Fostex T50RP Incremental Mods and Measurements

Discussion in 'Headphones (full-size)' started by bluemonkeyflyer, Jul 15, 2012.
  1. tommywantwingy
    The plates with the holes act as magnetic field guides (a "pole piece" in a loudspeaker driver) to concentrate the field where the diaphragm coil sits (and increase the sensitivity). The field lines attracted by the guides follow the material around the holes within the guide material to exit the guide sides and get back to the magnet. If you want to start carving at the holes I'd suggest maintaining the wall thickness between the holes - I imagine excessive field concentration (i.e. closer to saturation) makes the guides less effective. I don't imagine an acoustic benefit to be had by modifying this aspect of the driver, although deburring or rounding the edges of the (probably) stamped holes might make me feel better about what little flow is there. Make sure you scrape all the metal bits stuck off the magnets afterward.

    Below the first natural frequency (everywhere <1kHz in my approach) the operating shape is basically the same as the first mode shape (see above). If you wanted to damp 100ish Hz by damping the center of the center magnet you'd be damping potentially everything up to around 2kHz (and other modes - 3.2kHz included - that show antinodes in the center). That sounds like an interesting experiment.

    Today I still maintain my expectation that the 100Hz rise is a mode arising from the interaction of the diaphragm "Vas", and the mass, stiffness, and terminal impedance on both the cup and pad sides of the diaphragm. The diaphragm itself appears to resonate beginning at 1kHz, and the extra mass and lower compliance of the "stuff" around it would generate the lower frequency responses. My model doesn't incorporate that at the moment, but if someone wants to suggest an equivalent mechanical system of springs and masses I could couple that with the diaphragm and see what happens.
     
  2. William Song
    I put it together last night and it's had some pretty drastic effects
    I have a side with no mods, complete stock with 840 pads, and a side with some changes. (Plasticine in baffle wells, and strips of sound dampening foam on the back It's called response super sound deadening foam, with 1.5mm butyl and 3.5mm foam)
    Without changing the inside of the driver, everything basically the same on both sides, with a little better high end extension on the modded side.
    After the tape inside the driver, it sounded like the soundstage just started to push itself a little wider. The frequencies are much smoother than stock, very linear, except for a spike at ~7khz which immediately died down at 7.4khz
    It seems to be working well on my drivers, it tamed ~10khz frequencies
     
  3. MF Kitten
    I have to see some measurements of this now!

    I wonder if there's a way to target the 1k peak too... I swear, I am gunna ruin these drivers, but I refuse to not try this out! :p
     
    tommywantwingy likes this.
  4. MF Kitten
    I see the 9.5khz nodes are in three long ridges. Is it safe to assume those are aligned with the magnets?

    Also, does anyone have the exact distance from magnet to membrane, so I know how much I can apply to the magnets?
     
  5. tommywantwingy
    I measure 0.018+/-0.002" between the cup-side magnets and the clamp surface, and the diaphragm assembly contributes another 0.003" of clearance to that. I'd call it a solid 20 mils in assembly, so about half of that gap remains with the tape on one side.
     
  6. William Song
    Another thing I noted is that, DO NOT overtighten the screws like I did
    I thing I may have given it a little too much and the entirety of the thread and the plastic came off
    The screws of the driver are on one side connected to the 'cover' with the white backing paper, the other side are 8 protruding holes which align the diaphragm as well as house the screws.
    If you overtighten it, the whole plastic bit comes off, so be firm yet gentle.
    I don't know what's been said about this but just be weary
     
    tommywantwingy likes this.
  7. tommywantwingy
    The center two node lines in the 9.8kHz mode are right at the magnet holes (you can distinguish the circles in the mesh image), and the outer two nodes are just outboard of the next row of holes (i.e. at the inner edge of the outer magnets).
     
  8. hemipowered007
    Good to see some new ideas, I'm really happy with the extensive mods I did earlier this year, but I'm really interested in this tape mod. I really don't want to open mine again! Maybe I need to buy another pair and start over.
     
  9. rikk009
    That's a 472MB file. What is the length of the video? Can you downsize/re-encode it? I am on limited data plan.
     
  10. MF Kitten
    I'm curious, tommywantwingy... What do you think happens to the frequency response if I were to cover the entire length of every bar magnet with that tape? I might have a fetish for damping all resonances and reflections :p
     
  11. tommywantwingy
    That is really difficult to predict based on the mode-superposition analysis I showed. The modal and operating shape results only tell half of the story since the phase varies over the diaphragm in operation at intermediate frequencies. Differently phased components may constructively or destructively interfere to alter the averaged magnitude response - this can be seen with the 8.8kHz operating shape comparison above. If I had to guess I might suggest that it would darken the signature and reduce the sensitivity, but my guesses are generally unreliable - that's the reason for the software.

    Lucky for you I've had a full-harmonic (i.e. not mode-superposition) model simmering this week with 32 localized dampers, so I might be able to give a reasonable prediction of what would happen soon. First things first, though - I must tune it to reproduce the general FR baseline.
     
    Last edited: Oct 26, 2017
  12. MF Kitten
    That's amazing, you're a legend for that!

    I wonder if there's a mod one could apply to remove those sibilant "planar wall" ringing frequencies you see in CSD graphs of planar measurements.

    If you think covering all three magnets would darken the sound, I really wonder in what way... Maybe we could potentially get a better type of darkening than the simpler felt over the driver!
     
  13. nick n
    One thing to consider. Many tape adhesives over time with elevated temperatures= either from long term wearing or ambient temperatures= will get viscous/gooey and (possibly )weep all over the place resulting in a sticky mess.
    Also be prepared to deal with sticky residue on parts should you want to remove that later.
    Try to keep anything with glue type adhesive away from critical areas.
    Unless you actually test them for temperature resistance you will not know how they will react long term.
     
    Last edited: Oct 26, 2017
    bluemonkeyflyer likes this.
  14. Bubblejuice
    Does anyone know if there is a map or diagram of all the parts of the T50rp and their exact measurements?

    I want to make some 3D models for cups for 3D printing, but I'm new to it, so it'd be hard to just come up with measurements on my own.
     
  15. William Song
    Mr Speaker has released his Alpha dogs, which can be found on the thread
    https://www.head-fi.org/threads/open-alpha-t50-3d-printed-headphone-project-from-mrspeakers.825868/

    You can get the dimensions for the baffle from the parts, which from what I've heard was used to make the Mad Dogs (Pros?)
    And then you can base your measurements for driver, screws, cups from those.

    There's a few parts on thingiverse, but I don't think they'll really help in designing new parts.
    Other than that, I don't think there's much else.
     

Share This Page