Class 1capacitors with NPO, COG,
dielectric, or tempco is youyou're limited to the high
picofarad range. You're you'renot gonna get much past the
nanofarads range withoutgetting, you know, a 1206.
(28:07):
And then at that point, youknow, if if your circuit can
handle it, great.
Parker Dillmann (28:11):
But They're expensive too.
Stephen Kraig (28:13):
They're they're expensive. Class 1 is usually
your low value stuff.
Parker Dillmann (28:18):
Yeah. When I'm building a board or designing a
circuit, I'll use a class onefor, like, my loading capacitors
on my oscillators, like, mycrystal oscillators.
Stephen Kraig (28:30):
Oh, yeah.
Parker Dillmann (28:30):
Because those are also they're, like, 0.0047
picofarad or something likethat. I can't remember on top of
my head.
Stephen Kraig (28:36):
Well and and what's funny is at a certain
point, it's actually hard to notget a class 1 capacitor. Yes.
Like, at a certain point, theyjust all are class 1. So Yeah. I
shouldn't say all, but
Parker Dillmann (28:48):
it's 27 picofarads or something like
that. Right. Depends on thecapacitor in the circuit, of
course. But
Stephen Kraig (28:53):
Yeah. Okay. So so so back to the back to that
little game we're playing. If Iif I'm making an 0402, if I want
to up the capacitance, so I sayI'm the designer of the actual
capacitor, I can add more platesto a capacitor, or I can get the
plates closer to each other. Ireally can't change the size of
the plates too much because I'mfixed to my 0402 size.
(29:17):
Right? If you think about it,those actions that I'm talking
about, adding more plates orgetting them closer together end
up being the same or having thesame result. And if my voltage
coefficient of capacitance ishighly dependent upon your
capacitor spacing, then just byupping that capacitance via
(29:39):
those options that I've given,I've actually made my voltage
coefficient of capacitance evenworse. So something to keep in
mind when selecting a capacitor,if you have a given package
size, say, 0402, if you'relooking at capacitances on the
high end so, basically, takeyour little you the slider on
(30:00):
Digi Key or the filter slider.Look at the far extreme side
where it has the highest values.
Those are gonna have your worstvoltage coefficient of
capacitance. And then you haveto weigh, do I pick an 0402 or
do I pick whatever size that hasa terrible VCC or do I just bump
up in size? Because if you thinkabout it, like, the next knob,
(30:23):
in terms of me being amanufacturer of the next knob I
would have would be making myplates bigger. If I make my
plates bigger on my capacitor,then I don't have to get them
closer together. I can get abetter VCC by just upping the
size of my capacitor.
So just like those charts Ishowed where I have an 0402 and
(30:45):
0805, it's proof of this showingthat the 0805 has a much lower
VCC than the 0402. Because on inthe 0402 to get the same
capacitance density, they haveto shove more plates closer
together, which leads to a worseVCC.
Parker Dillmann (31:02):
I gonna bet you that's because with the plates
being closer together and moreof those plates probably, you
have more the effect of thattitanium atom shifting around.
Stephen Kraig (31:14):
I think so,
Parker Dillmann (31:15):
like, those interactions between the charge
and those atoms are closertogether.
Stephen Kraig (31:21):
You also have a higher density of the e field in
a more localized area. So theeffect that each individual e
field from the crystallinelattice gets compounded. Mhmm.
So if you want a lower VCC, picka larger capacitor for the a
larger sized capacitor for thesame capacitance value. What's
(31:44):
interesting is kind of rule ofthumb here actually applies to
almost all the parameters on acapacitor.
If you think about it, gettingthe plates closer together will
actually reduce your voltagerating on the capacitor because
now you have less of an of ajump for things to arc across or
for, the dielectric to breakdown. So capacitors with a
(32:08):
higher capacitance value in asmaller package will inherently
have a lower maximum voltagerange unless those 11 herbs and
spices help out. There's alwaysexceptions to all of these rules
that, manufacturers have come upwith. But in we're talking
about, like, generic capacitorshere.
Parker Dillmann (32:29):
Yeah. I'm looking at this this chart
between this 0 805. So the worstcase for that 0805 is negative
40%. If you basically appliedlike let's say negative 40% was
like the worst you could sufferthat actually makes that 042
capacitor actually only reallybe a 16 volt capacitor instead
(32:52):
of 50 volt.
Stephen Kraig (32:54):
Right. Right. Yeah. And in fact, the 0 805 at
at 10 volt DC applied is almoststill the same value. Yeah.
Maybe there is some kind of ruleof thumb where it's like if you
apply 1 fifth of the voltage,then you're within some reason.
I don't know. I don't I'm notsure about that, but it is kind
of interesting that that simsurfing website that's the the
(33:18):
it's made by Murata. It's it'ssort you only find Murata parts
on it, but you can get all ofthis data and you can see all of
this data and you can get,printouts of every single
capacitor you have. So if youreally want to fine tune your
circuit and know every littlething about it, they have all of
their charts on there.
Parker Dillmann (33:36):
Yeah. My rule of thumb is I just I I guess
I've never really used ceramicsand high voltage high voltage.
I'm doing quotes, like, over 24volts. I always pick, like, a 16
volt capacitor and whatever sizeis working on my board for,
like, a 3.3 or 5 volt rail. Thatway I know I'm in the better
part of these curves.
Stephen Kraig (33:57):
Yeah. Yeah. You're in the the closer to 0
part. Yeah. Yeah.
So
Parker Dillmann (34:04):
50 volts seems like a lot for, like, a point 1
microfarad part.
Stephen Kraig (34:08):
It is. It is. But I just picked
Parker Dillmann (34:11):
a voltage fan.
Stephen Kraig (34:13):
Yeah. But but yeah. I mean, you could run this
exercise with basicallyanything.
Parker Dillmann (34:17):
Mhmm.
Stephen Kraig (34:17):
Now this this brings up another point that we
hammered home in episode 141,but I wanna I wanna talk about
it again. And it was do not pickthe extremes or avoid the
extremes. So, say, you you'regoing to find an 0603 extremes.
So, say, you you're going tofind an 0603 cap, and you have
voltages from 6.3 volts all theway up to 500. Don't pick 6.3
(34:40):
and don't pick 500.
What I mean by that is you havethe extremes that the
manufacturers say they canmanufacture all the way across.
If you pick something that's onthe edge of them, first of all,
it becomes more difficult tomanufacture those parts
typically at the edges of theparameters. And if that part
(35:01):
ever goes obsolete, the chancethat you would find a substitute
for that becomes that much moredifficult. That actually kind of
plays into what I was sayingearlier, like, the decisions you
make flow downstream to everyoneelse. So your purchasing
department or your contractmanufacturer is gonna get kinda
grumpy if you pick an 0201 1000volt capacitor and they have to
(35:24):
match those parameters.
I don't even think that exists,but you get where I'm going with
that.
Parker Dillmann (35:29):
Right? Yeah. I I wonder the these those edge
case capacitors. I would love tosee the VCC curve on, like, you
said the 6.3 volt capacitorbecause it might just be, like,
flat. Right?
It could be really
Stephen Kraig (35:43):
good. Yeah.
Parker Dillmann (35:44):
Yeah. It could just be they bend up that part
can be, you know, really good,and we just don't know about it
because we have to go in a simsurfer
Stephen Kraig (35:52):
and check it out. I'm actually on it right now.
Parker Dillmann (35:55):
But I actually only pick parts that I can get
that curve from. Oh. Well, Ijust like I reward the
manufacturer that's puttingforth the effort
Stephen Kraig (36:07):
I Like that that's a good way of putting it
So
Parker Dillmann (36:09):
because I think it's a very important value that
you know, this is also somethingthat should be taught in school
too because this was not taughtto me in school at all.
Stephen Kraig (36:17):
Well, no. I mean, the thing that was taught to me
in school was capacitance value.That's it.
Parker Dillmann (36:23):
Yeah. That's it. And then we had voltage as
well, but that was it.
Stephen Kraig (36:27):
Yeah. I don't even remember if I had voltage.
Okay. I'm looking at okay. Sothis is somewhat of a ridiculous
capacitor.
I'm looking at a 6.3 volt. It'sa 1206, so it's a pretty big
one, but the the cap value is220 microfarads. So it is a
gigantic cap. And across 6.3volts, it drops 80% capacitance.
(36:51):
Oh.
Yeah. But
Parker Dillmann (36:53):
So you could use that
Stephen Kraig (36:55):
at, like, 1 volt? Yeah. Or just be okay with the
fact that you're gonna have aserious
Parker Dillmann (37:02):
But I'll put this way is, you know, you're
getting what? You're at a 220microfarads, you're getting 44
usable microfarads out of thatcapacitor. So if you've
calculated your feedback loop onyour op amp, right, to and
(37:23):
you're like oh, yeah, I need 220microfarads, and oh, let me go
pick this capacitor. It'sactually gonna be 44
microfarads. And so you'rewhatever you're doing with your
feedback loop is gonna be wayout of whack.
Stephen Kraig (37:36):
So I I found another cap. This is an 0 6031
microfarad, and it is a 6.3volt. And across 6 volts, it
only drops slightly less than20%. So it it has so much to do
with capacitance density. Youknow?
If you're shoving to get 220microfarads into a 1206, that
(38:00):
thing is jam packed on theinside, and I guarantee you
those plates are really closetogether.
Parker Dillmann (38:05):
Yeah. I I'd love to everyone out there
that's listening, all youbreakers. What would you a
capacitor like that for? A 6.3volts 220 microfarad capacitor
no, ceramic that loses near itstop end 80% of its capacitance.
(38:25):
I'd love to know what theapplication is for that part.
Stephen Kraig (38:28):
You know what what what comes to mind is maybe
some kind of, like, voltagereference filtering if that
voltage reference was, like, 1.2volts or something like that.
Parker Dillmann (38:38):
Yeah. That could be it.
Stephen Kraig (38:39):
I could see something like that.
Parker Dillmann (38:41):
Yeah. A very low voltage filtering. Yeah. But
let's I'll let them know ifsomeone out there has used these
parts. I'd love to know what theactual application was.
Stephen Kraig (38:51):
And they're like, oh, that's why my circuit wasn't
performing like I thought.
Parker Dillmann (38:54):
I could be.
Stephen Kraig (38:56):
So so yeah. So stay away from the extremes.
Don't pick the lowest voltageand don't pick the highest
voltage of capacitors with whenit comes to values, do the same.
Don't pick the highest valuecapacitor in the smallest
package you can. There's a lotof reasons behind that.
I mean, we just went through allthe VCC problems, but there's
(39:19):
also a lot of difficulty inmanufacturing with those. And
what's interesting, this isthere's difficulty in
qualification of those parts. Soa lot of parts have to go
through life, qualificationtesting, a lot of temperature
testing, and a lot of extrastuff in order to meet whatever
extra ratings they have, likeautomotive and things like that.
(39:42):
And if what's funny is if you goto Digi Key and you just select
aecdashq200. So the this thispassives automotive rating, it
automatically eliminates a wholebunch of capacitors that would
have been available.
In fact, it takes about 2 thirdsof your capacitors out of Digi
(40:04):
Key, and that's not necessarilysaying that those parts are are
bad in any way. But what it doesis it kinda gets rid of the
extreme cases that would notpass the qualifying test for
being automotive. So it sort ofgets rid of the extremes and the
edge cases for you by justselecting that. But James Lewis
(40:26):
had a really interesting pointwhen we talked to him last about
this, about how capacitors startout without a capacitance value,
and then they kiln them and theykind of level out to a
capacitance value after they'vedried out in the in the ovens.
And that's a fully chemicalprocess.
(40:46):
It's not like something that youcan just turn the knob and,
like, move a thing in there.It's done. Right? It's a
chemical process, and so they'retrying to hit hit a target and
if you're trying to hit a targetthat's on the edges, that has
that's the hardest for themanufacturer to actually hit. So
it makes way more sense to tryto pick a value for both for not
(41:07):
both, for capacitance, for yourvoltage, and for your tempco
that fits more within the centerof the bell curve for any
capacitance size.
So in other words, like, a in a402, if you're trying to pick a
10 microfarad or 402, that is byfar an edge case. So that's not
going to be something very easyfor the manufacturer to create.
(41:31):
And so hitting all of theirtargets with that, all the
parameters with it is verydifficult. That would be a
really good indicator to notselect that size of capacitor.
Now if we're talking about a 100picofarads in an 0402, now we're
talking.
Now that's something that'sright in the center of what
makes sense for that size ofcapacitor. Now, at the same
(41:51):
time, if you have, say, find it,like, a 1210, a big cap, and
it's a 1 picofarad cap in there,you have the exact opposite
thing even though the resultsare the same. Like, how do you
make 1 picofarad in a bigcapacitor? You're gonna have 2
plates inside. I'm just I Idon't know exactly the amount,
(42:13):
but you're gonna have, you know,a lot fewer plates and they're
gonna be spaced differently.
It's that something like that isalso hard for the manufacturer
to hit on the low side ofthings. So that would be a great
example of just like the 10microfiber 4 2, the opposite
picking a really low valuecapacitor in a big package is
(42:34):
hard for them to hit. So it allcomes down to get a gut feel for
what the bell curve of everypart looks like in terms of
their spread and try to sticknear the center of that. And
don't be afraid of bumping up tothe next value or the next size.
Yes.
(42:54):
It is nice to have smallcomponents. They fit into your
layout and, you know sure. Iagree. I've I've been there. But
in terms of being able to getparts that consistently work,
and if they go obsolete, you caneasily find substitutes for them
and you don't get bit in thebutt by problems with VCC, try
(43:16):
to stick to the center of thebell curve.
That's sort of the big maintakeaway here.
Parker Dillmann (43:22):
Larger components also enable you to
route between pads.
Stephen Kraig (43:27):
That's true. That's true.
Parker Dillmann (43:29):
I think we we used to call those the sneaky
traces on this podcast a longtime ago.
Stephen Kraig (43:34):
I remember that. I remember being asked, are you
doing sneaky tracing?
Parker Dillmann (43:38):
Yeah. Because at o four zero two level, you
it's I don't think you can fit atrace in there. I need to try,
actually. I wonder if
Stephen Kraig (43:46):
there is With standard manufacturing, yeah, I
don't think
Parker Dillmann (43:49):
you can. With 3 mil, 3 mil, I wonder if you can
sneak 1 in. Maybe not. That's, 9mils total.
Stephen Kraig (43:56):
I don't think you can. Unless you compromise on
your footprint a lot.
Parker Dillmann (44:00):
Yeah. And then then you get more tombstoning,
which is not really acompromise.
Stephen Kraig (44:06):
No. I you lose way worse.
Parker Dillmann (44:09):
Yeah. That's way worse. Yeah. I do like
designing with one size in mind,but capacitors are always a
different game. Usually, I'lltry to pick I do usually do,
like, 042.
Sometimes I'll bump up the sizebecause of the like, let's say a
point 1 microfarad or whateverbecause it makes sense for,
like, the trace it's near. Like,let's say it's on my output of a
(44:31):
LDO or whatever, and so I'llhave, like, a 10 microfarad cap,
which is like a, you know, 1206size. But then next to it, you
know, I need to have, like, asmaller value, like, a 0.1
microfarad or 1 microfarad. AndI'll actually pick probably the
same size if I can just forthermal loading when it reflows.
(44:53):
So that little part doesn't justget overcooked.
Right?
Stephen Kraig (44:57):
Yeah. Yeah. Yeah. That makes sense. So that's
There's also heat shadow.
Right? Yeah. So it might not getcooked enough if it's next to a
big component.
Parker Dillmann (45:08):
Yeah. That that's weird stuff. You just
have to learn by just designinglots of boards. Yep.
Stephen Kraig (45:15):
So Yeah.
Parker Dillmann (45:17):
I got nothing else to add to this.
Stephen Kraig (45:19):
Yeah. I think I think I've vomited enough
information out.
Parker Dillmann (45:22):
Let us know. I really wanna know people who use
those edge case components,especially that, like, 6.3 volt
220 microfarad that loses 80%over over that 6 volt range.
Stephen Kraig (45:38):
Also, did you know that it's losing 80%?
Parker Dillmann (45:41):
Yeah. Yeah. Let us know. What did you use that
in the design for? So so thankyou for listening to circuit
break for macro fab.
We were your hosts Parker Dolan.
Stephen Kraig (45:53):
And Steve and Craig.
Parker Dillmann (45:54):
Later everyone.
Stephen Kraig (45:55):
Take it easy.
Parker Dillmann (45:56):
Thank you. Yes. You breaker for downloading our
podcast Tell your friends and coworkers about circuit break the
podcast from macro fab leave usa review, too If you have a cool
idea project or topic you wantus to discuss, let Steven and I
in the community of breakersknow. Our community where you
can find personal projects,discussions about the podcast,
(46:17):
and engineering topics and newsis located atform.macfabdot com.
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com