Search This Blog

Tuesday, March 12, 2013

On the modification of Monoprice 8320 & 9927 [Updated]

[Updated on 3/20/13: Final version has been rolled out, with a cyanoacrylate coat applied]

As previously analyzed, Monoprice 8320 & 9927 are formidable contenders for the throne in the budget-fi war of IEMs. Although their electroacoustic performance are excellent, two problems need to be addressed:

1. Short bandwidth
2. A quarter-wavelength resonance at 5 kHz

While MEElec's large triflange definitely mediates the issue, not everyone can fit such a huge sleeve in their ear canals. Should internal modifications be able to take care of the issue, There is no doubt that Monoprice 8320 & 9927 are true kings among the others.



What you need to know beforehand

First of all, it is best to choose 9927, and stay away from 8320. Not only 8320's frequency response bandwidth is slightly shorter than 9927, its quality control is mediocre at best: I bought two pairs, and they all turn out to be defective. Moreover, while 9927 comes with a nice package, 8320 comes in a plastic bag with only a single pair of sleeves. 9927 is a better value, period.


Opening 8320 & 9927 is easy as a cake, since the two-part housings are loosely joined together with adhesive. Simply stick a razor blade at where the strain relief is located, and gently wiggle your way through, all the way around the housing. On the other hand, as the rim of the diaphragm is glued directly to the housing, the driver itself is virtually unextractable.

8320 & 9927 are well-vented IEMs, and there are total of 3 vents that can be tweaked with: One on the front, and two on the back of the driver. By default, the rear vent at the center is clogged with glue. While the front vent low-passes the sub-bass range, the rear vent high-passes the same region. In other words, two vents cancel themselves out acoustically, while yielding a good amount of depressurization for better transient characteristics. 
Since all of the vents do not directly affect the problematic frequency response range, such modification does not improve the sound quality of the IEMs. Changing the diameter of the bore helps, but rather than damping the peaks, it works like a shelving filter in the high frequency range. Interestingly, a polymer + agar coat (Or any type of cyanoacylate-based adhesive) does exactly what's needed by adding a mass to the moving parts. As the inner loss increases, peaks are effectively settled & shifted downwards.


1. Irreversible method: For intermediate users only.
Materials needed: A low-viscosity polymer, a syringe with long nozzle for in-nozzle injection


Make sure your low-viscous polymer is made of a material that adheres to a plastic film very easily. Over-application of polymer will clog the front vent mesh. If it gets wet, the mesh must be removed, and replaced with an electrical tape with a pinhole punctured.


Once successfully modified, Monoprice 9927 shall have a tonal balance equivalent to that of IEMs at higher price range, possibly surpassing HiFiMAN RE-400.


2. Irreversible method: For advanced users only.
Materials needed: A low-viscosity polymer, a syringe with long nozzle, a needle

This modification technique is same as the above, but with an extreme twist. Once the modification is completed, make a pinhole-sized ventilation hole on the diaphragm itself. This will definitely mess up the harmonic distortion figure, but will greatly damp the overall resonances.


As an additional ventilation kicks in, the reflection between the driver and the backplate @ 250 kHz becomes more emphasized, and  the harmonic distortion figure reaches up to 2% @ 1 kHz at 100 dB SPL output due to the overdrive. 


Please keep in mind that this technique is very risky as it may permanently damage the diaphragm itself, and must be done in exactly same manner on both units in order to avoid any chance of interchannel deviation. 

29 comments:

  1. You work on this is just amazing. Thank you. Basically, the mod is syringing a thin coat of polymer on the filter at the end of the nozzle. Am I correct? How do I find this type of polymer? What are your impressions of the sound quality after this mod?
    It's unfortunate that you have had issues with both pairs of 8320s. I have 4 pairs and they all work fine. I guess it could be batch related. Thanks again. Excellent work.

    ReplyDelete
    Replies
    1. Thnx! Any type of non-water based polymer(as hard as natural latex) should do it. However, the polymer has to be quite well diluted, as the entire sheet of diaphragm is to be coated only with a small amount of polymer, no more than few drops.

      Delete
  2. I am a bit confused about what the polymer is being applied to, from "adding a mass to the moving parts" I am guessing agar is spread over the transducer membrane, which is meant to stiffen the membrane, or perhaps a glob of glue sticking to the transducer membrane adds inertia and loads down the transducer to reduce the resonance?

    ReplyDelete
    Replies
    1. A small amount of low viscous polymer spreads the agar on to the entire sheet of 9927's diaphragm with a few micro meter thickness. A glob of glue is not ideal, as it can interfere the uniformity of the cone's motion.

      Delete
    2. I see, the most common household "low viscosity polymer" I can think of is actually clear nail polish, or maybe instant adhesive? I'll try a bit of nail polish cut with volatile cyclopentasiloxane on some plastic wrap and report back.

      I am sure we can get creative by dissolving different resins in rubbing alcohol, but I'd love to know what solvent you have used for the agar.
      A quick search showed that agar isn't soluble in alcohol, and I suppose a water based solution would take a long time to dry, perhaps agar was first saturated with water then cut with the polymer?

      Delete
    3. You know what, I just went with your nail polish idea. Voila, the effect is even great than my solution. Thnx for the inspiration, I'll update this data ASAP with the nail polish coat! Gotta give you some credit too.

      Delete
  3. Crazy how flat the FR is with the coating on the Diaphragm. You wouldn't expect this in such an inexpensive IEM. Nice work here..

    ReplyDelete
  4. I just want to make sure I clearly understand Method #1. In short, you squirt Krazy Glue through the mesh and let it fall on the driver membrane? The picture looks at though you put the syringe on the earbud side and squirt it through the mesh. If that is the case, what needs is there to open then IEM in the first place? Additionally, how would the mesh not get wet if you effectively squirting right through it?

    I plan on doing a recable and this mod, I just want to do it correctly.

    ReplyDelete
    Replies
    1. Since I completely cut off the nozzle for replacement, I never bothered to use the above syringe, or opened the housing to apply the nail polish. Repeated this couple of times in order to achieve optimum damping: applied several drops with a plastic rod, and swooshed the unit around for even application while not trying to get the mesh wet.

      FYI, the hardest part is to get both units done equally, and I lost two IEMs due to mismatch. Good luck.

      Delete
    2. So how do you thin the nail polish to apply it properly and are you just coating it with the applicator that comes in the bottle? How many coats. What's the technique for replacing the noozles...any tips...lol(pun sorta intended)? Any further thoughts on accomplishing this mod. I have a couple of 9277's on the way. Would you recommend Meelec triple flanges. I can fit large dual flanges pretty easily. I have fairly massive earplug foam sleeved modded dual flange on my RE 400's. This is just to damn tempting not to try!

      Delete
    3. The one I modded is now in Austria, LOL

      You only need to apply a few drops(3 times coating), and the adhesive will disperse by itself.

      And as far as replacing the nozzle goes, I cut off the nozzle using a sharp razorblade, and replaced it with a plastic cylinder that is wider/longer than the original. Of course, the opening must be widened up to match the diameter of the bore too. (I recommend using a precision handpiece drill)

      Delete
  5. Rin

    I have Meelec triple flanges on my 9927's. I tweak the tips by putting foam earplug cores in each flange and then cut back the very tip so that there just
    a 1.5 mm lip holding the smallest foam core in place. I used a similar approach with my RE 400's on their double flange. I use a handkercheif filter replacing
    the stock one and then I put in a foam damper into the noozle it's about 4mm long and 2mm wide. It's a very open soft foam. Yanked it out of some foamed silicone tpeos stock tips. I was using a foam I had that was very similar to what is in the RE 400. This foam controls sibilance much better. Everything is clearer and more evenly balanced with a hint more warmth and liquidity. No loss in air, detail, or space. Bass has a bit more snap and weight. Just wondering if I should still cut off the nozzle and replace with a wider bore?
    Will their be any benefit. Not sure if cyanoacrylate will benefit them at this point but am willing to do it if it might be worthwhile. They sound quite amazing right now with only 15 hours of Jlabs burn in from that site.
    Right now Inks is borrowing my RE 400's so I can't do direct comparisons but these in present form are clobberng the 400's in every way.

    James(jgwtriode)

    ReplyDelete
    Replies
    1. Oh wow, that's great James. More damping + Meelec triflanges should solve most of the issues. I won't recommend replacing the nozzle, since the improvement you can expect from the mod isn't really that much. On the other hand, cyanoacrylate coating provides extra stiffness to the driver, which is great IMHO. Just be super careful not to get the front vent mesh wet- that will end the life of the IEM.

      Delete
  6. Rin:

    Okay I made some extensions from plastic straws cut narrow and rolled to insert in as a guide to drip cyanoacrylate down into the nozzles. May not have got all the drops in since their may be some side residue in straws. I hear a bit more detail a touch less resonance and a bit more even character with more bass snap. Is that about what it should do? Vocals particularly female vocals I notice have more focus and presence than just prior to putting in the glue. I don't hear any imbalances.

    Just out of curiosity what would changing the nozzles to a larger bore size actually do?

    ReplyDelete
    Replies
    1. As long as there's no channel imbalance, you're good. The effect of increasing the bore size is shown on the above graph. Check out the "Thick nozzle" plot.

      Delete
  7. Thanks Rin!


    Could you also improve high frequency response ang bandwidth by building a low order resonant circuit to boost the high frequencies a bit in level? I know they make filters for loudspeakers Baffle Step Compensation I believe. Someone mention a 68ohm resistor and a 0.33uf cap. Would you wire that in series and wouldn't that be just a basic first order filter? It's been a while since I studied electronics. Would that be worthwhile in your opinion, either with or without the wider bore nozzles?

    Thanks for your feedback

    James

    ReplyDelete
    Replies
    1. A simple RC filter would yield a boost starting from 7 kHz. However, I don't 'openly' endorse the idea of integrating passive components to headphones- Not only it is lossy, but also it beats the purpose of acoustic optimization. Yes, headphones are MP systems, but frankly, I don't give a dang :P

      Delete

  8. Hey Rin ok...then!

    Would you say that with the wide bore and meelecs the high frequencies are still a bit rolled off or are they about right for this type of application. BTW do you know of any triple flanges that have a 3/16" bore?

    Thanks,

    James

    ReplyDelete
    Replies
    1. As long as the bore diameter remains same, there's no treble loss. What we can expect from Meelec's long stem is the quarter wavelength resonance shift so that the peaks are more evenly distributed along the spectrum.

      Once the nozzle diameter is enlarged, the sleeve's bore size must be widened up in order to keep the treble level intact.

      Delete
  9. Okay I appreciate that. The Meelec's measure 1/8" aproximately 3.5mm at the tip.

    I have some lostear clones from Ebay coming that are 4.5mm at the tip. That should give me the other 1/16". Then I will try doing the wider bore and tell you what I hear since I can't measure it. If I want to do the needle width hole
    in the diaphragm, do I poke the hole center nozzle...it doesn't appear it can be done actual center of the diaphragm since the nozzle looks offcenter?

    Thanks again Rin,

    James

    ReplyDelete
    Replies
    1. No problem! I made it in the center when I cut off the nozzle stem and widened up the opening. Please lemme know how everything turned out, can't wait..

      Delete
  10. Rin:

    Got my lostear tips from ebay. It suck they are only 3.5mm just like the melecs. They are about 1mm ir 1/16" to narrow on bore width. They do however sound better than the Melec's despite being essentially the same design. The silicone looks slightly more opaque and is softer. I get a better seal and everything is more organic and smoother but subtley more detailed and the bass is sanppier and more tuneful. Soundstage is a bit better and images are hint more focused. This is the best sound I have ever heard from earbuds. Still looking for a 4.5mm bore width on eartips for wider nozzles. May just do the wider nozzles, punch as pin hole in the diaphragm and stretch these to fit until I can find triple flanges with a wider bore. I have tried Hifiman double flanges that would be the right size but they don't sound as musical or as detailed as the triples. Will do the final mods in next few days and let you know.

    Thanks for all your efforts,

    James

    ReplyDelete
    Replies
    1. Try to grab a pair of Yamaha EPH-100 stock tips- they are super wide! Good luck to you, my friend.

      Delete
  11. Actually just found another alternative. Sennheiser triple flanges for IE 8 are quite large bore. Yalokin makes some and there is an even cheaper british version available in black and clear on ebay. Sweet.

    Im' just out of money until end of month. I have figured out an easier way to enlarge the bore size without having to completely cut off or drill the nozzles as much. Might not even need glue to hold it together. Cut down a ball point pen tip cap. I found some that are 1/4" with a 1/8" inch hole that can be drilled out to 3/16". Then cut the noozle off at the eartip groove. Bore out the remaining nozzle to 3/16". Now cut the tip cap which has been bored out and sanded smooth to a uniform 1/4". Make a locking sleeve out of a 1/4" plastic flexi straw. Slide that over the remain nozzle on the earbud. Put a little cyanoacrylate gel around top of remaining nozzle. Make sure the is just enough overlap of the straw sleeve to hold the pen cap tip in place. Now let glue dry and remove or leave on. I am then going to cut silicone tip core out of other tips and slide over assembly to make a stop for the core of the eartip...to prevent it slide all the way down the nozzle. No filter holder now but I just use a bit of glue around the top edge and then trim filter as
    necessary. Slide IE 8 clone triflange tips on. Oh yeah and of course punch a pin hole in diaphragm before I lengthen the nozzle. No alignment, no repositioning no excess messy glueing. The straw maintains alignment and locks the pen tip cap in place till the cyanoacrlate is dry.

    Just an idea to simplify the process.

    Will do this at the end of the month when I have some funds again.

    Thanks,

    James

    ReplyDelete
    Replies
    1. Way to go, my friend! I only provided the basics, and the rest of micro-detail is up to the readers haha.. IMHO as long as both channels sound same after the mod, you are good.

      Delete
  12. Hey Rin:

    Had a chance to play with microfibered 400's and the 600 with the Tour Inks has sponsored. The microfibered 400's are at least as good as the 600's. Waiting for a bit more burn in on the 600's to make a final determination. I want to get some of this same material for the monoprices and see if it can do the same for those.
    Would you mind providing what type of microfiber cloth or brand it is to make finding one a bit easier?

    Thanks, James

    ReplyDelete
    Replies
    1. The microfibered RE-400 vs RE-600 challenge is for fun! haha..

      Oh, simply search for the term, "light weight microfiber fabric" or "heavy weight non-woven fiber" on ebay/Amazon! Any of em should do it.

      Delete
  13. I Would like to buy a Pair of your "Fixed" Monoprice 9927's if your selling them on ebay.

    ReplyDelete
  14. Rin, I love your site!
    Thank you very much for taking the time and put so much work into this site :)
    I have a question about Monoprice problem #2 "A quarter-wavelength resonance at 5 kHz"
    I would like to understand the reason for this resonance to occur. Which part of the earphone is responsible for this to happen and why?
    thanks for any help to understand the issue!

    ReplyDelete