1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

A very high damping factor=Overdamping headphones?

Discussion in 'Sound Science' started by gustavmahler, Jan 17, 2015.
2 3 4 5 6 7 8 9 10 11
  1. GustavMahler
    Can i overdamp headphones by a really high damping factors? Is there a maxinmum recommended damping factor or only a minimum?
  2. jodgey4
    I think it would really depend on the parameters the engineers designed the cans on. I would assume most engineers assume we'll be using amps that are trying to be close to ideal voltage source - output impedance of <2 ohms or so. Sennheiser says the HD800 was designed for ideally zero output impedances, though their own amp is a bit higher.
  3. Dogmatrix
    I don't think "damping factor" has much real impact on sound either way
    I own amps with high output impedance (over 100) and low (less than 1) and cans with impedance of 32 , 300 , 600 and I don't hear anything I could attribute to "damping factor" when I mix and match
    Each headphones character is unchanged by amp impedance
  4. jodgey4
    http://www.innerfidelity.com/content/musings-headphone-amplifier-output-impedance agrees with @Dogmatrix mostly for electrical damping.
    SilverEars likes this.
  5. Silverprout
    It just change the frequency response of the heardphones...
    Well... it's like bending an archer, how much force is needed to hit the center of the ciber ?
  6. GustavMahler
    Isn't the varying frequency response of the headphones created by the varying impedance?
  7. stv014
    Not really, since electrical damping is limited by the resistance of the voice coil. In other words, driving headphones with an impedance of 250 ohms with an amplifier that has an output impedance of 0.25 ohms instead of 25 actually only increases the damping by about 10%, and not by a factor of 100. Although the latter figure obviously looks better for the purpose of marketing.
  8. castleofargh Contributor

    he was probably talking about headphones with non flat impedance responses over frequency.
    just like dogmatix opinion isn't true in a general sense, but would work fine with headphones that have a relatively flat impedance over the frequencies. and to be true it also needs the amp to be ok with loads of low values. be it for providing enough current or simply at some point to get rid of the excess energy, the amp would have to be able to do it, and all in all I don't imagine much situations where a measurably better sound can be obtained by not following impedance bridging rules. amps don't love small loads, that is always true. but that's more a problem with underdamping so not really what you were asking.
    I personally would tend to go for low impedance amps whatever the headphone I need to drive it with. in fact I would go for low impedance because I plan on using several headphones with different specs. having a good damping if anything is the best way to keep FR variations in check. so going for 1/200 appeals to me more than the usual 1/8, as a precautionary measure for the headphones where I ignore all about the impedance response.
  9. SilverEars

    Interesting section about the 32ohm AKG K420.  Thanks for pointing it out. 88ohms OI providing better response than 0ohms? The AKG has high efficiency and 32ohms. How to explain this phenomena?
  10. SimonPac
    Significant amplifier output impedance has two separate major effects.

    First it will alter the frequency and phase responses of the headphone from those obtained with zero output impedance, unless the headphone has completely flat impedance with frequency. This effect may or may not be significant depending on the phones and impedance involved. It is often most pronounced with low impedance dynamic headphones. Typically a higher output impedance will result in bass lift as the headphone bass resonance boosts phone impedance at the corresponding frequencies resulting in less attenuation of the electrical signal. You may prefer the sound that way and the designer may have factored it in.

    Second, amplifier output impedance affects the amount of the electrical contribution to damping (driver motion settling after a stepped input). Amplifiers can dump current out of the transducer as well as provide it and this allows electrical damping. Damping dumps unwanted energy (here due to voice-coil/membrane motion). It is contributed to by both electrical and mechanical factors and is never ideal across the frequency band and for different amplitudes. Damping may be over, under, or 'critical'. Critical damping stops the driver in the shortest time and would generally give the highest fidelity to the electrical waveform. In practice, headphones are complex electrical and mechanical systems with various resonances and often untidy step recovery. The optimum degree of electrical damping they require may not be achievable under all signal conditions and so is not really a clear cut issue. I would expect that few headphones would be over-damped at the mechanical level; I might be wrong. If they are under-damped, sometimes the amplifier will not be able to correct this because of voice coil impedance.

    The rate at which electrical energy can be dumped out of the driver is indeed limited by voice coil impedance which limits current and therefore energy flow. Very often this dominates the situation and the amplifier can't do that much about it. What is called 'damping factor' for an amplifier is usually a measure of the additional loss of damping due to the amplifier output impedance. The higher the number, the less the extra loss of electrical damping.

    Subjectively I feel many dynamic headphones sound more natural/forgiving with a low-ish electrical damping factor. HF resonances can become less noticeable and timbres more realistic. Conversely, at very high electrical damping some headphones seem to sound thin. In reality there is not necessarily a 'right' answer because audio replay transducers are still very far from perfect, and in all cases the way the sound is recorded is a big factor.
    skwoodwiva and Jon Sonne like this.
  11. ev13wt
    There are differences between 1 and 5. They are not really audible.

    I think I remember reading something about the voice coils being responsible for higher damping factors basically being "irrelevant" in a system.

    Found it: Toole again. :) Do buy that book of his! Its a great read.

    Last edited: Oct 17, 2017
  12. 71 dB
    There is damping factor ( = load impedance / source impedance) and damping ratio . ( = actual damping / cricital damping) .

    In control theory damping ratio of 1 is usually desired. A damped swing door with damping ratio 1 will return to it's default position the fastest, which is good. In audio things are a bit different in my opinion. Slow return of signal level to zero happens at very slow (infrasound?) frequencies. That's why I am not worried about overdamping (damping ratio > 1) in audio.

    Some headphones remain overdamped even when damping factor goes to zero. For example, AKG 550 has according to my calculations a minimum damping ratio of over 9 meaning it's strongly overdamped no matter what the output impedance is.
    ev13wt likes this.
  13. castleofargh Contributor
    electrical damping vs mechanical damping.
  14. ev13wt
    With mechanical damping, do you mean the surround of the cone or the air itself? :p
  15. SimonPac
    The innerfidelity (Musings...) article mentioned above purports to be looking at a high (for audio) frequency electrical resonance. The author has modeled it using an equivalent electrical series LCR circuit by assuming that the mechanical factors can be fully represented by 3 lumped series resistive or reactive elements. The ringing he shows as a result of electrical under-damping of this resonance is at about 6KHz. Very many headphones do indeed have a pronounced acoustic frequency response peak at around this frequency. If the transducer mechanical behaviour (to include air resistance, surround/suspension compliance, membrane breakup and anything else relevant) is fully modeled by this LCR electrical network, then it is possible, for some headphone resonant equivalent LCR values, that a high amplifier output impedance could improve this particular situation and bring it closer to a technically optimum critical damping situation. If this was achieved, the acoustic peak would be greatly diminished. In practice, I don't think all the electrical and mechanical factors are fully represented by the electrical model at this sort of frequency.

    Usually we are talking about low frequency primary resonance, which is more typically at around 100Hz. Headphones usually have reasonable mechanical damping here. This LF resonance has been typically modeled by others as a parallel LCR circuit. A pure parallel LCR circuit will be damped very quickly with a low output impedance.

    It seems to me that it follows from all this that it depends on what the electrical model of the electrical/mechanical resonance you are concerned about looks like. Most headphones have several peaks/resonances. Certain resonances may be close equivalents to a series LCR circuit, perhaps the HF ones. If so, those particular resonances may be better damped and therefore less obtrusive with a higher amplifier Z out. The primary LF resonance of a headphone, if, as seems likely, best modeled by a parallel LCR circuit, will very likely be optimally damped by a low amplifier Z out.

    if the innerfidelity article author is correct, I would think that higher amplifier Z out might produce better HF damping for some peaks. It would likely give worse LF damping. Steady state sine wave testing with various z out values does not seem to change HF response much. My conclusion would be that the mechanical resonances in headphones cannot be accurately mapped into simple electrical parameters presented to the amplifier. They cannot be corrected much by choosing output impedance. That is not to say that output impedance might not affect them somewhat, especially in the dynamic music signal situation.
2 3 4 5 6 7 8 9 10 11

Share This Page