Fostex T50RP Incremental Mods and Measurements

Discussion in 'Headphones (full-size)' started by bluemonkeyflyer, Jul 15, 2012.
  1. peter123
    The Monoprice M1060 seem to be pretty promising and might be worth reading up on. Haven't heard it myself though so take it with a grain of salt.....
  2. hemipowered007
    Ya I've thought about those before, but unless they're modding to a closed back, which ill look at, i dont have much interest. Im fairly certain my he4 would still suit my tastes more ( for open planar) having read up on them before people started modding those 1060. Cool though, now I have a new thread to read, spent last week re reading a whole lot of this thread, noticing how many times I repeated my own questions without knowing I had already asked. My memory is crap since having kids I swear.

    Edit: just saw you also said 560 which is a planar closed back. I have reading to do!
    Last edited: Aug 22, 2017
  3. hemipowered007
    The monolith thread is so long, and annoying, that I decided to just write monoprice and ask if I could just buy the drivers from the 1060. Those things are hideous and I would just end up wanting to build a new setup anyway. I realize this isn't a monolith thread, but it'd be fun to see how they do against a t50 driver when FULLY modded. The new headfi still has me wondering how to search within a thread. Anyhow, thanks for that headsuo
  4. hemipowered007
    So, wave guides, plus sealing the driver to baffle, plus other tweaks, I am so happy right now. The instrument separation has definitely improved. What I have now is a very neutral, open sounding, black background closed back perfection for me. I cannot find one single issue with the sound signature now, I just need to find better pads as I don't like the 840 pads. Will have to break down another 60 bucks and try the ori pads. I'm amazed what this driver can do, even without having a complete memory of their exact performance with the last set of mods(with stock cups/baffles), what I know is they NEVER sounded this good. And I know it's not source, because I had my audio gd10se and an sx1010 with these prior, as a few other amps. The decay is perfect, treble detail is there, not as much as he4 or my ex1000, but it's there, maybe more like an hd650 from the little I recall. The mids aren't forward, but placed in just the middle. I don't think anyone's been successful in making the t50 driver a lush mid signature though. Bass is perfect for me, it hits hard for as clean as it is. I really, really, dont know how closed cans could vastly improve over this. Sealing the driver off really improved the bass quality and punch, I'm honestly surprised. I punched holes in the felt opposite the treble reflector, which I added more to. Pictures below. All in all, I'm probably not done, but don't see what more I could squeeze out of the t50 driver without being a completely different signature of course. All in all its about enjoying the music, and that's what these are letting me do, not overanalyzing, not nitpicking, just really enjoying the sounds. 20170823_174628.jpg 20170823_174628.jpg 20170822_170606.jpg 20170823_175557.jpg
    Last edited: Aug 24, 2017
    GREQ likes this.
  5. rikk009
    Anyone know about corrugating the headphone drivers?
  6. Aphex72

    Is a good price the MKII version (new unit sealed) for 70 €?
  7. tommywantwingy
    Hey folks, forum newb here. I decided to model the RP driver in some software I use at work. It had been bugging me that some uber-damping mods here appear to preserve the high frequency character. I wondered in particular whether the ~10kHz peak is due to a resonance of the driver membrane itself.

    TLDR: I believe it is, and I think it can be damped in relative isolation with focused driver treatment.

    I modeled a 0.001" rectangle of the measured diaphragm dimensions. I set the poisson's ratio to that of copper (copper's is 0.36, and polyimide's is 0.34), left the default elastic modulus, and tuned the density to match the fundamental. Judging by the 20/40/50's influence on the mid bass hump I assumed the 100Hz response was a "cabinet" alignment resonance. The next response peak is at 1kHz, so I tuned for that. What I found was that there were modes with large mass participation ratios (i.e. they couple strongly with excitation) in most of the right frequencies...

    The model (note the circles in place of the magnet vents for reference):

    The first mode was typical shape (1kHz):
    The 5th mode (3.2kHz):
    and the 17th mode (9.8kHz):
    [​IMG] [​IMG]
    As a side note, one reason these couple well with input (the same reason they produce measurable difference in average sound pressure) is that there is an odd number of antinodes in the mode shape - 1 at 1kHz, 3 at 3.2kHz, 5 at 9.8kHz)

    In the harmonic analysis I applied a unit pressure load to the area covered by the magnet arrays:
    I set mode-specific critical damping ratios to match Tyll's measurements. It followed a logical progression of over damped 1st mode, near critical damping by the 5th mode, and tapering down to about 4% critical damping by 20kHz. I evaluated response magnitude (and phase, shown below) as the velocity averaged over the entire diaphragm area:

    I'd trust the measured data before this analysis, but the way the 3-4kHz hump and the 10kHz spike are so prominent without any funky damping values seem to support the notion that a driver mode is to thank for the 4kHz rise and to blame for the 10kHz zing (better word?).

    I ran a second harmonic analysis with the 9.8kHz mode damping value doubled, and extracted frequency response and operating shapes at the frequencies of interest. Original left, and targeted damping right:
    3.2kHz (note the nearly equal max velocity values):
    8.9kHz (note the nearly equal max velocity values even though the FR magnitude was boosted):
    9.9kHz (note the max velocity is about half of the original):

    I disassembled one driver and played an 8-10kHz sine sweep with the diaphragm setting on the baffle magnets. My headphones had a spike at 10kHz and a huge deadspot at 9.5kHz. I messed around holding a finger over different spots on the diaphragm with the sine playing. The outer hotspots indicated in the analysis mode shape gave the most noticeable attenuation at 10kHz, and an increase in response at 9.5kHz. This appeared to match the analysis result. Note that I did this on a pair with only one driver (my T40 parts beater) - since the half-magnet setup gives a really weak response it would be drowned out by the other side.
    Armed with this I chose the outboard magnet bars to apply small damping pads. It is an area where the 3.2 kHz operating shape shows velocity of about 20% of that at 9.9kHz.

    I give you the hockey tape mod. The tape measured 0.010" and filled about half of the stock gap between the diaphragm and the magnets:
    I put these two strips on the cup-side magnets, reassembled, and played the sweep again. 10kHz was panned about 60% away from the hockey tape, and 9.5kHz was panned about the same towards the hockey tape. After I did the same to the other side I played the sweep again, and my repeatability-challenged ears observed a linear response. I'd call it flat, but it might taper down into the top end. I listened to them for a while, and I'd judge the dynamic range to have expanded in the upper registers. The separation is just as good as it was before. I added the tape to the ear-side magnets as shown above on one side, and I think that was too much. I removed it, and I don't intend to adjust it any more. If I had to do it again I would use a hole punch to produce a repeatable shape.

    Happy listening.
  8. GREQ
  9. tommywantwingy
    I used Howies brand tape - which is described as a "tightly woven poly-cotton cloth and an aggressive natural rubber adhesive"

    Since it is bonded to the magnet I don't think permeability would make much of a difference with this. I would guess that surface roughness (hockey tape is good at that) would be key with the air oscillating across the surface rather than through it.
    GREQ likes this.
  10. Armaegis
    As a Canadian, I feel slightly offended...

    ... but as a Canadian, I got over it =P
    peter123 and GREQ like this.
  11. William Song
    Great work, but how do you open the drivers without breaking anything.
    I tried it yesterday, and I accidentally broke of the two smaller hooks on one side of the driver. I don't think I bricked it, as it is screwed down quite securely.
    Do you have any tips of opening the driver without damaging anything. I'm scared of doing it now after my mishap with them.
  12. tommywantwingy
    If you'll notice in my photo - all four of those posts on my test driver were gone. I couldn't arrive at a decent way to flex the hooks out of the way when disassembling. I would bet those serve as an assembly aid only since there is some gap left when the hooks are the only thing holding it together. I found the same thing you did: the screws collapse the gap nicely when gently tightened. It might also be some sort of sneaky warranty-void-if-tampered-with design feature.
  13. bluemonkeyflyer
    This video shows how to open the drivers. I made the video when conducting a failed modification experiment.

    The 4 snap tabs are easy to break so try this at your own guarantees! They will definitely break if you just try to pry them off. If you break the snap tabs, you can use dabs of epoxy to permanently secure the driver frames.

    ~ BMF
    rikk009 likes this.
  14. bluemonkeyflyer
    Very interesting and impressive work! I'll have to give this a try. Thanks for sharing.

    rendyG and tommywantwingy like this.
  15. MF Kitten
    Holy crap that's interesting! This is the kind of thing I wanna see more of! I'm kinda worried about disassembling the drivers, but that is exactly the kind of result I want!

    Is there a specific reason to the metal grille having those teeny holes, when the bar magnets are spaced that far apart? What if you made the holes bigger, or cut slots out instead?

    Also, running that simulation, can you see a way to reduce the mid bass specifically, without damping the entire driver? I put these drivers into DT990 housings, so they have a massive acoustic space to play around in. They can play deep bass well, but the mid bass hump is the same. I think the driver just REALLY wants to vibrate at that frequency.
    tommywantwingy likes this.

Share This Page