Heavy Copper Webinar Sept 2019

Heavy Copper Webinar Sept 2019


Lisa: Welcome everyone I’m gonna go
ahead and get started today thank you for joining today’s webinar my
name is Lisa Holmes and I’m the marketing director at advanced assembly
today we are discussing designing with heavy copper this is actually a
presentation rag delivered at PCB West a couple weeks ago but before we begin I
wanted to go over a few logistics you should have a control panel on your
screen if you have any questions during the presentation please use the chat box
to ask them we will try to answer the questions during the webinar but we’ll
also have time for questions at the end now I’d like to go ahead and introduce
our presenter Greg Ziraldo. Greg is the director of operations at
Advanced Assembly with over 14 years of experience in PCB manufacturing. Greg is
responsible for all areas of engineering and manufacturing at the company. Under
his guidance advanced assembly maintains a ninety eight point eight percent
on-time performance record with orders shipping in five days or less on average
prior to an advanced assembly Gregg was manager of manufacturing at TTM
technologies. Now I’m gonna go ahead and turn this over to Gregg to get us started. Greg: Good morning everybody good afternoon to the people on the East coast. Again my
name is Gregg glad everybody can join so this presentation is going to cover
several facets of design and fabrication with heavy copper generally heavy copper
is considered at about one point five ounces per square foot I want to talk
about material properties how to determine copper thickness for your
designs how manufacturers fabricate your board and then challenges that we have
when we are actually assembling it like said at the end of the presentation be
glad to take questions so if you have anything write them down or feel free to
use the chat box and if I could hit that in the middle of the presentation I
absolutely will let’s get started so copper has a few basic material
properties that make it an excellent choice for electronics and I want to
kind of preface this discussion as this presentation is going to start really
kind of at the entry level one-on-one 101 class and then we’re gonna kind of
ramp up to a few more kind of technical key points that I want you guys to take
away from this but again we’re gonna start basic and kind of go from there so
again copper has excellent material properties that does make it a the first
choice for electronics it has high electrical conductivity as a very high
thermal conductivity and it’s also relatively inexpensive compared to other
materials materials that conduct electrically well also condemned tend to
conduct heat well copper conducts heat and electricity better than any other
common PCB material it’s perceived as better than aluminum better than steel
better than gold silver is a slightly better thermal and electrical conductor
than copper but its cost is at least 100 times greater than coppers cost per
pound that really makes copper the primary conductor of heat on most PCBs
designers will sometimes include aluminum heat sinks to help dissipate
heat into the environment via convection but that’s really a choice based on
copper excuse me based on cost but at the end of the day copper is the
superior conductor so we’re going to some material
properties copper and core material expand as the board temperature
increases the coefficient of the thermal expansion is really one to two orders of
magnitude greater in the core layers than of is and the copper layers so
whenever the board changes temperatures these two materials really expand at
different rates and materials can then experience mechanical stress you can
start inducing latent defects immediate defects obviously all these things need
to be considered when you’re designing the glass fibers constrain in the
expansion in the X and y directions of the material but the z-axis expansion of
the core materials is really where your trouble is going to set in the core of
the material can actually expand much more than the copper can this is going
to lead to a variety of failure modes and played it interconnects that will
cause your board to fail somewhere down the road as we discussed just a minute
ago with latent defects these failures are also not immediate if they happen
weeks or months at the end this is ultimately going to leave into a failure couple of different examples of
different via failures if your plating is it done correctly your board is gonna
fail when it comes to heavy copper these images show cross-sections of vias that
are both post and mid failure via failures are not always an open circuit
failure sometimes they will remain as a measurable resistance defect for a time
before actually failing completely even if your plating is done correctly high
temperature thermal cycling and harsh environments can also cause your boards
to fail as well so I’m gonna kind of go through a couple of these different
images real quick so this first image in this micro section is there significant
cracks and even cracks they’re starting to propagate on the left-hand side of
the barrel you see an entire fracture that’s completely disconnecting on the
right-hand side you start to see stress fractures going into that as well so
this can all be from heavy copper this can be from high heat this can be from
mechanical stress as I said a minute ago what’s gonna happen with this is that
you can obviously achieve an immediate open within that specific net at the
same time to the expanding and contracting of the copper as it goes
through your thermal cycling can connect and disconnect all during the same test
it’s gonna throw your four-year readings off it’s gonna throw you’re all just
basically of your data off resulting in failures this second image right here is
an entire fracture where this would be a complete disconnect from the whole wall
to the actual interconnect itself is a complete separation from the signal
itself so now we’ve seen kind of a couple of different things at what
thermal expansion can do with copper on the PCB you probably realize that it’s
best to avoid large changes and rapid change changes in temperature
what temperature change you decide is appropriate for your design as a matter
of design requirements and the material properties of the core material that you
choose some important laminate properties to consider when it comes to
protecting circuit boards especially from high temperature situations include
I’m gonna go over a couple of just key points here and some definitions that
are really going to kind of help you and your journey as you go through this so
one of the first things that we typically look at is what we call T G
which is the glass transition temperature and this is really the
thermodynamic shift of the polymer of the core from Ridgid to Saab so
basically when you are going through lay up your board is going through its
Precure if this is really just what point the material becomes soft and
excuse me from rigid to soft and then starts to actually flow the z-axis
excuse me the z-axis expansion this is the thermal expansion along the z-axis
and this is usually expressed as a percentage or CTE which is a coefficient
of thermal expansion next thing that we also look at is what we refer to as TD
or the decomposition temperature the TD is the measurement of weight loss due to
the resin degradation because of high heat so it’s that point of which 5% of
the mass is lost due to decomposition next we look at moisture absorption this
is actually how well a material can absorb
excuse me absorb moisture from the environment and we’ll talk about a
couple of different materials that actually have significant problems if
they’re not addressed quickly with that and then the lastly that we look at is
what we call time to delamination and this is just basically the time dilemma
the ulema nation takes given out certain temperature slides are moving
sorry about them so standard materials that are pretty common in the PCB world
like 85 n fr 4 3 7 th are they all do well with heat however they’re really
heavy so if you guys are designing for something that has a payload requirement
sometimes these need to be reviewed for consideration if the stack up is too
large 85 MT is a laminate that actually uses aramid fibers as a base
construction rather than woven glass so this is going to support weight
reduction and is very strong however it will absorb moisture very quickly and
will require extensive pre baking and vacuuming prior to post lamination so
basically with this material it’s almost kind of like particle board if you’re
looking at like a wood reference so instead of actual grains of wood it’s
essentially fibers almost looks like cardboard that’s pressed together and
bonded with with resin and what this material does is that it’s incredibly
strong it’s very very lightweight but it absorbs moisture like a sponge so there
is a significant amount of work that needs to be done in the inner layers
stage as well as the outer layer stages and sub assemblies when you’re
manufacturing with 85 NT so like I said a minute ago vacuuming pre-baking post
baking are all things that need to be done to mitigate moisture in this
because what’s gonna happen is that you’re gonna go through and especially
if you’re manufacturing heavy copper high temperature in the HDI world you’re
gonna get to the final assembly of your build and you’re gonna have high
resistance shorting all over the place so just be careful with that more heat and electricity with your PCB
while keeping the temperature down you must increase the cross-sectional
area of your PCB trace there’s two ways to increase the cross-sectional area of
the trace and that’s to increase the width or the thickness or both
increasing the width consumes valuable circuit board real estate which leads to
fewer boards for panel which is ultimately going to increase your cost
increasing the thickness of the traces increases production time a bit but
allows you to put more PCBs on a panel theoretically the more PCBs per panel
results in a lower production cost and at the end of the day that means money
that you’ll save in theory heavy copper also makes for a stronger board thicker
copper helps to stiffen the boards but you need to ask yourself is this really
what you need so you’ve got a couple different options to simulate these
things you have two that you can really use reliably one of them is you can
simulate or you can estimate estimation is really the cheapest but it’s really
the least accurate option it’s more so with this as experience in time with the
design it’s only gonna be really useful here if you’re working with something
that is brand new a new type of technology that you guys are working
with this is not going to be a viable option simulation is going to involve
software sometimes hardware but it’s time consuming to set up and run but
it’s going to really provide the best approximation for how your board is
gonna behave in the real world most modern design and camp software
platforms can really were done I can run a conductor simulation for impedance
values either by Gerber layer or stack up typically what you can do on the if
you design a Gerber layer and there’s a certain trace thickness copper weight
that you put in you can essentially highlight that specific trace that
specific net and kind of run a simulation test on how that is going to
perform controlling impedance through dielectric dielectric and trace width
and space needs to be considered for high frequency designs as the designers
you guys can call out values in ohms with typically a plus or minus 10% which
is standard on these conductors and the testing of course is gonna be done what
we call a TDR process which is short for time and domain reflectometry if you are
designing for harsh environments like high altitude space down hole locations
where convection isn’t available to cool your board you should really simulate
your design simulation is going to provide reasonably accurate information
of how your board is going to behave in the real world tools that I prefer our
sim scale SolidWorks Autodesk mint or all of those things are going to provide
a variety of simulation methods and really what you choose depends on what
you can budget and how accurate your simulation needs to be if you depend on
convection or forced air to cool your board you might find that you need to
choose a tool that can handle more of a fluid dynamics route otherwise your your
data and your results are gonna be inconclusive they’re not going to be
accurate if your budget doesn’t allow for a fancy simulation software package
you’re gonna ultimately have to estimate calculate and and calculate your trace
thickness excuse me estimation is often good enough for simple designs if you
can give yourself a suitable margin for error by derating your design based on
expected operating conditions a couple things to keep in mind is that
if you guys are using the old equations from the IPC 2221 to calculate thickness
and width you can potentially be wasting a lot of copper money and time I’m gonna
kind of go into some details with this but I think it’s really worth noting
especially since there is a perceived gospel of information that’s out there
that most design calculators use but is not entirely accurate I’m gonna go to
some details with them so the equations from the IPC to in 20 to 21 are derived
from data that was formulated in 1956 actually by a pair of pretty much
unfunded researchers who really listed their results as tentative the idea that
an internal trace must be d rated by 50% appears to come from assumption in those
charts but not empirical evidence it kind of turns out just through our
research for this talk it was really wildly inaccurate and internal traces
can actually be berated as little as 5% in many cases those results somehow
survived decades to be incorporated into formal standards and every single online
calculator that we found when we were preparing for this talk and it just like
I said a minute ago it just really became gospel even though there was
nothing specific out there that actually confirmed or denied that these were
actually accurate IPC 21:52 attempts to fix the problem with more recently
derived data but you don’t need to purchase the standard to get the results
and learn what the curves do a couple gentleman Douglas Brooks and Johann Adam
I wrote an entire book on the subject and actually the most useful chapter of
this book is actually free on dr. Adams website and I’ll post that here in a
minute you guys take a look at that if you’re designing for a specific
temperature change that number becomes a constant as your current increases the
trace width and/or thickness must also increase to compensate increased trace
width does a slightly better job at dissipating heat than increased
thickness back to Douglas Brooks’s website is this ultra CAD URL that’s
posted at the bottom of this slide and there’s more details on that and then
issues with the original 2221 and really discuss at length how to use the new
equations and curves from the newly derived IPC 21:52 so in this graph right
here this is a couple of examples and just really simply using the simple
equation that I posted a minute ago I’m really kind of where this is going to be
at so this was simply arranged to provide clarity for a specific design
temperature at an increase of 20 degrees C so for a known current value that you
guys have this graph can really be used to determine trace width and copper
weight so for example with this one is you have a 5 amp current that can be
carried by a 430 mil wide trace at quarter ounce copper by using the same
method and equation this would go down to you could use 130 ml wide one ounce
trace and then you could use a 40 mil wide 4 ounce copper trace so remember that this assumes that this
is unobstructed convict because to be convection of service traces at a
standard temperature and atmospheric pressure these equations are going to
need to be reviewed if you guys are using space and down whole product
applications so when we talk about fabrication so printed circuit boards
are fabricated in dozens of steps you have basic circuit boards to layer four
layer that are built in you know ten to twenty steps and then you can have HDI
multi lamination microvia copper fill that are 150 to 200 steps the basic idea
for internal layers is to remove copper in places that we don’t want it to
we don’t want through etching the outer layers of copper are built up through
electroplating and then they’re etched after that so etching is used on the
internal and external layers of PCBs to remove copper so if your board stack-up
requires 2 pieces of 2 ounce copper on internal layers basically what the fab
house is doing is gonna be pulling out a sheet of 2 ounce copper clad material
and then just depending on your core substrate requirements from there the
core is imaged and then is etched away with the copper remaining to actually
create the pattern or what we call artwork and spaces of your PCB from
there you’re going to laminate those two sheets together in a press that is
controlled by time thermal rise and cure along with your outer layers and then
you’re going to build up the outer layers and via holes with electroplating
electroplating is used on the external layers of a PCB to add copper so if your
board stack it requires 2 and 3/4 ounces of copper we’re gonna pull out a thin
foil from our stack and then build up the copper on the surface and in the via
holes to reach that two and three-quarter ounce per square foot of
thickness common issues that we run into on the
fabrication side is that heavy copper is just honestly it’s just very very
difficult to fabricate whether it’s a Qing internal layers that are heavy
copper that are two three four ounce six ounce plated up or anything that is on
the outer layers that are significantly plated up at the end of the day a Qing
can be one of the most difficult things to do anytime at the end of the day
anytime something that’s really difficult fabricators gonna raise the
cost of your boards as insurance against a lower yield it’s very common for
boards to scrap do to over edging or under etching for those panels to end up
in the garbage and ultimately they need to make sure that they’re gonna be
recouping that cost and so that cost is gonna be pushed back to you heavy copper
etching is often sometimes unreliable and time-consuming boards can end up
over edged on one side like I mentioned under edged on the other side there’s
deep and narrow spaces that can trap a chin and keep fresh session from being
applied plating is also very difficult but for different reasons an electric
plated board doesn’t always plate at right the way that it’s at right angles
as far as on the board surface typically especially with today’s technology and
plating you need to go more of a post plating approach which is using
rectifiers and actual programs to be able to throw copper from the anodes
onto the panel itself rather than the old-school way of using DC plating at
least with poles plating you’re gonna be giving an even more of an even layer of
copper plating across the panel itself from edge to edge as opposed to DC
plating where you typically build up heavier copper around the the corners if
you’re not using the thieving properly then after the boards are etched and
plated there’s gonna be other problems edge boards require a special high resin
prepreg to fill the deep gaps and the voids that are created by etching
basically with this and we’ll talk about this a little bit more as we get to the
presentation is that when you have heavy copper and you’re etching if you look at
the z-axis in a micro section I kind of refer to the signal layers as they look
like train tracks and so on the internal side if you’re trying to fill those
voids with prepreg if you’re not using enough
it’s going to create voids which ultimately add a micro section level is
going to scrap your product once the boards have been laminated the
electroplated outer sides can be difficult to cover and solder mask and
then difficult to subscreen as well so just like on the inner layers where you
have deep pockets to fill the same thing goes on the outside so if you were using
a heavy copper outer layer you might need more than one coat of solder mask
to ensure that you have full copper encapsulation typically what can happen
especially if you’re using a standard LPI or even a dry film solder mask on
the outer layers and you’re only using one layer or one coat of that you’re
gonna get the deep pockets filled but the crest or a crown of the trace is
gonna be extremely thin and are actually exposing the copper which is gonna
result in either scrap or it’s gonna result in a significant amount of touch
band rework on the other side if that you’re using heavy copper on the outer
layers and you do decide to go through a double pass with solder mask you’re
ultimately increasing the thickness of your board too so if you have a very
tight restraint as far as what chassis the board is going into you could be
compromising that just have to think about assembly is the process of
actually populating PCBs with your parts this is referred to as populating
stuffing assembling it’s got a multitude of different terms for it heavy copper
PCBs especially Rojas compliant PCBs can be difficult to assemble due to the
solder temperature required so some issues that heavy copper PCBs have when
it comes to assembly is that obviously heavy copper conducts heat a little bit
too well and essentially they act like a giant heatsink they take much longer to
heat up and much longer to cool down then regular boards so this becomes a
problem especially if you’re using higher technology parts BGA’s requires a
certain amount of ball collapse the QFN packages require a certain amount of
phillyd and voiding is allowed so most people would think that you just do
simple change to a profile or slow the conveyor belts down through the reflow
oven and there’s some that’s no problem right wrong so now we’re we’re really
introducing more heat and really heating up the boards and it’s really at this
point possible to cause the epoxy or fiberglass to decompose prematurely like
we talked about and this is where Villegas can also experience some types
of cracking and a host of other problems typically with a heavy copper and large
layout board there is a significant amount of pre heat that is required
because ultimately what’s going to happen with that is that you’re gonna
have areas of cold solder if it’s not done properly
so this all needs to be done through solder samples and excuse me
solder samples and your temperature profiling into reflow oven if this is something that you guys are
gonna be working on with hi car excuse me heavy copper you should probably be
starting to have long conversations with your PCB manufacturers engineering
departments regardless of the size of your board heavy copper is gonna be a
challenge really for anybody depending on your design and small seemingly
insignificant design decisions can lead to huge cost bikes again like I said
latent defects or birds that just don’t meet your design requirements so I want
to kind of go over a couple of things I want to keep you guys want to keep in
mind so generally you want your layer stack to be symmetrical so on a four
layer board if layer one is 2 ounce and layer 2 is 1 ounce then layer 3 should
be 1 ounce at layer 4 should be 2 ounce this is really going to create a board
with greater stability if you have a board with uneven symmetry it’s gonna
have greater flexstretch on one side versus the other and that means when the
board heats up as soon as it goes through that finger I mean when the
board heats up it just gonna flex cup it could bow I kind of alluded to looking
like a Pringle chip and there’s a lot of criteria that especially for aerospace
weapons in the fence for flatness for bowing twists and if something is uneven
copper weight and especially heavy copper in that it’s it’s not gonna pass
it’s look ton nothing gonna be rejected in the final QA next I want to talk
about cross hatch or cut outs to reduce your heat transfer and this is kind of
something that’s a little bit interesting and not everybody always
things about this in uninterrupted ground plane is always the top design
goal but sometimes it’s necessary to prevent heat transfer to sensitive
components there’s a couple of tricks that we can have when we’re designing in
our back pocket when we’re working with high-end parts such as precision voltage
references so first you can use a hatched ground plane to reduce the rate
of heat transfer to sensitive areas of the board and this idea is
used in wide areas of copper elsewhere and that the heat will flow from hot
components to cool areas and hopefully be dissipated there leaving your
sensitive part alone so as you can see on this first image of this just this
gerber artwork so if you look next to the through holes on the left-hand side
instead of having a solid ground you have a crosshatch pattern which is
ultimately gonna you know still create connections still serve the purpose of
the ground but the crosshatch design is going to alleviate heat and dissipate
heat much better so secondly this is kind of the fun one and this is the one
that I feel a lot of people kind of overlook if you need to protect a really
critical part you can actually create a clearance and just make basically three
to four sides around it using just enough copper to run your power and
signal traces so what you’re seeing in the area that’s arrow does the white is
obviously the clearance and so the component that’s gonna be placed on that
footprint right there if there is a sensitive part to that component you can
clearance still maintaining total connection to the rest of the ground
artwork but you can leave that area free and clear so that that heat is not going
to affect it I want to go over a couple of questions
before we get into questions from the chat
and this is just a lot of common questions that we routinely see that
might kind of clear up a few basic things for you all before we get started
on the next part so one of the biggest things that we get asked is how do i
specify heavy copper in my stack up should I tell my fabricator that I want
to mount plate it up to three outs or do I let the board shop decide of course
the easiest answer for that is gonna be that completely depends on your board
house so we submit your fab drawing and your stack up that you want unless you
are ordering a board from legacy prints that specify precisely how the board is
fabricated most of your fab houses today you’re gonna have an array of
off-the-shelf materials that are coming to comply with the thickness that you
require however older designs that asked for two over two material that’s plated
up to four over four ultimately this is just gonna end up in additional steps
that the fab process has to assume and then ultimately is just going to cost
more so being what I meant with that is that you know I’ve seen drawings from
the 90s early 2000s that have their legacy designs typically these are DoD
related to where they want to Hobbs inner layers as a base material and they
want them plated up to four ounce well obviously now in today’s world you know
four ounce copper layers is kind of common and they can be pulled off the
shelf but typically if you need something that’s too over to that wants
to get plated up before before as a significant amount of processing that’s
involved with that not only the plating of plaiting inner layers is very tedious
and then also actually etching those as well so another question that we get to
is discussions of embedding copper or using busbars and it’s really window
would you recommend a busbar embedded copper instead of using heavy
conferences so really what we talked about with this is really embedding
anything into a PCB is gonna add a significant amount of extra steps and
therefore significant cost so I’d really avoid it at all possible I know a lot of
people like to really design them this thing in the entire world to build
it just kind of more of a proof-of-concept but ultimately that
doesn’t add any value to the budget of the project if you are designing a board
that has two or three high current traces in it and everything else is low
current low voltage high speed digital traces that need control impedance and
you don’t really have a choice sometimes embedded copper or bus fires
at that point is gonna have to be required but ultimately it just comes
down to time and cost cycle time is going to be significantly added to these
part of these projects so the one that we run a lot are what is your minimum
trace in space and so this is a pretty generic detail of what fab companies are
going to tell you it’s gonna go over obviously your copper weights that are
on the left-hand side it’s going to tell you what they’re able to achieve for an
earlier trace and spacing external trace and spacing but again it also just
depends on what type of equipment they have and just the general capabilities
of the facility we’ve seen people go down to two and two obviously there’s
nanotechnology and micro etching that you can do to ensure that extremely
small conductor widths and extremely small spacing can be achieved it’s just
at the end of the day is going to be a conversation with your fab house and
what they can do and certainly what they can’t do and obviously they need to be
honest too and say look this is something that’s out of our wheelhouse I
don’t want to give you guys you know expectations that we can and then
ultimately fail so as far as the presentation goes that is the end of
what we have we start going through some questions hi this is mark I’d like to
encourage you if you have any questions please ask Gregg now he’s got a ton of
experience and we’ve got plenty of time left to answer a few of those so Gregg
the first question we have is from an anonymous attendee what is your opinion
and/or experience with epoxy resins specifically for filling the gaps
between heavy copper traces in a separate processing step so there’s a
multitude of different prepregs that are going to be used for that as far as you
know specific call outs or Izola materials or our alum materials as
really anything off-the-shelf is going to be fine it’s really just a matter of
actually how many pieces of prepreg you’re gonna be putting in between them
and ultimately what your design goal is for what the copper weight is it’s
really again going to be a conversation with your fab house and you know telling
them what you’re trying to achieve and you know what they can do and what they
recommend all right we had a question in the chat window that I think you
answered but I’d like to add on to that if the commenter doesn’t mind when do
you know let’s say you decide you need three ounce copper when do you know know
that you need like a 1 ounce base 2 ounce plate versus say a 2 ounce base
and a 1 ounce plate so there’s really not a there’s not a scientific answer on
why it’s really just personal preference again in my experience it comes down to
the designer at the end of the day I mean me personally if it was up to me
and I wanted a 3 ounce final thickness I would just be pulling a three ounce
copper clad piece of material but again there are older designs and some people
design them where they want a base that’s plated up but again at the end of
the day if if copper thickness is what you’re trying to achieve then it’s best
to use that that clad material right off the shelf because plating up if it’s not
done correctly there’s gonna be like a demarcation line between the base
copper and what’s been plated up there’s a risk for more fracturing at that level
if it’s not been prepped and plated properly there’s just a host of other
issues that kind of come along with it and from a risk mitigation standpoint I
would just prefer to go with the original clad material at the thickness
that you need all right what other questions do we have out there from TV
land ask him now or you’ll have to get back to work all right from Howard Evans
we’ve got gray beards used to tell me that plated copper was not as thermally
conductive as clad a kneeled copper recently I’ve read this is not true so I
think the question is do you have any knowledge on that on the thermal
conductivity differences for that’s a really good question I actually I’ve
been a gray beard myself I don’t I don’t think I’ve actually ever heard one
versus the other I think it’s one of the to me I don’t know if that’s an old
wives tale or what obviously plated copper if you look at it more at the you
know microscopic level the grains are slightly different than clad material
but as far as it actually being able to have a measurable difference of how it
transfers heat and conducts heat I have yet to see anything that proves or
disproves that one is better than the other again I think you know going back
to the gray beard comment you’re going to have people who or I guess more old
school that are used to doing it one way and then newer processes equipment and
data has supported that things are not necessarily what they seem
so it’s one of those things where you got to take it with a grain of salt and
just don’t take it for face value but I have yet to see anything in the industry
whether it’s Publishing’s white papers or anything like that that show that one
is truly better than the other and to answer that – Howard from a physics
standpoint I can’t imagine that the difference would be so great that you
would be able to measure it in anything other than a controlled laboratory
environment for all practical purposes on a circuit board maybe you’re looking
at the change of maybe 0.1% but coppers got a thermal conductor
of 430 watts per meter Kelvin you know are we talking about 430 verses for 30
point-0 – it would have no practical difference because once you get to the
air interface you’re looking at like a point zero two watts per meter Kelvin
you know it’s a two to three order of magnitude difference I just can’t
imagine that it would be meaningful in any sense at all from the physics and
the material standpoint alright from Kelly Minton what size vias can be
drilled and plated with two ounce or greater copper so what what about via
size when we’re doing heavy copper the Assizes obviously we via size is going
to be critical when you’re dealing with heavy copper microvias are completely
out of the question and your standard mechanical drill size of you know seven
ten eleven mils even those are gonna get extremely dicey because at that point
you’re throwing thousands of basically mils of copper on to the internal hole
wall I really wouldn’t start I wouldn’t touch
anything that’s under 10 mils all right hopefully that answers Kelly’s question
if not please feel free to follow up what about cost Greg what if I’m
designing with heavy copper is it cheaper if I use all external heavy
copper all internal a mixture what can I do to see cost savings with this with
this technique the biggest thing with heavy copper from a cost-saving
standpoint is really making sure that your stack up you’re able to use as much
off-the-shelf material as you possibly can because you know like we have talked
about it in the the presentation just ago is that if you are actually using it
base material and then plating up or doing a significant amount of plating on
the external layers all of those are more processed it could be more drill
cycles ultimately if you have a typically an heavy copper there’s going
to be separated drill from the vias to the actual plated through holes to where
there’s different play steps there’s different imaging steps
and so as much of those as you can reduce the steps that you can reduce
it’s ultimately gonna be cost savings unfortunately cost savings and heavy
copper kind of the antithesis of each other so you’re not out of the gate you
kind of have to expect you’re gonna be paying a little bit more for it than a
standard process of you know half ounce or one ounce or standard one meal
plating in the whole and everything like that but there are certain ways like I
said with off-the-shelf materials and things that are readily available to you
before you started going the customer out customs never cheap all right what
else would you guys like to know we’ve answered all questions okay then I’ll
continue to ask some in unless the community that would like to join in
what about custom reflow profiles advanced assembly I think you guys have
some ovens that allow some targeted temperature measurement and very quick
custom reflow generation well you can almost a little bit about how that works
so an advanced assembly we actually have two different ways that we reflow or two
different capabilities of reflow so we have conventional reflow which I
referred to as basically your standard conveyorized pizza oven with the
multitude of different heating zones and cooling zones for reflow one of the
actual you know and these work great for a multitude of different industries
different types of designs because we can change that profile to do a quicker
and higher pre-soak prior to the reflow point so that way you know from the
thermal standpoint you’re not putting too much initial thermal stress on the
board when it goes through the reflow zone the other thing that we also have
available to us too is what it’s referred to as vapor phase reflow which
is not nitrogen reflow and so there’s a few manufacturers out there that use it
but we use it regularly and we love it and basically what it’s actually using
instead of heat zones it’s using actually a chemical conversion to create
heat and vacuum inside of a sealed chamber so from a thermal standpoint
with heavy copper it reduces the ability for the board to
be used as a heat sink and more just again as a carrier for the parts so once
this this chemical conversion happens in this chamber it immediately creates the
appropriate amount of heat for the board itself and then the vacuum actually
pulls the part down onto the footprint so from a voiding standpoint especially
that is rampant with high heat and heavy copper from a standard conventional
reflow we were able to see 90 95 percent void free bottom terminal and leadless
parts highly recommended all right well with that perhaps we can bring Lisa back
and yep I’m here so we will go ahead and wrap it up for now if there is anything
that you think of that you would like to ask Greg we will be sent emailing a
recording of today’s presentation along with his contact information so you can
please feel free to reach out to him with any further questions and so thank
you everyone for joining us today and we will see you next time and Kelly and Lou
I do see your questions we’ll get the answers for that to you thanks all right
sounds great thanks everyone

Be the first to comment

Leave a Reply

Your email address will not be published.


*