[SI-LIST] Re: Inductance of Spiral Inductors - why not to us a network analyser for Q measurement

Steve:

If a network analyzer is used to measure R and wL to calculate the Q(f) of
an inductor in a broad frequency range, your reasoning is right. I think an
impedance analyzer is better than a network analyzer to measure Q(f). HP
4294 can measure Q up to 110 MHz.

Another definition for Q can be used if Q at only a few frequencies is
needed. Q = fo/(f2-f1) where fo is the resonance frequency, and f1, f2 are
3dB frequencies. With this definition, Network Analyzer can be used to
measure very high Q. If a bare inductor is measured, the Q at self-resonance
frequency (SRF) can be obtained. If the Q's at other frequencies are needed,
a high Q capacitor can be used to bring the resonance frequency down.
Capacitors with much higher Q than inductors are readily available, so that
1/Q_meas = 1/Q_L + 1/Q_c ==> Q_L ~ Q_meas.  If the capacitors are tunable
(also available), you can measure Q at many frequencies. This way you can
use Network Analyzer to measure much higher Q since 1) you don't have to
measure the tiny phase angle accurately, and 2) you do have to measure
frequency accurately, which, fortunately, is not a problem with network
analyzer.

Regards,

Yaping Zhou
Motorola Semiconductor Products Sector

----- Original Message -----
From: "Steve Rogers" <SRogers@xxxxxxxxxxxxx>
To: "'Ozgur Misman'" <omism@xxxxxxxxx>; <si-list@xxxxxxxxxxxxx>
Sent: Wednesday, November 21, 2001 3:21 AM
Subject: [SI-LIST] Re: Inductance of Spiral Inductors - why not to use a
network analyser for Q measurement


>
> Hello Ozgur,
>
> Heres my reasoning.....
>
> First I must be specific in that this is a problem with high Q inductors
or
> more to the point the big problem is where the imaginary part is much
larger
> than the real part (This gives you high Q).
>
> The way the network analyser tells you the real (resistance) and imaginary
> (reactance) of the inductor you have connected is to measure the magnitude
> and phase of the forward and reflected components. Now.... draw yourself a
> phasor diagram for the Real and imaginary of a very high Q
> inductor.......... have you done it yet????? Now what sort of size is the
> angle you are trying to measure accurately. More than this... do some
quick
> sums on how much the calculated Q factor moves for a very small angle
> shift...... you will be shocked (assume Q factors of around 100 or more.
> Network analysers are ok just so long as the ratio of real to imaginary is
> not severe!!! I hope this helps..... By the way I have found that most
> engineers are not aware of this, me included until recently when I tried
to
> measure some VHF spirals with a network analyser.
>
> Regards,
>
> Steve
>
> -----Original Message-----
> From: Ozgur Misman [mailto:omism@xxxxxxxxx]
> Sent: 21 November 2001 00:51
> To: SRogers@xxxxxxxxxxxxx
> Cc: si-list@xxxxxxxxxxxxx
> Subject: Re: [SI-LIST] Re: Inductance of Spiral Inductors
>
>
>
> Hi Steve,
> I was curious  why one should not use a network analyzer to  get the Q of
> the inductor. A few years back when I was working on spiral circular
> inductors ( or printed inductors in general). I always did network
analyzer
> ( S-parameter )  measurements.  Once you have the 1 port or 2 port
> S-parameter data you can calculate the Q factor. It is a ratio of
> imaginary part/real part of input impedance which you can easily deduce
> from S-parameter data whether it is a 1-port or 2port inductor.
> I have read  lot of both journal / conference papers in the area that use
> same S-arameter techniques to get the Q , and  L of the printed inductors.
> You seem to have worked extensively in this field. I just would like to
> know very briefly  why network measurement  is not a good approach?
>
> dont use a network analyser to get the Q of the spiral (if the
> Q is high) I Wont bother giving you the what whens and whys but believe me
> when I tell you it wont work. I suggest either a boonton meter (not much
> chance you have one) so in the real world use a tracking generator, a
> spectrum analyser and resonate your spiral with a high quality capacitor
>
> Best Regards,
> Ozgur Misman
>
>
>
>
>
> Steve Rogers <SRogers@xxxxxxxxxxxxx>@freelists.org on 11/20/2001 10:22:10
> AM
>
> Please respond to SRogers@xxxxxxxxxxxxx
>
> Sent by:  si-list-bounce@xxxxxxxxxxxxx
>
>
>
> To:   si-list@xxxxxxxxxxxxx
> cc:
>
>
> Subject:  [SI-LIST] Re: Inductance of Spiral Inductors
>
>
>
> Hi Rafael,
>
> Heres some refs that may be of use for designing inductors
>
> 1. Inductance Calculations and working formulas ///Fredrick Grover
>    you may have some trouble getting a copy of this. This is the mother of
> all
>    inductor references.
>
> 2. Radio Engineers handbook ///// Terman. You should be able to get a copy
> of
>    this one. Much of the content is taken from Grover. This gives a number
> of
>    expressions for spirals. If you dont have a copy of this book then get
> one.
>    It may be old but it cuts the mustard.
>
> 3. Simple accurate expressions for planar spiral inductances /// Mohan,
> Hershenson, Boyd and Lee. Get this off the net. This is and IEEE journal
> paper Volume 34 #10 octobler 1999.;
>
> 5. Do a search for "C PATRICK YUE" he has written lots of stuff re spirals
> on silicon. Also Stanford university are very much into planar spirals on
> silicon. I suggest you have a dig on their site.
>
> 6. Free EM simulator analyses spiral inductor on silicon /// J Rautio.
This
> is a paper by James Rautio of sonnet. This paper is worth reading (the guy
> knows his stuff)
>
> 7. Inductance Calculation techniques part 2 /// Marc T Thompson. This kind
> gent has put a short paper on the web with lots of useful inductor
> formulas.
> There are a number here for spirals.
>
> 8 I almost forgot the classic book by Wadell. I think its called
> Transmission line design handbook. Anyway this book is a real gold mine.
> Get
> one now.
>
> All of this lot is worth a look. Most of it will give you a good stab at
> getting a theoretical inductance prediction. Q is another story.
>
> Not to put you off but I have done a lot of work in this area and it gets
> pretty complicated. Heres some things to think about
>
> (a) Lots of things affect the Q. Some less obvious than others.
>
> (b) If you print the inductor, your currents are confined to a thin
> rectangular strip conductor. The current distribution in this type of
> conductor at high frequency has the majority of the current crowded
towards
> the edges of the strips. The effects are will described (if you like
maths)
> in Collin's micrwave engineering book using conformal mapping. The
> phenomena
> is skin effect.  The problem here is that at high frequency you end up
with
> large conductor losses (I squared R) due to the high current density over
a
> small area)
>
> (c) You have to put tracks next to each other to increase the mutual
> inductance (is there another way?) by doing this you alter the current
> distributions even more making the bad situation even worse (this is
> proximity effect)!!!!! I would not dream of trying to sort this out with a
> pen and paper (E-M solver needed - but!!!!!!! see later)
>
> (d)The PCB traces you can fabricate in reality are not perfectly smooth at
> their edges. This is a big big problem that is overlooked all the time. I
> have made lots and lots of measurements and performed lots of simulations
> using various E-M tools. The rough edges of real life conductors push the
> losses up!!!!!! Wadells book covers the subject in some detail. The effect
> seems to be at max a degradation by a factor of 2. You could try to model
> the surface roughness (if you were barking mad) remember that you would
> need
> to have detailed mesh down to sub skin depth with a model size that would
> encompass your spiral.
>
> (E) What about dielectrics. If you are running a high ish frequency, you
> will need to consider the tan delta of your substrate.
>
> (F)Wheres the spiral going???? If its only ever going to run in free space
> your fine. The reality of it is that it will be in some form of enclosure.
> If you have any conductor near the spiral you will have great fun with
eddy
> current effects. In addition to this if theres a poor dielectric around
the
> E fields will give you grief as well!!!!
>
> (g) Another one to watch out for is the coating nightmare. I made some
> spirals once and they had 60/40 tin lead coating. This is not the best
> approach to High Q inductor design given the conductivity of the coating.
I
> wouldnt like to model this either given that it will form a convex
miniscus
> on the rough surface of the conductor.
>
> (h) Return current. Heres another good one that gets overlooked. You pump
> current through your inductor, but it must return to the supply. Where is
> this return? If your inductor is far from this return you have a good
> antenna. Putting a ground plane under the inductor as a return will be a
> problem. Eddy current will circulate in the ground plane, this will be an
> image of the current in the spiral but in the opposite direction. This
> image
> current causes a negative mutual inductance which drops your spiral
> inductance (Q therefore goes down - oops). In additon the eddy current
> gives
> I squared R loss in the ground plane. You might think you can make a
> patterned ground shield, trust me these only work on silicon (terminate
the
> E field before it gets into lossy silicon)
>
> ---------------
>
> So an E-M tool will give you the best stab at the inductance and Q for a
> spiral. It will account for all but not the surface roughness. Your guess
> on
> that one!
> E-M is how I would go if I lost the use of my test equipment. What I am
> saying I think is use whatever method to design a basic spiral for around
> the inductance you need. Then, build one and measure it. One final
> warning..... dont use a network analyser to get the Q of the spiral (if
the
> Q is high) I Wont bother giving you the what whens and whys but believe me
> when I tell you it wont work. I suggest either a boonton meter (not much
> chance you have one) so in the real world use a tracking generator, a
> spectrum analyser and resonate your spiral with a high quality capacitor
>
>
> Good luck, let me know more about what exactly it is you are doing
>
>
>
>
> Regards,
>
>
> Steve Rogers (Purveyor of many fine and some really useless novel planar
> inductors)
>
>
>
>
>
> Good luck and I hope you have not lost all hope......
>
>
>
>
>
>
> -----Original Message-----
> From: Rafael Martinez [mailto:rmartinez@xxxxxx]
> Sent: 20 November 2001 15:57
> To: si-list@xxxxxxxxxxxxx
> Subject: [SI-LIST] Inductance of Spiral Inductors
>
>
> Hi there,
>
> I am a postgraduate student and I need to design a couple of PCB printed
> spiral inductors for my project. To do that we have purchased Ansoft
> Ensemble 8.0 full version 3D simulator included. So, I would like to know
a
> couple of things about this.
>
> 1.- As far as I know I need to get any sort of empirical formula to relate
> the dimensions of my inductor with the value of the inductance. Does
> anybody
> know any reference or any website useful for me?
>
> 2.-The second step is to simulate the model... Does anybody know if with
> the
> Ansoft package there is any straight forward method to extract the value
of
> the inductance and the Q? I mean the kind of simulation I have to do, what
> kind of analysis.
>
>
> Thank you very much,
>
> ___________________________________________________________________
>
> Rafael Martinez Catala        Rossa Avenue, Bishopstown,
> Postgraduate student          Cork, Ireland
>
> Cork Institute of Technology        Fax: +353 21 4326625
> Electronic Engineering Dep.         Tel: +353 21 4326858
>
> http://www.cit.ie
> ___________________________________________________________________
>
>
> ------------------------------------------------------------------
> To unsubscribe from si-list:
> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
>
> or to administer your membership from a web page, go to:
> http://www.freelists.org/webpage/si-list
>
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>
> List archives are viewable at:
>           http://www.freelists.org/archives/si-list
> or at our remote archives:
>           http://groups.yahoo.com/group/si-list/messages
> Old (prior to June 6, 2001) list archives are viewable at:
>           http://www.qsl.net/wb6tpu
>
>
>
> _____________________________________________________________________
> This message has been checked for all known viruses by the
> MessageLabs Virus Scanning Service. For further information visit
> http://www.messagelabs.com/stats.asp
> ------------------------------------------------------------------
> To unsubscribe from si-list:
> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
>
> or to administer your membership from a web page, go to:
> http://www.freelists.org/webpage/si-list
>
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>
> List archives are viewable at:
>           http://www.freelists.org/archives/si-list
> or at our remote archives:
>           http://groups.yahoo.com/group/si-list/messages
> Old (prior to June 6, 2001) list archives are viewable at:
>           http://www.qsl.net/wb6tpu
>
>
>
>
>
>
> _____________________________________________________________________
> This message has been checked for all known viruses by the
> MessageLabs Virus Scanning Service. For further information visit
> http://www.messagelabs.com/stats.asp
> ------------------------------------------------------------------
> To unsubscribe from si-list:
> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
>
> or to administer your membership from a web page, go to:
> http://www.freelists.org/webpage/si-list
>
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>
> List archives are viewable at:
> http://www.freelists.org/archives/si-list
> or at our remote archives:
> http://groups.yahoo.com/group/si-list/messages
> Old (prior to June 6, 2001) list archives are viewable at:
>   http://www.qsl.net/wb6tpu
>
>

------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List archives are viewable at:     
                http://www.freelists.org/archives/si-list
or at our remote archives:
                http://groups.yahoo.com/group/si-list/messages 
Old (prior to June 6, 2001) list archives are viewable at:
                http://www.qsl.net/wb6tpu

Other related posts: