Maker Pro
Maker Pro

What is microstrip?

D

Dummy

Jan 1, 1970
0
What is microstrip? What does it do.
I bet I've never seen one in my life before.
 
J

John Larkin

Jan 1, 1970
0
What is microstrip? What does it do.
I bet I've never seen one in my life before.

Bet? How much?

A microstrip is a planar conductor on the surface of a dielectric,
usually but not always with a big conductive ground plane on the other
side, or embedded part-way in the dielectric. So every PC board
surface trace is a microstrip, pretty much.

Fast boards usually have an interior ground plane, and there are
equations that let you figure out the impedance of microstrip traces,
given the geometry and material properties involved; some of those
equations are reasonably accurate, sometimes. Agilent's free Appcad
program does this pretty well.

Try something adventurous, like googling "microstrip." Do "stripline"
and "embedded microstrip" while you're at it.

John
 
B

Bob Penoyer

Jan 1, 1970
0
What is microstrip? What does it do.
I bet I've never seen one in my life before.

Microstrip is a lot like a regular, run of the mill PC board. On a PC
board, traces are typically laid out in whatever manner gets things
connected as the designer might like. Things aren't so simple using
microstrip.

Microstrip typically uses a solid ground plane on one side of the
board material with traces only on the other side. What makes
microstrip distinctive is that the board material is very consistent
so that its dielectric constant (in a simplistic sense, its insulating
characteristic) is very predictable. On such a board, traces are
constructed with particular widths so that they have a controlled
characteristic impedance. That way, traces behave as transmission
lines, something like coax cable.

Clever things can be done with microstrip. It can be used as a
transmission line as just described. By simply controlling its
geometry, microstrip can be used to create filters, power splitters,
power combiners, impedance transformers, couplers, planar antennas,
etc. The dimensions of microstrip are dependent on the substrate's
dielectric constant and the frequency of operation. The higher the
frequency, the smaller the dimensions. For this reason, microstrip
tends to be used at frequencies roughly from 1 GHz on up.

If you've looked at an RF module with something that looked like a PC
board but with some strange looking geometric traces, you've probably
seen microstrip. To the uninitiated, micrstrip is a lot like black
magic.

Stripline is like microstrip except that stripline traces are
completely contained inside the board (substrate) with ground planes
on both outer layers.
 
C

Chris Carlen

Jan 1, 1970
0
Dummy said:
What is microstrip? What does it do.
I bet I've never seen one in my life before.


How old are you? Might you consider (if of appropriate age) a college
degree in EE? If you do that, or even if you take just the single
relevant course that explains all this, then you will understand deeply.

That course is one in electromagnetic field theory, based on calculus,
which implies that you need a certain degree of calculus competence in
hand as well before you take the course. In it you will be exposed to a
thorough derivation of the wave equations that describe EM wave
propagation on a transmission line, and you will learn through Maxwell's
equations how to calculate the per unit length conductance, shunt
conductance, inductance, and capacitance of typical transmission line
geometries such as microstrip, and thereby calculate characteristic
impedances, as well as get into all the fun of reflection diagrams,
impedance matching, etc.

Study it well, not just to get the answers but rather go exploring the
topics in depth. Ie., get yourself a TI-92 calculator, or get on a PC
with Mathematica or MatLab or something like that, so you can develop
visualizations of the concepts, and you will not only ace the course but
really get a "feeling" for what it all means. This understanding will
revolutionalize your understanding of all aspects of electronics design,
such as cabling signals, PCB design, and why wires aren't always just wires.

If possible take an EM course that emphasized transmission lines, and
treats it in the first part of the course, rather than at the end.

Good luck!


--
_______________________________________________________________________
Christopher R. Carlen
Principal Laser/Optical Technologist
Sandia National Laboratories CA USA
[email protected] -- NOTE: Remove "BOGUS" from email address to reply.
 
J

John Larkin

Jan 1, 1970
0
That course is one in electromagnetic field theory, based on calculus,
which implies that you need a certain degree of calculus competence in
hand as well before you take the course. In it you will be exposed to a
thorough derivation of the wave equations that describe EM wave
propagation on a transmission line, and you will learn through Maxwell's
equations how to calculate the per unit length conductance, shunt
conductance, inductance, and capacitance of typical transmission line
geometries such as microstrip, and thereby calculate characteristic
impedances, as well as get into all the fun of reflection diagrams,
impedance matching, etc.

Do you use Maxwell's equations to calculate microstrip impedance?

The last time I used calculus was in estimating a mosfet's switching
power dissipation, maybe 5 years ago. It came out close to a quickie
graphical estimate, so wasn't actually necessary at all.

I did take a year's worth of field theory in college, and we finally
got up to the full expression of Maxwell's equations (divergences,
gradients, curls, all that nasty vector field stuff.) I remembered it
just long enough to struggle through the final exam.

John
 
Å

ånønÿmøu§

Jan 1, 1970
0
Agilent's free Appcad
program does this pretty well.
Do you have a link for it or maybe you could post it to A.B.S.E. ?

I tried a search for it and all I could find was PDF's with reference to it.

thanks!
 
C

Chris Carlen

Jan 1, 1970
0
John said:
Do you use Maxwell's equations to calculate microstrip impedance?

Maxwell's equations are needed to derive the formaulas to determine the
incremental (per unit length) R', G', C', and L' for a particular
geometry. Once those formulas are in hand, then Maxwell's equations
aren't needed anymore, unless you encounter a new geometry (which
wouldn't be called microstrip of course, which is just one particular
geometry.)

The values of the incremental parameters plug into the characteristic
impedance formula for a TEM transmission line to give the impedance.
That formula can be derived from circuit theory applied to an
incremental length of transmission line, which leads to the pair of PDEs
that are the wave equations. Solution of the wave equations leads to
the characteristic impedance relation.

In practice, one needs only the formulas. But I consider all formulas
suspect unless I work through the derivations myself. Also, if you are
the sort of person that gets deep satisfaction out of the way
mathematics so precisely and eloquently desribes physical phenomena,
then the experience of going through the math is very fulfilling.
The last time I used calculus was in estimating a mosfet's switching
power dissipation, maybe 5 years ago. It came out close to a quickie
graphical estimate, so wasn't actually necessary at all.

That is often the case, for this sort of thing.
I did take a year's worth of field theory in college, and we finally
got up to the full expression of Maxwell's equations (divergences,
gradients, curls, all that nasty vector field stuff.) I remembered it
just long enough to struggle through the final exam.

John

What really impressed me was the transmission line theory and the plane
wave propagation theory. I wish I had the time and patience to continue
on with the plane wave stuff, because we stopped just before getting
into the derivations of the Fresnel equations, and other optical issues
which are of great interest to my laser work. Ultimately, I'd like to
be able to understand plane waves in anisotropic media, which would
reveal the inner workings of frequency doubling crystals and optical
parametric oscillator crystals, and other non-linear optical and
electro-optical stuff. But I don't know if one can get that far without
tensors.

I had to learn the vector calc. on the fly as I took the course, because
I didn't have a formal class in it. But it wasn't difficult.

Trouble with all this stuff is you do indeed loose fluency in the
subject quickly without practice. But the overall impression and the
fundamental understandings remain, and are far deeper than anything I
could have gotten without the math treatment.

Good day!



--
_______________________________________________________________________
Christopher R. Carlen
Principal Laser/Optical Technologist
Sandia National Laboratories CA USA
[email protected] -- NOTE: Remove "BOGUS" from email address to reply.
 
Å

ånønÿmøu§

Jan 1, 1970
0
Thanks for the link! Google ... I should of known! I went to Agilent and searched there and I left a
message for some support. I got this email back four times! (for a single inquire...)
***********************************************
Hi ånønÿmøu§,
Thanks for contacting us at Agilent Technologies.
You can contact the following number to confirm it and the number is as follows:
Product Purchase
Hours:6:00am - 6:00pm MT, Mon - Fri
Check prices or request a formal quotation, place an equipment, upgrade or retrofit order, verify
delivery availability, check order status or expedite an order.

Thanks and Regards,
(***Name removed***)
Phone: +1 (800) 829-4444
FAX: +1 (800) 829-4433
**************************************************
Oh well! Thanks again!
 
J

John Larkin

Jan 1, 1970
0
Google is amazing. When I want a datasheet, instead of going to the
mfr's web site, I usually just punch the part number into google and
hop directly to the right page. I can't understand how this much data
could be indexed as fast as it does... googling "lm2991" got the
National product folder first hit, in about 2 seconds. I suppose it
did about a million other lookups at the same time.

I think google is like ebay: it only makes sense to have one auction
site (why would you want to list on #2?) and I think the dynamics is
that one search engine ultimately dominates.

John
 
W

Walter Harley

Jan 1, 1970
0
Chris Carlen said:
That course is one in electromagnetic field theory, based on calculus,
which implies that you need a certain degree of calculus competence in
hand as well before you take the course. In it you will be exposed to a
thorough derivation of the wave equations

Oof. I took "Fields, Matter, and Waves" as an undergrad. The main thing I
remember was that it was a large classroom on the other end of campus from
the class I had just before it, so I was always stuck somewhere way in the
back of the room, panting from the across-campus sprint; and the professor
had an almost incomprehensibly thick German accent, and a habit of drawing
vector diagrams with colored chalk so that he could get a massive amount of
information into each drawing, information which I could barely read from my
distance and which I anyway had no good way of capturing in my notes unless
I remembered to bring colored pens. So each class was just a blur for me.
I got a "C", if I remember, which was uncharacteristically poor; and I've
always wished since then that I understood the topic better. I keep
thinking of trying to take it again, but of course now the calculus has
faded too.
 
C

Chris Carlen

Jan 1, 1970
0
Walter said:
Oof. I took "Fields, Matter, and Waves" as an undergrad. The main thing I
remember was that it was a large classroom on the other end of campus from
the class I had just before it, so I was always stuck somewhere way in the
back of the room, panting from the across-campus sprint; and the professor
had an almost incomprehensibly thick German accent, and a habit of drawing
vector diagrams with colored chalk so that he could get a massive amount of
information into each drawing, information which I could barely read from my
distance and which I anyway had no good way of capturing in my notes unless
I remembered to bring colored pens. So each class was just a blur for me.

It is at this point that one would be wise to begin learning how to
learn on one's own. There have been a few lecturers in my academic
experiences who really had something to offer that went beyond the text.
But most of the time I could ignore the lectures, and just study the
text. Ultimately, at the graduate level, this is an essential skill.

I am fortunate to work with folks who can pick up advanced texts outside
their fields, and ina few weeks start functioning at a respectable level
in the new material. Such as a bunch of mechanical engineers doing some
pretty sophisticated spectroscopy. They are very impressive.


Good day!
 
K

Keith R. Williams

Jan 1, 1970
0
It is at this point that one would be wise to begin learning how to
learn on one's own. There have been a few lecturers in my academic
experiences who really had something to offer that went beyond the text.
But most of the time I could ignore the lectures, and just study the
text. Ultimately, at the graduate level, this is an essential skill.

Wow! I find just the opposite. I pick up far more from a
lecture than by reading. In fact I do *very* well listening to a
good lecturer, enough that I can aid the instructor in explaining
the material. I get virtually nothing from a text alone. A good
lecture will get me over the hump such that the text makes sense.

Perhaps we're on opposite ends of the learning stick. ;-)
I am fortunate to work with folks who can pick up advanced texts outside
their fields, and ina few weeks start functioning at a respectable level
in the new material. Such as a bunch of mechanical engineers doing some
pretty sophisticated spectroscopy. They are very impressive.

Certainly. Thet would impress me! ..and I've known several like
this. Sadly, I'm not one. However, a simple lecture does
wonders.
Good day!

Indeed it was! ;-)
 
G

gwhite

Jan 1, 1970
0
So every PC board
surface trace is a microstrip, pretty much.


I find I do grounded co-planar wave guide (GCWG) more
often these days, although as you know they are "close."
GCWG seems convenient because using a standard ground fill
gap creates it automatically. Microstrip requires special
keep-outs or design rules if the top is ground filled.
 
Top