Modding Planars: Voodoo, Alchemy, and Science

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

  1. MF_Kitten

    MF_Kitten Banned per own request

    Banned
    Joined:
    Jan 1, 2016
    Likes Received:
    181
    Trophy Points:
    43
    As per usual, I had some bright ideas while relaxing on the couch. Imagine the classic Fostex Tx0RP driver. It's a square membrane, with a uniform magnetic field over it. However, the actual membrane doesn't actually move like a wall. That just isn't possible unless the actual membrane is rigid with a softer outer edge, like those square subwoofers people put in their cars.

    The membrane will have a longer excursion near the middle, which means it will bend and flex non-uniformly.

    So what if you were to dampen the driver to make the movement MORE uniform, by restricting airflow in the centre, and allowing more of it around the outer edge? Or if you were to take it even further, you can make it behave a bit more like a rigid round cone driver by dampening the corners as well, since the corners will have some rather irregular movement compared to the sides. A typical dynamic driver only has sides. There simply are no corners. And the best ones are the ones that move more uniformly and that physically distort and flex less.

    Of course, I'll try this out myself, but if anyone has specific ideas that could help me develop this, let me know! Right now most of the mods that people perform are rather simple, and treat the driver as a single factor, with very little thought put into what it's doing on the inside.
     
  2. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,461
    Trophy Points:
    113
    Location:
    Winnipeg
    You are not the first the ponder on this. There have been many configurations of damping schemes and many of them place extra things directly in the middle. Sure some of that is simply to absorb, but the ones directly placed on the driver will affect the airflow. The most extensive testing there was taping/gluing paper down over the centre square and putting holes into it with a pin

    There have also been waveguide type things built. I believe discussion of different sized holes for the waveguides depending where on x/y axis were brought up, but I don't think they ever made it into prototypes.

    My Speakers also has his pleated diaphragm thing and even has an animation to show how it's supposed to behave more linearly, although physics doesn't actually work like that, but I digress.

    Despite all the ruminations, the other question you have to ask yourself is whether the more "planar" desired action of the diaphragm is actually significant from the viewpoint of emanating soundwaves, and whether the mods you create to affect the air pressure voodoo will have negative consequences on the action of the diaphragm.
     
  3. spwath

    spwath Hijinks master cum laudle

    Pyrate BWC
    Joined:
    Dec 13, 2015
    Likes Received:
    7,894
    Trophy Points:
    113
    Location:
    Madison, WI
    I was thinking about that as well. However, unless there is some sort of surround, no matter what airflow stuff you have, the edges will always move less, because they are affixed in place. I wonder if one could make a planar driver with a surround...
     
  4. MF_Kitten

    MF_Kitten Banned per own request

    Banned
    Joined:
    Jan 1, 2016
    Likes Received:
    181
    Trophy Points:
    43
    the main point is that you can LIMIT certain types of movement, thereby FORCING other types of movement, eliminating certain resonances and modes caused by the curved square shape of the diaphragm. Simply limiting flex and distortion will really change how it sounds in practice.
     
  5. Bill-P

    Bill-P Level 42 Mad Wizard

    Pyrate
    Joined:
    Sep 25, 2015
    Likes Received:
    4,801
    Trophy Points:
    113
    Location:
    San Jose, CA
    Eh... the diaphragm shouldn't flex or excurse that far. That's the point.

    A good planar diaphragm would technically not move around much, if at all. The movement of the diaphragm should be almost invisible to the naked eye. At higher frequencies, excursion is very limited, and at lower frequencies, you'd ideally want the diaphragm to transmit most if not all of the movement to the housing, which will then transfer the energy to your ears. If the diaphragm moves a lot, then you know something is wrong. Audeze diaphragms after a while exhibit what you're describing, but other planar headphones don't (at least they don't if they're still functioning properly). Because the magnets on both sides of the diaphragm are supposed to prevent the diaphragm from moving too much in either direction. Hint hint. This is also why planars with one-sided magnets were thought to be a "cool idea" but then was ultimately abandoned for good ol' dual-sided magnets.

    And while damping a planar headphone, or actually, any headphone for that matter, you absolutely do NOT want to block the middle of the diaphragm with anything at all. Not only does it restrict airflow, it creates standing waves that will augment the waves generated by the driver (if those waves do not affect the diaphragm at lower frequencies where they start moving air, that is), thereby reducing linearity. You'll get weird upper mid and treble response if you do.

    Honestly, you can just try it yourself and you'll see what I mean.
     
  6. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,461
    Trophy Points:
    113
    Location:
    Winnipeg
    I know what you're getting at... but this is a bit of "putting the cart in front of the horse". The flexing/bowing of the diaphragm is the most natural behaviour of that mechanical system.

    If you're restricting airflow in some way, you are changing the atmospheric forces acting on that system. More critically, you are affecting those forces in a non-uniform manner across the diaphragm, which creates a whole host of nonlinearities in the behaviour. Standing waves are complex enough without adding a variable component. Furthermore, sound is essentially just air flow/wave propagation, and typically the less discombulated the better before it reaches our ears. Changing that air flow for the sake of altering diaphragm behaviour to produce unstable results which will then again go through that altered flow regime is a bit of a chicken vs egg type thing.
     
  7. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,461
    Trophy Points:
    113
    Location:
    Winnipeg
    Do you mean that at low frequencies the diaphragm doesn't move and transmits all the energy into the housing and it's the housing that generates sound? or conducts vibrations through to the skin? ... ?
     
  8. Bill-P

    Bill-P Level 42 Mad Wizard

    Pyrate
    Joined:
    Sep 25, 2015
    Likes Received:
    4,801
    Trophy Points:
    113
    Location:
    San Jose, CA
    No, I mean... ideally, at low frequencies, I think we want the diaphragm to not move much (if at all), and it should transmit most of the energy into the housing, which then transmits the vibration over to the ear pads, which then conduct it to our skin...

    Realistically, both of us know that that doesn't happen, and there will always be some amount of excursion, but I think we want to get as close to the ideal model as possible. Nobody wants a loose diaphragm flapping back and forth after all.
     
  9. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    @MF_Kitten, you can try stuff. But I don't think forcing the driver to displace uniformly is going to make things more linear in all cases necessarily.
     
    Last edited: Mar 19, 2017
  10. SSL

    SSL Friend

    Pyrate
    Joined:
    Nov 12, 2015
    Likes Received:
    1,239
    Trophy Points:
    93
    Do we? That doesn't make any sense to me at all.
     
  11. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,461
    Trophy Points:
    113
    Location:
    Winnipeg
    Bill, there are these things called the laws of physics...

    ... the ideal model is for the housing to be as stationary as possible for maximum efficiency and linearity in transforming an electrical signal into mechanical motion which in turn becomes acoustic energy. Ideally the diaphragm should be stiff/rigid as to present a uniform pistonic surface (barring large excursions and high frequencies, which is not relevant at the headphone and acoustic scale). No we don't want something flapping like grandpa's buttcheeks on taco night, but we still want that diaphragm to move and do the job that an acoustic transducer is designed to do.

    Where in the world did you get the idea that we want to transmit low frequency information via conduction? That works when you're tapping out morse code or sitting on a sub woofer to tickle your jollymaker, but soft tissue absolutely sucks for transmitting anything in the audible realm. Run a bi-amped headphone with proper tactile shakers that supplement the audible portion if that's what you're going for.
     
  12. Bill-P

    Bill-P Level 42 Mad Wizard

    Pyrate
    Joined:
    Sep 25, 2015
    Likes Received:
    4,801
    Trophy Points:
    113
    Location:
    San Jose, CA
    I think I should clarify. I actually wrote like... 5 paragraphs and then I decided to just post that "short" response, but I can see now I have left out quite a few things.

    Yes, I understand that the ideal model is for the housing to be as stationary as possible for maximum efficiency. This is what we want for dynamic headphones, with drivers that inherently move at every frequency no matter what you do to the housing. And yes, with dynamic headphones (and speakers), I would agree that having a completely still housing should be the goal. Focal had the right idea. They went with carbon fiber and their headphone has the right technicality to prove it.

    However, I think the situation for planar headphones (e-stats and orthos) is a bit different. The diaphragm is not free to move like with dynamic headphones. It is bound to and by the housing, and thereby, you can restrict the diaphragm's movement by making the housing too stiff/rigid. If making the housing super still was the goal, I'd think most manufacturers would be going for denser materials for their top-end planar headphones. Or say, in the case of e-stats, you would think that the manufacturer could go with a denser, heavier material for the housing since the whole construction already offset the weight. But why didn't they? Hell, why did Sennheiser use wood with HE90? Or aluminum with HE1? In fact, if aluminum was always on the table, why not use it with HD800, where it would make the most sense?

    I'm actually suggesting a design that I think the JPS Labs folks were going for, and that AKG was going for, and that somehow you're not seeing: the Abyss without seal, and/or the AKG K1000, where, because of the law of physics, a part of the housing becomes the acoustic transducer themselves.

    Perhaps my wording was confusing (or I had brain fart), but that was what I meant. ;)

    On that note, transmitting low frequency energy via conduction is probably the opposite of what I'm suggesting, though for some reason, it seems like that's what I implied from my previous post (yeah, I admit it reads like that, too). Even for in-ear headphones where that's the #1 way to get low frequency response out of the tiny drivers, I actually try to... impeach it by hacking away bits of the ear tips to reduce the conduction.

    So, in a nutshell, I'm suggesting the opposite of what OP is saying: instead of dampening the middle, undampen the sides.
     
  13. SSL

    SSL Friend

    Pyrate
    Joined:
    Nov 12, 2015
    Likes Received:
    1,239
    Trophy Points:
    93
    Nobody wants a couple bricks strapped to their head.
     
  14. n3rdling

    n3rdling Friend

    Pyrate
    Joined:
    Sep 26, 2015
    Likes Received:
    929
    Trophy Points:
    93
    Location:
    Los Angeles
  15. Elnrik

    Elnrik Super Friendly

    Pyrate
    Joined:
    Jan 10, 2017
    Likes Received:
    8,973
    Trophy Points:
    113
    Location:
    Denver CO
    Home Page:
  16. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    Props to @Bill-P for not loosing his mind at the suggestion of the laws of physics.

    However, I think the laws of physics reference applies. The housing IMO is not the one that translates electrical power to mechanical power which then results in air pressure changes. It is the transducer. And that applies to dynamics, planars, and electrostatic headphones.

    The housing, together with the driver and the ear enclose a volume which will be compressed by the driver. All of the materials enclosing the volume react differently to different frequencies.

    Anyhow, the reason I don't feel uniform displacement will necessarily result in linear response (assuming one could readily do that anyway), is that some areas of the diaphragm may have more difficulty than others in moving at a certain frequency at a certain amplitude. If pushed, they will proly start to break down and vibrate at other frequencies and inject non-linear distortion. There are, from what I read, all kinds of trade offs on the stiffness, elasticity, geometry and so forth of a driver. One could possibly pick up a book or a few journal articles about it and get a bit more informed about it. Assuming that forcing a driver to move all around at the same frequency, in-phase, and at the same amplitude is optimal is probably assuming too much.

    I remember @Arnaud posted some findings in simulating how the diaphragm of an electrostatic transducer moves as a function of frequency. It was not uniform from what I remember. But I don't think it necessarily has to. As long as the diaphragm moves at the same frequencies that the electronics "command", and the resulting air pressure is roughly at the right amplitudes and phase, relative to other frequencies, life is good. There may be differences in air pressure (slight or not) at different locations inside the volume. One would have to do a simulation and measurements to figure out how well behaved things are. The density of the air volume, the geometry of the headphone, and materials would probably play a non-negligible part of the results.

    In other words, I think things are more complex than just making sure the driver moves uniformly all over. And it some cases, forcing this my be detrimental as one may be forcing the driver into non-linear behavior. Say, maybe the edges of a particular driver may not "like" to displace a lot of volume at certain frequencies, meaning it will break down and inject all kinds of non-linearities.
     
    Last edited: Mar 20, 2017
  17. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    That makes it even more confusing to me. Without a seal the diaphragm will just move much more. Most if not all planars to me seem to be highly damped with super low Fs for flatline bass extension. I don't really think you can compare the K1000 to the Abyss with no seal.

    I also don't think the Utopia's housing, especially the earpads, are designed for maximum stiffness. At low frequencies the housing seemed to vibrate much more than the HD800s. The driver is also tuned for much more excursion.
    Then what are you suggesting? Bass rolloff?

    Breaking the seal isn't really comparable to simply reducing the conduction. It really just reduces the bass, conduction or not.
    I also don't think conduction is the #1 way to get bass out of the drivers. They're not "tiny" for the volume they have to drive. And I don't think there's a lot of vibration with IEMs, especially not for those with eartips. CIEMs might be a bit of a different story, but I also don't think the drivers can move the housings much if at all.

    That just makes it more loose all around. And I don't think it'd be a good idea to have a rigid planar diaphragm. That would probably result in resonances all over the place.
     
    Last edited: Mar 20, 2017
  18. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,461
    Trophy Points:
    113
    Location:
    Winnipeg
    Dynamic transducers are not "free to move"; they are bound by the surround/spider/whatever that restore them to place in the absence of signal. The cone is stiff but will suffer it's own set of resonances, both radial and transverse. In an ideal case the cone will have no such breakups/resonances, but of course the real world is not ideal. The stiffness of the cone or cone damping/doping will reduce the magnitude of the distortion products within the cone itself. The housing still matters because again... physics. You have Newton's 3rd law, so you still want some mass to hold the driver in place. Then have housing stiffness and damping to reduce the unwanted body resonances which will affect the driver's location in space. The chamber damping is different from housing damping obviously, all the fluffy stuff affects the acoustic wave rather than body/vibration/resonances.

    Yes a planar diaphragm is bound by the housing, but your concept of this housing "restricting" the movement is completely backwards. We *want* the housing to be non-moving so all the energy is transferred to the diaphragm so the diaphragm moves as much and as cleanly as possible. The housing/magnet assembly/etc are already orders of magnitude heavier than the diaphragm, so this is mostly the case. True, a heavier and even stiffer less-resonant housing would be better, but then weight/comfort and cost of the headphone becomes a factor. Metal is great for weight and stiffness, but it sucks for damping out unwanted vibrational energy that does make it in. A planar diaphragm still has tension which provides restorative force. A planar diaphragm is not "stiff", but it approximates this with the traces and diaphragm tension. However, the diaphragm will still suffer transverse distortion products (mostly transverse, but highly more complex depending on the trace pattern).

    I still don't see the point you're trying to make with the Abyss or K1000, unless you mean the housing is acting as a sort of waveguide... but that's not the same as being part of the transducer.
     
  19. Bill-P

    Bill-P Level 42 Mad Wizard

    Pyrate
    Joined:
    Sep 25, 2015
    Likes Received:
    4,801
    Trophy Points:
    113
    Location:
    San Jose, CA
    You were saying?
    [​IMG]

    Sorry, couldn't help it. :p

    But honestly, it has happened before. It's not like I'm suggesting something new and innovative. I'm not that smart.

    Haha, was I supposed to lose my mind? :p

    But seriously, the housing itself still experiences the force from the transducer. Newton's third law, as Milos pointed out.

    The thing is that I think the "housing should be solid, super dense, and should be perfectly still" situation is almost like "resistance is futile" if you know what I mean. At some point, the housing will still have to vibrate. It's just to what extent that will happen. For cheaper headphones with plastic enclosures, the housing vibrates a hell of a fruggin' lot whenever there's any sound.

    Abyss has magnets only to one side. So whether you break the seal or not, it's already not having much control at higher excursion, because the magnetic field is not constant throughout the whole range of excursion.

    Carbon fiber at the same density is actually much stiffer than aluminum or steel. HD800 is like some composite plastic IIRC.

    The only reason why Utopia's housing vibrates more is again, due to Newton's 3rd law. The driver has more excursion and is being pushed much harder than the HD800's drivers, so it ends up pulling the housing along as well.

    Dammit, now you guys are forcing me to be the one to reference physics' laws. :p

    On that note, though, I think ear pads stiffness is the other object of debate that I think we can put off for another time.

    Not necessarily but... lemme remind you that most speakers only reproduce as low as 80Hz, and though they still respond lower than that, I don't think the rest is "audible" per se.

    Mm, breaking seal reduces contact, which reduces conduction, essentially. Put your keyboard on a piece of foam, which has lower density than your keyboard's solid plastic/aluminum/whatever housing, and type like normal, or actually, hammer at the keys. You probably will feel less vibration on the desk, and it'll be less loud, but you're exerting about the same amount of force on each key anywho. I think this will prove that breaking the seal does impeach conduction.

    But that's the physical explanation of it. Probably less fun now that it's over-used. :p

    Yes, some bass roll-off will probably occur depending on how "strong" the force exerted by the diaphragm is, or wait, scratch that, my bad. Since I should be like... extremely politically correct in my wording in this situation, allow me to (hopefully) correctly state that it depends on how strong the force exerted ON the diaphragm, which causes the diaphragm to accelerate and cause a force from itself, which is what will be exerted on the housing joint/thing due to Newton's third law. Happy? :p

    Okay, joking aside, bass roll-off probably won't matter because as you know, flat bass ain't everything.

    Hoho, you'll be surprised. Whether IEMs have bass or not depend entirely on seal, and as proven above, seal is the key to conduction.

    Honestly, I think you're underestimating just how much "force" these things are pushing.

    Finally, here's a visualization of what I'm talking about, because it seems I'm having a hard time describing it succinctly enough. ;)

    [​IMG]

    ^ give that configuration some extreme magnets and I think we'll see interdasting stuffs. And if it has extremely powerful magnet to offset the efficiency, then I'd wager that the housing itself will become almost "transducer-like" at lower frequencies where the vibration is inevitable. Basically, the flat/rigid surface will probably be pretty significant in mass, so accelerating it causes it to exert a much greater force on the housing than if we were using some thin, nearly massless (nanometer-thickness!!!!!! :p ) diaphragm. That much greater force, I'd guess, will be pretty darn hard to dampen so instead, we could just let it be in free air. So no ear pads or anything. Probably no need for dust filter in the front of the diaphragm in this case if it's much more rigid than the usual material.

    ^ edit: by the way, I know that's gonna create some insane amount of bass with conduction ;)
     
    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
    No, I guess I should have said that the housing in those cases "feels the effect of the transducer (as in the transducer acts upon the housing) and is somewhat forced to follow the same path, albeit not exactly the same due to time lag and due to the housing being much more rigid than a thin plastic membrane". So yeah, you're right. Not really "as a transducer" per se, but I think the housing would at least "feel" the effect.

    In practice, I have put my fingers on the Abyss' housing while it's playing back lower frequencies (bass hits), and yeah, the entire housing, especially around the ear cups, vibrates a lot whenever bass hits, and without seal. Putting it on my head does help reduce those vibrations significantly, but still not to the point where they disappear completely.

    And honestly, I just don't like the way the Abyss sounds when it's on my head. It sounds much more compressed to me with that use case.

    K1000 when I heard it seemed to do the same thing, but I didn't put my finger on the housing to test my hypothesis.
     

Share This Page