Planar Speaker Asylum

Welcome! Need support, you got it. Or share your ideas and experiences.

Return to Planar Speaker Asylum

Planar Speaker Asylum

attenuation in RF region

198.26.119.84

[ Follow Ups ] [ Post Followup ] [ Thread: Display Email ] [ Planar Speaker Asylum ] [ FAQ ]

Mart

In Reply to: Re: to make a long story short ... posted by monk1 on October 08, 2001 at 12:23:23:

is due to skin-effect. You can witness the beginnings in some audio componenets in the audio spectrum. EG: I've measured the impedence of some conventional coils. Above 1kHz, the impedence still looks normal but inuductance is being replaced by resistance. You see, as the current migrates to the surface of the wire, less of the wire crossection is used. This increases the effective resistance. Simultaneously, while less current is drawn less magnetic field is created. So, the inductance is less. Thus, the magnitude of the impedence curve looks normal while the phase doesn't.
Obviously this skin effect increases with frequency until by the MHz region the resistance shelves the output & any remaining signal is smeared by the dielectric of the insolation. Hope this helps to see how my linear model breaks down outside it's design region? However, this does explain a lot. BTW, I still don't know why you can't download the zipped excel file which models a few lumped models with a 201th order model. Damn, I wish you could because I included Zobels across the amp & speaker for which you can expirament.
POLES: I don't see how using a Zobel adds a pole since it doesn't become a short that would zero the dominator, unlike my disc cap tweak the shorts out the couple of MHz region you witnessed doesn't conduct anyway.

... just my 2�
� moderate Mart � �Planar Asylum

Follow Ups:

Re: attenuation in RF region - Monk1 11:01:51 10/09/01 (5)
I would agree the skin effect is a factor at the higher frequencies but, the skin penetration at 20 kHz for copper is nearly 0.5 mm. This is more than sufficient strand penetration, even for the skimpiest stranded conductor audiophile speaker cable that I have seen. At 20 kHz, the conductor strand diameter would have to be greater than 1 mm before the skin effect begins. Most sranding in audiophile cables are much finer than that so skin depth should not be an issue.
For high frequency work, there are many different forms of inductance to consider. Self inductance, mutual inductance, internal inductance, partial inductance are only the tip of the iceberg. The dominant form of inductance in a speaker cable is self inductance, which is entirely dependent on path geometry. The resistance due to skin depth is accounted for separately and does NOT replace self inductance, even at radio freqencies. You can easily account for the skin depth resistance in the lumped-circuit model without having to resort to the transmission line model.
On poles and zeros, it may appear on initial inspection that the simple RC "zobel" network would not, by itself, create a pole. However, we're dealing with a complex network with many circuit elements, some of them are parasitic. One would have to consider all variables on a case by case basis because not all speaker cables and loudspeaker crossovers are the same regardless which model is chosen. Use of the "Zobel" directly across the amplifier output should be done with caution. let's not forget that the resistor in the "Zobel", even a low inductance one, turns into an inductor (typically 10 nH or more) at high frequencies. This inductor can resonate with the circuit capacitance to produce... well, more resonances.
BTW, I searched the site and did find the plots and Excel models afterall. Good luck with your experiments.
Monk
- 1mm penetration is a 1D model - Mart 11:34:36 10/09/01 (4)
  which doesn't account for axisymmetry & also doesn't account for material E & M parameters each field of which has independent permitivity & permiability. Although, this does give the engineer a ballpark figure for qualitative design decision purposes.
  Yes, I did measure inductor variances in 12ga round-wire large-guage air-core chokes. However, one also notices this behavior in dome tweeters in frequencies above 10kHz too. Most impedences are asymptotic to ~45�. Even the ones without a steel parts that comprise a magnetic circuit to increase inductance & blame hysteretic lag.
  
  ... just my 2�
  � moderate Mart � �
  Planar Asylum
  - Re: Skin depth - Monk1 14:04:04 10/09/01 (3)
    Even if we factor in a dielectric like polypropylene with a relative permittivity of 2.5, the skin depth is still an inconsequential 0.3 mm. Deeper penetration is preferred. What this result tells us is that most speaker cable strands would have sufficient skin depth to ensure full penetration thru the skin surface to avoid the skin effect when the cable construction does not involve magnetic materials. My main point was: skin depth resistance, no matter how inconsequential, can still be accounted for in the lumped-circuit (RLC) model. The transmission line model is neither necessary nor appropriate for loudspeaker cables.
    As a general rule, if the cable length is less than one tenth of the wave length at the frequency of concern, then it is for certain not a transmission line.
    - Re: Skin depth - Mart 14:55:10 10/09/01 (2)
      I wasn't even too worried about insolation permeability & permitity in my previous post but rather the interference of speaker cabling projected E & M field penetration & dynamic interaction in addition to itself.
      I wasn't intending to bring up transmission lines which can experience reflections & other non-linear circuit phenomena. I don't know why you keep bringing it up. Care to enlighten me on this?
      
      ... just my 2�
      � moderate Mart � �
      Planar Asylum
      - Re: Transmission lines - Monk1 15:42:42 10/09/01 (1)
        
        In Reply to: Re: Skin depth posted by mart on October 09, 2001 at 14:55:10:
        
        I was merely pointing out that the circuit diagram you posted earlier was the equivalent circuit for a transmission line and... that transmission line effects do not apply to audio speaker cables. I believe we are really barking up the wrong tree by applying the transmission line model to loudspeaker cables. That does not mean you can't try it. (=:
        Monk
        
        just substituted a long cable length - Mart 00:42:58 10/23/01 (0)
        
        In Reply to: Re: Transmission lines posted by Monk1 on October 09, 2001 at 15:42:42:
        
        and around 125ft the RF reactance instability seemed to go away.
        ... just my 2� � moderate Mart � �
        Planar Asylum

This post is made possible by the generous support of people like you and our sponsors:

You can not post to an archived thread.