13119	2016-09-04 03:25:51	Ashhar Farhan	Inductor Q in LTSpice
Message	Date	From	Subject
how does one specify inductor Q in LT Spice? If Q is also expressed as the ratio between its reactance and resistive losses, then will Q linear drop with increasing frequency? - f
13121	2016-09-04 05:57:52	chuck adams	Re: Inductor Q in LTSpice

13122	2016-09-04 06:46:40	Ashhar Farhan	Re: Inductor Q in LTSpice
chuck, to establish the context, i am designing some low pass filters with notches in them. these are not possible with the GPA.EXE of the EMRFD hence I am having to use LTSpice. I had developed a blind spot to the simple relationship that Q = X / R for an inductor. As X is the reactance that increases with frequency, then the Q is frequency dependent too! Should we talk of 'this toroid has a Qu of 200 at 10 MHz' instead of saying 'this toroid's unloaded Q is 200'? - f
13124	2016-09-04 07:20:35	Claudio Girardi	Re: Inductor Q in LTSpice
yes, inductors Q is frequency dependent; see these Q charts for common toroids for some typical curves of Q vs frequency http://www.micrometals.com/appnotes/appnotedownloads/ipc4hqi.pdf . Not only X changes with the frequency, also the parasitic R changes with frequency too. If you are interested in the losses over a narrow band only you can just use a series or parallel resistor, as mentioned earlier. With a more complex circuit you can model the inductor behavior over a wider bandwidth, see http://m.eet.com/media/1142818/19256-159688.pdf . 73 de Claudio, IN3OTD ----Messaggio originale---- Da: "Ashhar Farhan farhanbox@gmail.com [emrfd]" Data: 4-set-2016 1.46 PM A: "emrfd@yahoogroups.com" Ogg: Re: [emrfd] Inductor Q in LTSpice chuck, to establish the context, i am designing some low pass filters with notches in them. these are not possible with the GPA.EXE of the EMRFD hence I am having to use LTSpice. I had developed a blind spot to the simple relationship that Q = X / R for an inductor. As X is the reactance that increases with frequency, then the Q is frequency dependent too! Should we talk of 'this toroid has a Qu of 200 at 10 MHz' instead of saying 'this toroid's unloaded Q is 200'? - f
13127	2016-09-04 08:28:05	davidpnewkirk	Re: Inductor Q in LTSpice
> As X is the reactance that increases with frequency, then the Q is frequency dependent too! > Should we talk of 'this toroid has a Qu of 200 at 10 MHz' instead of saying 'this toroid's unloaded Q is 200'? Yes, Q is frequency dependent. Look up inductor specs at an electronic-parts supplier--say, Digi-Key--and when you see Q values you'll see a measurement frequency associated with each. Lab-standard commercial Q meters use/d particular frequencies to measure Q for different inductance ranges, so you will tend to see particular Q-measurement/specification frequencies over and over--790 kHz, 2.52 MHz, 7.9 MHz, and so on, depending on the inductance of the part evaluated. Yes, frequency should be mentioned whenever measured Q, or a value of Q necessary to achieve a particular aim, is called for--but context implies this. If we're designing for 7 MHz and need an inductor with a Q of 200, we need that inductor to exhibit that Q in the vicinity of 7 MHz, not at some remote frequency of non-interest. Circuit simulation software that supports characterization of inductor Q will therefore allow input of Q value and the frequency at which that value of Q obtains. Some simulators even allow selection of different Q models (skin effect operative or not, and so on.) As other correspondents have said, however, especially for narrowband simulation work, seriesing with an ideal inductor the (usually small) value of resistance that equates to a given Q for that inductor at the modeling frequency can give accurate results. Note, however, that this technique may make the behavior of each such inductor less realistic at dc, a possible issue if dc through the inductor is high and accuracy of dc behavior is important to the simulation. Best regards, Dave amateur radio W9BRD radio pages: http://dpnwritings.nfshost.com/ej/