Modding Planars: Voodoo, Alchemy, and Science

Discussion in 'Modifications and Tweaks' started by MF_Kitten, Mar 19, 2017.

  1. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,467
    Trophy Points:
    113
    Location:
    Winnipeg
    The housing will feel the opposite motion as the transducer, because of Newton's 3rd law. We typically do not want this, because this both takes energy away from the diaphragm (which is what transmits that energy to our ears) and worsens the performance of said diaphragm.

    Just because you feel vibration with your hands does not mean it was part of the design.

    Putting the headphones on your head mechanically couples the housing to your (supposedly much heavier) head, making it a larger mass system which will lower to resonances. Clamping pressure and pads will alter the amount of mechanical coupling.

    If you prefer the sound of the Abyss floating rather than on your head, I'm 96% certain it has nothing to do with the housing vibrating and more to do with the acoustic properties of sealed earpad/earside chamber.
     
  2. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,467
    Trophy Points:
    113
    Location:
    Winnipeg
    Bill... this thing that you're trying to describe... flat-ish rigid surface with mass that emanates sound with all the conductive traces to the side... that's what we call a dynamic driver.

    No we still don't want to the housing to behave like a transducer, because again that Newton's 3rd law means it is going in the opposite direction. Even if the surface of the housing was emanating that energy as acoustic waves, it is the opposite of what we want.
     
  3. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    One may want the housing to absorb some frequencies at the expense of efficiency if the driver is particularly more efficient in producing some frequencies than others. In effect, the housing may act as an equalizer.
     
  4. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,467
    Trophy Points:
    113
    Location:
    Winnipeg
    In a dynamic driver maybe just barely I can see it... but not to start MacGyver'ing the body when the effort should go to controlling the driver instead of compromising everything else. For a planar driver I don't see it.
     
  5. SSL

    SSL Friend

    Pyrate
    Joined:
    Nov 12, 2015
    Likes Received:
    1,239
    Trophy Points:
    93
    In an ideal world, the opposing forces of the drivers cancel out when coupled to the head.

    At the end of the day, soundwaves are an alternating compression and rarification of air (or whatever). I don't think a housing, even if made from a relatively "resonant" material, can be anywhere near as efficient at this as the driver. As mentioned previously, this would just reduce the efficiency of the primary driver. So, I have a hard time believing that engineering the housing in this way would be the intention of any headphone designer.
     
  6. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    It might have a bit more distortion, but saying that it doesn't have much control...
    I think the shape matters much more than the actual materials here. And the way the driver is coupled to the housing also probably matters more than the actual housing itself. And then again the earpads probably matter a whole lot more than that. I was just talking in terms of Utopia as a headphone vs HD800 as a headphone, both including their earpads.*
    That doesn't really make sense. No matter how big a driver is, for the same level at the same frequency the driver has to move the same amount of air. Sure, the smaller Utopia driver has to move more for the same level, but the overall force will still be the same.

    Uhh, saying that most speakers don't extend below 80Hz is a bit far-fetched. And I'm not talking about anechoic chamber measurements, but even then speakers can go as low as 50, 30 or even 20Hz flat. It's just a matter of efficiency because most bass drivers are too small.

    I just tried to say that the driver is much more efficient at moving air (which weighs almost nothing) than moving your skin or the housing. The different eartips also sort of decouple the housing from your skin.


    That looks sort of like a BMR driver or something like that, except with a rigid plate.
    You will absolutely need damping, especially with strong magnets or the whole thing will ring like a kitchen plate. Better yet, make it into a cone or dome shape to move the resonances upwards and try to keep the size small enough to push them above 20kHz. A flat plate, even a tiny one made from diamond will have horrible resonances in the audioband. See Fig 4 here: http://www.troelsgravesen.dk/Jenzen-D/DiamondStudy.pdf
    BTW: Even a big cone, if it is deep enough and made from a highly damped material (paper) can have treble without resonances and probably lower distortion than most tweeters (due to the huge size). Price to pay are cancellations and general phase issues in the top octave. Yes, that's a whizzer cone.

    Trying to design a headphone driver that will put out some bass without a seal will wreck your efficiency (or else it will be tipped up) and it'll still have to be very close to your ears or the whole bass vanishes from dipole losses. The K1000 needs speaker amps for a reason (and even then it has tons of bass rolloff).
    That's just because without an air cushion for the diaphragm the resonance will have an extremely high Q (and also higher in frequency). Same thing happens to all stats.

    EDIT: *The Utopia cups aren't even made of carbon fiber. And I'm not even that sure if carbon fiber is less resonant than some other matierials.
     
    Last edited: Mar 26, 2017
  7. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    Well, there were all those housing mods that made the T50RP cans listenable though:

    http://www.innerfidelity.com/conten...st-winners-best-measuring#3987tr4J7tSliPBt.97

    (obviously one can misshapen the sound just as well by fooling around with the assembly - not to mention the looks and the integrity of the final deal).

    I also know that adding and removing ports (i.e. little holes) to the housing affects low frequency response since this may have an effect on the pressure as the driver moves. I know that works for both planar and dynamic drivers (like my HD202). Seal the little port inside an HD202 and kiss goodbye to it's bass (becomes a ball-less HD201).

    There are limits to this, and I'm all for controlling the driver with digital/analog equalization once the options for assembly modifications become impractical or unattainable though.

    Better yet. Buy a headphone that needs little to no modifications and be happy.
     
    Last edited: Mar 20, 2017
  8. MF_Kitten

    MF_Kitten Banned per own request

    Banned
    Joined:
    Jan 1, 2016
    Likes Received:
    181
    Trophy Points:
    43
    basically what I want to do is to lower the mid-bass bump that the Fostex driver has, and instead let it do more sub bass. It seems to me that the mid-bass bump is caused by the driver being allowed to flop about too freely. The more you let it move undampened, the more it centres around the mid-bass. The more free it is near the middle, the more emphasis is put on this. So I'm thinking by allowing it to move MORE freely along the edges, but LESS freely near the middle, the energy will make the entire diaphragm move more, instead of the middle taking over most of the movement. If you can compensate for the loss of low end QUANTITY by simply filtering the high end on the ear-side of the driver, you might get something interesting.

    basically, it's the mid bass-to-sub bass ratio I'm interested in.
     
  9. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,467
    Trophy Points:
    113
    Location:
    Winnipeg
    Yes, controlling the acoustic response via the venting configuration of the chamber is one thing. I thought you meant designing the housing/baffle/etc to vibrate sympathetically to muck with the FR, which is all sorts of wonky.
     
  10. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,467
    Trophy Points:
    113
    Location:
    Winnipeg
    Nope. Just... nope. A "hump" in the diaphragm (utterly miniscule as it may be) does not create a hump in the frequency response.
     
  11. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    LOL! Nope, I don't think that would work all that well.

    One could sorbothanize, toilet-paper-treat, or dirty-socks-condition the cups to attempt to remove some frequency peaks or issues (not actively add). You proly know better than me, but these guys argue about the sound absorption properties vs. frequency of certain materials:

    http://www.paroc.com/knowhow/sound/sound-absorption

    From the materials stand point, which you know a bit or two, I bet different materials absorb acoustic sound at different frequencies more efficiently than other materials.

    Like I said, in some cases I would guess things may start to become impractical and unattainable.

    As far as active cups, the only time I saw something that sort of resembles active cups is this:

    http://www.innerfidelity.com/content/innerfidelity-august-update-0

    Check isobaric headphone mod, and even then I'm not sure how well it worked:

    Before:
    http://www.innerfidelity.com/images/KossKPH7.pdf

    After:
    http://www.innerfidelity.com/images/KossKPH7DIYModifiedObobskivich.pdf
     
    Last edited: Mar 20, 2017
  12. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    If serious about this, at some point one may benefit from picking up a decent book about headphone construction and theory of operation before making wild guesses.

    ... Or one could just try random stuff for science.
     
  13. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    This hypothesis is incorrect. There are actually headphones that don't change much in the bass with the seal broken. Some early Audezes were like this, but they were not consistent in this behavior.

    The reason this happens is because of the Q of the driver. Q is the damping factor, where a lower number means increased damping. Total damping = mechanic damping (surrounds, tension) + electrical damping (from magnet, charge). Some drivers are like the STAX Lambda don't have much inherent damping (high Q) and require a small volume of air for this damping. This is why the Lambda bass changes so much when measured without a good seal (midbass bump and massive rolloff) than with a good seal. The early Audezes didn't require much seal to extend flatly down to 30Hz (low Q). The Abyss are somewhere in the middle, where a poor seal results in a 50Hz emphasis with slight rolloff below, and a good seal results in flat bass to 20Hz.

    There are several factors at play:
    1. Change in damping factor when the seal is broken. The leaky air cushion from a seal provides damping, much how like automobile shocks do with oil and tiny holes / valves.
    2. Dipole effect: When seal is broken. the movement of the driver further away from the ear and the larger opening allow the back wave to cancel to the front wave, thus reducing bass below a certain frequency. Bass tends to be affected because of its longer wavelengths which makes out-of-phase cancellation more noticeable.
    3. Changes in the resonant frequency of a driver when the seal is broken.
    Conduction has nothing to do with it.
     
    Last edited: Mar 20, 2017
  14. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    Actually, that is a very good point. Not all planars behave the same indeed. One parameter I remember is the thickness of the diaphragm. I may be misremembering, but I think the thinner the diaphragm is, the more sensitive to seal it is. All probably related to the Q of the driver.

    When I measured some HiFiMan planars in San Diego I remember getting all things mucked up because I assumed seal was not essential for planars. Turned out, it did mater for the newer generation ones at the time. @Marvey saw similar behavior if I remember right.
     
  15. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Yes. We had a discussion on this. I was just as puzzled as you, until I realized that the Q of drivers is different. Therefore measurements on sealed plate are absolutely necessary to to get proper bass measurement. Some of the early Audezes seemed to have more damping. You can also argue that the HD800 and HD650 are fairly well damped too.
     
  16. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    BTW, adding this link to the discussion:

    http://www.head-fi.org/t/498292/my-diy-electrostatic-headphones/585#post_8738868

    It shows some of @Arnaud old simulations on electrostatics. My interpretation is that @ 100 Hz radiation is mainly along the center of the driver. Not so it seems @ 300 Hz where radiation happens in three locations of the driver. This may be a function of the wavelength and the size of the driver, and perhaps also the geometry of the driver, cups and pads (materials and so forth). At some point further along the volume things may become more cohesive. Or not.

    Things may become more complicated as the frequency increases. Dunno.

    Trying to make the whole driver move uniformly through out its geometry, and across the whole audio frequency range, would probably be counter productive and possibly in the un-obtainum range (based on what I'm seeing in those simulations and intuition). It may also be asking for a solution to a problem that does not exist.
     
    Last edited: Mar 20, 2017
  17. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    It's a problem that doesn't need solving. Restricting excursion near the middle of the driver is best done by either not playing as loud or increasing the size of the planar diagram. Dampening the center is likely to introduce compression of transients and increase distortion. May as well brace the baffle with different colored Plasticine, with red meaning bass and blue meaning treble; or run over a planar diaphragm with a tank, to create accordion like pleats.
     
  18. Bill-P

    Bill-P Level 42 Mad Wizard

    Pyrate
    Joined:
    Sep 25, 2015
    Likes Received:
    4,801
    Trophy Points:
    113
    Location:
    San Jose, CA
    The flat-ish surface by definition means it's a planar (well, "plane") diaphragm, though. Whether it's dynamic or not doesn't really matter much because I think all of these technologies are somewhat intertwined, hence the "correct" name IMO for "planar magnetic" headphones should be "orthodynamic". (on that note, I don't think "pleaded diaphragms" count as "planar" anymore, but I digress :p )

    And nah, I don't think that's the case. Newton's 3rd law works differently here. It's not that the diaphragm is "pushing away" on the housing such that the housing will go in the opposite direction. Note that the interaction here is sideways, not head-on. The reverse is true. The diaphragm is "pulling" the housing along with it.

    There may be some phase/timing issues involved with such a system, but I think it's more efficient than essentially "asking" the housing to be able to "resist" the pull of the diaphragm, in which case, the diaphragm itself may be strained enough such that it will eventually stretch. Think Audeze here. In fact, I think Audeze is a prime example of how "resistance is futile", and hence why I'm suggesting this model.

    Maybe not necessarily. See Marv's response below:

    Yeah, I'd agree the hypothesis that conduction is related to bass is incorrect in some cases. I have observed for myself how conduction is not necessarily related to bass response (not in the grand scheme of things at least). And that may not even be true for some IEMs, like say, I have observed that behavior in Andromeda.

    But I think conduction still causes some amount of energy to be transferred either way, be it at bass frequency or not. So I think sometimes breaking the seal may be preferable as it prevents extra conduction from needlessly altering our perception of a driver's response, and it won't necessarily alter bass response significantly. This is hard to explain, but perhaps I'll send over some modded/cut-out IEM tips for you to measure and test to see if they will have an effect on the frequency/distortion/spectral decay/spaghetti dimension domain, or if it's all in my head.

    It's just that @Serious insists that seal be there for bass to exist. And it's not like I haven't had the same discussion with him multiple times now. ;)

    Thanks for pointing out what I'm missing!
     
  19. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    I just insist that for good efficiency you will need a seal for good bass extension. Simple as that. All the non-sealed headphones are hugely inefficient (K1000, weird stat thing I heard, etc.). The K1000 is tuned for a high Q so it still has some bass.
    Sure, there's a difference between no seal and a perfect seal and the IEM eartip mods won't completely break the seal, but that's a different story. Over-ear headphones without earpads require high Q drivers with low Fs, which practically means low sensitivity (and thus high distortion) by definition. I'd be more concerened about phase issues from the high Q and distortion than weird conduction resonances with headphones.

    The weird estat thing that I heard probably had a super high Q resonance at maybe 50Hz or so (which would already be pretty impressive) and was extremely inefficient with a ton of diaphragm excursion to get bass. Distortion and resonances will be much worse from a non-sealed system.
    I'm not really suggesting that it's not possible to get bass from a non-sealed system, it's just that I think the seal is one of the major advantages of headphones compared to speakers, allowing fantastic bass extension and super low distortion, because you don't need a lot of excursion.

    Partially sealed systems are a different story and this seems to be what you did with your eartips. I think most open dynamic headphones fall into this category.

    That sounds a bit like the ADEL (or apex or whatever) technology to me, which rolls the subbass or infrasonics off to prevent pressure changes in the ear canal.
     
    Last edited: Mar 20, 2017
  20. Bill-P

    Bill-P Level 42 Mad Wizard

    Pyrate
    Joined:
    Sep 25, 2015
    Likes Received:
    4,801
    Trophy Points:
    113
    Location:
    San Jose, CA
    Well, if efficiency is not high enough...



    ^ oh, you totally shoulda seen that coming. :p

    Bass will be literally thunderous then and efficiency would be the least of your concerns.
     

Share This Page