Talk:Dielectric loss

Proposal: merge with Dissipation factor

There is a big overlap between these two articles. More, I saw a very strange redirect. Loss Tangent was not pointing to Loss tangent (which exists) but to Dissipation factor. Now it points to the right one, but there is the overlap problem yet. I believe these can be merged together. Sorry I can't help much, beacuse I don't have much time and I don't know very much about this subject, but only a little bit. Kar.ma 18:20, 29 December 2008 (UTC)

This article is best kept as Loss tangent, with Dissipation factor pared down to a stub. Dissipation factor uses electrical Loss tangent as one example of the more general property of Dissipation factor, and probably should not go into such detail of Loss tangent. Dissipation factor should instead concisely explain additional examples, such as mechanical resonators. This concentration on Loss tangent in Dissipation factor probably orginated due to specifications given with commercial capacitors. —Preceding unsigned comment added by 128.84.45.231 (talk) 15:54, 8 January 2009 (UTC)

Dielectric loss tangent is the same as dissipation factor but I am not sure about loss tangent on its own. —Preceding unsigned comment added by 90.209.185.205 (talk) 17:14, 9 April 2009 (UTC)

Rather than lumped in with Loss Tangent, this article should be expanded. The lumped element model given for the capacitor shows resistance in series. The example should be expanded to include resistance in parallel as well. —Preceding unsigned comment added by Bmcnulty (talk • contribs) 18:34, 16 July 2009 (UTC)

Loss tangent has specific uses in radio engineering for loss in substrates, antennas, lenses, distributed circuits and the like whereas dissipation factor is predominantly used in other applications such as to describe the losses in a capacitor. Dissipation factor is used in radio engineering, but often in a much different way. The definitions should be kept separate. —Preceding unsigned comment added by 68.117.34.130 (talk) 06:20, 3 August 2009 (UTC) The Dipole Relaxation phenomenon (memory in dielectric) is a critically important concept for systems containing storage elements found in resonant structures. Most researched in capacitive elements of signal processing, the phenomenon will also be found to be a contributor in MEMS and other mechanical systems, as well as systems referred to as "Single Photon Optics," as we progress. This section warrants expansion when someone has time and refreshed awareness. Since the roots of our language and usage of these terms is a bit mixed, I'm thinking we should leave the two topics (Dissip' factor and Loss Tangent) seperate in order to best promote contribution by individuals focused on a region of expression localized to the roots of one topic or the other. In any case, the topics should reference one another for the most thorough understanding and association of the possible functions and properties. 138.210.84.242 (talk) 15:02, 31 July 2009 (UTC)

It is wrong to equate dielectric loss tangent with dissipation factor as .205 suggests above. A dissipation factor can be defined for a coil, which has nothing to do with dielectrics. Also for a resonant circuit, the dissipation now mainly due to the coil resistance with dielectric losses usually secondary. Equally, the phrase loss tangent can in principle be applied to things other than dielectrics, but usually it it is limited to dielectrics and capacitors (at least in my experience). I suggest that "loss tangent" is renamed to "dielectric loss tangent" and "dissipation factor" is left as the general article with a link to the specific case of dielectrics. SpinningSpark 14:40, 22 August 2009 (UTC)

Imaginary part of the complex permittivity

I would like to report an inconsistency in the definition of the imaginary part of the complex permittivity among different wikipedia articles: In this article imaginary part ε shall not include conductivity σ, while in "https://en.wikipedia.org/wiki/Permittivity" ε = σ/ω. — Preceding unsigned comment added by Rpetrillo (talk • contribs) 08:33, 20 May 2014 (UTC)

ESR in series with the ideal capacitor

This doesn't make sense. The voltage between the terminals of the leakage resistor and the ideal capacitor should be the same. 131.111.5.156 (talk) 16:28, 22 February 2018 (UTC)

Why would you say they should be the same? That is simply not true. The only voltages that need be the same would be the voltage across the real-world capacitor, and the voltage across the combined ideal capacitor and resistor. Which they would be. — Preceding unsigned comment added by Debeo Morium (talk • contribs) 19:29, 25 September 2020 (UTC)

Describing the difference between the two definitions of Loss Tangent

Hi Constant314, I feel this article needs some further clarification as is obvious from the edit I tried to make noted here which you revered: https://en.wikipedia.org/w/index.php?title=Dielectric_loss&oldid=prev&diff=980335056 rather than butting heads lets discuss your concerns and see where we can reach a good compromise between adding further clarity as well as addressing your concerns with my previous attempts. The lack of clarity in the article cost me about a week worth of effort in a paper I've been trying to write I intended to publish in a peer reviewed journal and I was hoping by adding additional clarity I could help prevent the same confusion befalling others. - Debeo Morium: to be morally bound (Talk | Contribs) 00:13, 26 September 2020 (UTC)

Hello. I have no problem with pointing out that loss tangent approaches ${\varepsilon ''}/{\varepsilon '}$ as either conductivity becomes small or frequency becomes high. That discussion should probably go right after the first equation of the section. You need to avoid language such as "we should" and "we must" and "we can see". That is text book voice. Wikipedia doesn't tell people what they must do, should do or can see. It is probably better to write the approximation equation this way

$\lim _{\omega \to \infty }{\frac {\omega \varepsilon ''+\sigma }{\omega \varepsilon '}}={\frac {\varepsilon ''}{\varepsilon '}}$

which reflects the most common usage. That is, frequencies are usually high enough that dielectric loss dominates loss due to conductivity. In typical materials regarded as dielectrics, that frequency is below 1 Hz, so, in practice, it is not very important. However, I think that it is definitely appropriate to point out the overall loss in a material serving as a dielectric can be a lot larger if the dielectric also has magnetic loss (for example dirt with iron in it). However, as ${\varepsilon ''}/{\varepsilon '}$ is defined as electric loss tangent you can say that the electric loss tangent is an approximation for the dielectric loss tangent but that is getting pedantic. So, I suggest, discuss approximations of the first equation, rather than the last two equations which are just definitions. Constant314 (talk) 01:12, 26 September 2020 (UTC)

That seems reasonable, I can take a stab at that. The other thing I'd like to point out, though im not as clear on this so didn't add it yet and am still talking with physicist friends to get some clarity, is an equivalent loss to the first equation as would be seen in magnetic circuits. I believe it would be some combination of magnetic loss and conductivity/resistivity as well. Any thoughts on that? - Debeo Morium: to be morally bound (Talk | Contribs) 12:01, 26 September 2020 (UTC)

I am actually starting to doubt my initial interpretation of the first equation that includes conductivity all together. If I actually measure the complex permittivity of a material at a particular frequency using a network analyzer, wouldnt the loss due to conductivity be rolled into the complex impedance? I am guessing in order to actually figure out what part is due to conductivity you'd have to separately measure DC resistance and then work the equation backwards? - Debeo Morium: to be morally bound (Talk | Contribs) 12:21, 26 September 2020 (UTC)

Yes. The impedance analyzer does not separate conductivity from dielectric loss. They both cause a phase shift relative to 90 degrees. You typically measure the conductivity using DC and assume everything else is dielectric loss.Constant314 (talk) 19:43, 26 September 2020 (UTC)