Request: Study of Damping Factor and Planar Magnetic Headphones

Discussion in 'Measurement Techniques Discussion' started by Klasse, Feb 16, 2016.

  1. Klasse

    Klasse Friend

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    117
    Trophy Points:
    43
    It's a well known fact that low damping factor leads to EQ in the shape of the headphone's 'impedance vs frequency' plot.
    (ie. the amplifier is not flat for headphones with variable impedance). This is a pretty simple phenomena that can be measured, it's audible and pretty easy to put on numbers.

    I'm interested in the measurability of the effects of Low Damping Factor in headphones with a flat impedance.
    Not sure if someone has done this kind of study in the past. If you know one, please share the link.

    I've used a K702 in the past, plugged into a Yamaha A-S500 (470 Ohm output impedance)
    Extreme case: damping factor = 0.13
    And the sound was still ok, not ultimately transparent, but not ruined either.

    So it would be really interesting to have a few measurements that put these effects in perspective.

    It would be ideal to measure a planar with really flat impedance like the LCD-2 plugged into an amplifier with really low output impedance (say, less than 2 Ohm) and the same headphone plugged into a similar amplifier with 50 Ohm or higher output impedance. Maybe a resistor can be used in series with the 2 Ohm amplifier to simulate the 50 Ohm, similar performing amplifier.

    The objective is:
    Frequency response, harmonic distortion and CSD in both cases:
    1- High Damping Factor (Less than 2 Ohm output impedance)
    2- Low Damping Factor (More than 50 Ohm output impedance)

    Thanks,
    Klasse

    Relevant Findings: #48
     
    Last edited: Mar 6, 2016
  2. philipmorgan

    philipmorgan Member of the month

    Pyrate BWC
    Joined:
    Sep 28, 2015
    Likes Received:
    3,790
    Trophy Points:
    93
    Location:
    In the wind, so to speak
    Home Page:
    I have a Dayton measurement mic coming today and will be putting together a measurement rig based on that asap. I also happen to have HD650, Alpha Dog, and Geek Out 1k v1 (with the 0.47 and 47 ohm output impedances).

    So... thought I'm a total measurement n00b, I'll give this a shot and report back.
     
  3. chakku

    chakku Friend

    Pyrate
    Joined:
    Oct 13, 2015
    Likes Received:
    676
    Trophy Points:
    93
    Location:
    NZ
    Damping factor doesn't really affect planars because they are purely resistive loads and that is precisely why they have flat impedance curves.
     
  4. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,464
    Trophy Points:
    113
    Location:
    Winnipeg
    If we're gonna argue that, then the flip side is that planars still technically have mass and a natural resonance due to tension. Just because nothing shows up on an impedance curve doesn't mean there isn't unwanted movement/energy to damp out.
     
  5. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,777
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    It affects power transfer. High source impedance relative to input impedance = lost power.

    I'll draw a simplified diagram when I get back into the office. Or maybe @OJneg will. Think of source resistance modelled as a resistor in series with the transducer.

    The notion that damping factor (Zout/Zin) doesn't do shit to planars (flat resistive load) is a misconstruction from the fact that FR will not be affected.

    Power and indirectly, distortion (and transient response) can be negatively impacted. Significantly if the damping factor is low.

    One only needs to try very low Z inefficient planars on amp like the ZD super with the Zout on high (20 ohms).
     
    Last edited: Feb 17, 2016
  6. chakku

    chakku Friend

    Pyrate
    Joined:
    Oct 13, 2015
    Likes Received:
    676
    Trophy Points:
    93
    Location:
    NZ
    I mean if you wanted to go really extreme with the output impedance for testing/measurements you could try them straight out of a DAC, right? Obviously not useful for real world stuff but the extremity of it would provide a clearer idea of what is happening. Would just need some more sensitive 'portable' headphones like the PM-3 or HE-400S.
     
    Last edited: Feb 17, 2016
  7. Colgin

    Colgin Friend

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    634
    Trophy Points:
    93
    Although it is pretty efficient for an ortho, my Oppo PM-2 sounds way better on the Cavalli Liquid Carbon than on my high output impedance Marantz amp (which is otherwise quite good with my high-impedence Senn HD 600). Maybe this is one of the reasons why.
     
  8. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,464
    Trophy Points:
    113
    Location:
    Winnipeg
    The furthest I've tested was on my HE-6 with and without a 15ohm resistor in series with the output (correcting for volume difference of course). I thought there was a difference, but it was small enough that I could have dismissed it as placebo. My only takeaway from that is 15ohms was not enough for me to confidently detect a difference.
     
  9. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,777
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Yes and no. The DAC output would be an extreme case and not applicable. It comes down to available power. If an amp has sufficient power to overcome the poor damping factor, then there might not be any deleterious effects. I have an HE-560 lying around. I'll take some measurements when I get a chance. Maybe add a resistor in series to the output of Vali 2 to simulate higher Z or something that like that. This week is super busy for me.

    FWIW, damping factor is a huge consideration in sound reinforcement. Long speaker cable lengths and high powered amps... you get the idea. This is where cable gauge and resistance is a BIG deal.
     
  10. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    I believe that for optimal power transfer the amp output impedance and the headphone impedance need to be matched. That is, be the same.

    This will work fine if the load is resistive. If the load is not resistive, then one needs to see if the driver is designed for voltage or current drive. Most are designed for voltage drive, so the load in this case should be a few times higher than the source resistance to avoid undesired coloration.

    As far as why impedance should be matched for maximum power transfer, one needs to understand that power is a voltage and current affair. If we increase the load quite a bit, the voltage across the load increases, but current decreases. If we decrease the load the current increases but voltage decreases. To get the result one needs to take the derivative of the expression describing the power across the load and find the load value that maximizes that expression. It turns out to happen when the load resistance matches the source resistance.

    As far as damping and mechanics, there is mechanical damping and electrical damping. In the context of electronics and load interactions I believe we would be concerned with electrical damping.

    In the context of long wire lengths, power transfer is indeed a huge consideration.

    This sort of stuff presents some problems when we are dealing with a headphone designed for voltage drive, whose impedance is complex and require quite a bit of power to drive. I believe some AKGs have these characteristics.
     
  11. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,464
    Trophy Points:
    113
    Location:
    Winnipeg
    Most decent cabling should get you at least 100' without appreciable loss (and if you're running a mic cable for power then you oughtta be slapped upside the head), perhaps half that length if you're bridging your amps and running some stupid amount of subs to crank enough oomph to help you pass a kidney stone. If you're going greater distances than that, then serious consideration should be given to placing amps closer to the speakers and keeping your long runs for signal lines only. If you're going even further, then fidelity is probably not a main concern and you should be switching to a constant-voltage distribution system where wire gauge is less of a concern.
     
  12. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,777
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Doesnt the max power transfer theorem apply in a different context?

    Doing a thought exercise. Say there is a voltage source of 100V into a load of 20 ohms, something like an LCD-X.

    If source impedance is 20 ohms, the power dissipated on the load would be 125W.

    If the source impedance is zero, then the power dissipated in the load would be 500W.
     
    Last edited: Feb 18, 2016
  13. feilb

    feilb Coco the monkey - Friend

    Pyrate
    Joined:
    Oct 5, 2015
    Likes Received:
    226
    Trophy Points:
    33
    It applies in the same context, but it is being looked at from the wrong end.

    Max power transfer theory says that given a source with a given output impedance, the largest amount of power it can transfer to a load is achieved when the load impedance matches the output impedance.

    It does not say that given a load, the max power transferred to the load is when the output impedance matches the load. From the perspective of the load, the greatest power transfer to the load is when the source impedance is 0.

    Thus, if the ZD has an output impedance of 20 ohms, the largest power transfer out of the ZD is possible when the load is 20 ohms. Above or below that number, you will see lower power transferred to the load.
     
    Last edited: Feb 18, 2016
  14. T.Rainman

    T.Rainman Acquaintance

    Joined:
    Oct 12, 2015
    Likes Received:
    57
    Trophy Points:
    18
    impedance matching is only needed for HF signals such as video, digital signals and RF.
     
  15. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    From what I understand, the max power transfer theorem assumes a known and fixed source impedance. If you source impedance is changeable then the optimum load (in regards to power transfer) also changes.

    So if the source impedance is 20 ohms, then the value of the load impedance for maximum power transfer is 20 ohms. In the limit, if the load impedance is zero or close (a short), power to the load is zero. If the load impedance is infinite or close (an open), the power to the load is also zero. Zero/Infinite impedance is sort of impossible and ideal though. I mean, there are negative impedance active circuits, but they are not infinite power generators.

    Anyhow, more realistically, if the source impedance is close to zero, say 0.5 ohms, then maximum power transfer happens if the load is 0.5 ohms.

    Here is an example with 100V from the source:

    1) Source is 20 ohms, then maximum power disipated through the load (20 ohms) would be 125 W.
    2) Source is 0.5 ohms, then maximum power disipated through the load (0.5 ohms) would be 5 kW.

    Obviously something is going to give for the 5 kW case. Audio amps capable of 100V and with 0.5 ohms of source impedance will likely current starve before they get to 5 kW (100 Amps). Possible the circuit breakers will kick in as well.
     
    Last edited: Feb 18, 2016
  16. feilb

    feilb Coco the monkey - Friend

    Pyrate
    Joined:
    Oct 5, 2015
    Likes Received:
    226
    Trophy Points:
    33
    This is not for power transfer reasons, though. It is for signal integrity reasons.
     
  17. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    If I remember correclty, impedance matching reduces reflections (echo). I think it's more of a transmission line consideration.

    It affects all frequencies, also from what I remember.
     
    Last edited: Feb 18, 2016
  18. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,777
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Got it. Depends on which way to look at it and what factors can be controlled, compounded by the fact that real world devices are current limited.
     
  19. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
  20. logscool

    logscool Friend

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    69
    Trophy Points:
    28
    It is only important to consider for high frequencies because these are the frequencies where nearly any length of cable run becomes a "transmission line". A transmission line is defined as having a length greater than λ/8. For this reason in the audio bandwidth (up to 20 khz) anything less than 1875 meters!!! is NOT considered a transmission line at these frequencies. Obviously when dealing with RF signals this is a hole different story and I'm also not suggesting that there are no signals above 20 khz being output by audio equipment.
     

Share This Page