[Bilavideo]

THEY'RE HERE!

The drivers have arrived! I'm so excited! My daughter helped me open up the package and we both were blown away by how truly tiny these devices are. I took some pictures, which I will share here: http://www.head-fi.org/forums/album.php?albumid=643. I haven't wired them up yet, but for fun I have squeezed them into a pair of foamies. I can't wait to wire them up and get some initial impressions of how they sound in the raw.

For the record, I'm trying not to get my hopes up too high, as these drivers have no crossover or filtering. If it turns out that they sound a bit harsh, that will be okay. I've ordered $36 worth of hearing-aid parts from Soniq, a hearing-aid supply company, way more stuff than I'll use here. I've got all kinds of acoustic tubing and other doo-dads on the way. It may be a couple of days before I'm able to add those refinements to the mix. For now, though, my next step is to get these bad boys wired for sound.

 

[ezzieyguywuf]

I thought you said you would post pictures as soon as you got them

 

[Bilavideo]

Quote:

Originally Posted by ezzieyguywuf /img/forum/go_quote.gif I thought you said you would post pictures as soon as you got them

Ask and ye shall receive. Head-Fi: Covering Headphones, Earphones and Portable Audio - Bilavideo's Album: DIY IEMs

 

[pyrokid]

wow, this is cool. I was just wondering why nobody had tried this on the car ride home, and I come back and see this.
Hope you do well. It would be really cool if you made something that sounded like a UE11 at this price.

 

[Billyk]

Are you planning on molding them into a solid unit somehow? Don't quit know how to visualize it though. Really like where it's going, those itty bitty drivers are very cool.

 

[Bilavideo]

Quote:

Originally Posted by Billyk /img/forum/go_quote.gif Are you planning on molding them into a solid unit somehow? Don't quit know how to visualize it though. Really like where it's going, those itty bitty drivers are very cool.

The neat thing about this is that there are any number of options. Tonight, I sat down with the CI-22955's and wired them up. I picked them because they're the bigger of the two and my soldering skills are less than spectacular. I also thought it best to leave the more expensive drivers for last. It was probably best that I did. Even after my adventures with the SOIC-8's, just soldering the CI-22955's was a tutorial. These little guys took me to school.

I just wanted to test them, first to make sure they weren't defective and then to see if they were even usable on their own, without crossovers or mechanical filters. I pulled the plastic center out of a couple of medium foamies and pushed the drivers through. As tends to happen in the middle of a project, I discovered that the jack I was using was crap, so I ended up carefully wiring these little guys to the speaker leads on my home theater. I didn't want to blow either my hearing or these drivers, but when I cranked the volume up, ever so carefully, I got proof that (1) these drivers work; (2) the wiring would hold for now; and (3) even without filters or a crossover, these drivers hold their own.

Because of their small size, these drivers can live inside an earplug. Among other things, that means they can be positioned more directly in contact with the eardrum. I've ordered $36 worth of acoustic tubing only to discover, to my delight, that I may not need any of it. A driver that resides in its own earplug is a driver that produces sound that doesn't sound like it has come out of a tube. As I suspected, the amount of bass produced by such a driver is impressive. As I listened to The Who's Pinball Wizard, I was impressed by the soundstage of this work-in-progress. To be fair, Pinball Wizard is a pretty airy track, but the clarity of my bare drivers, sitting inside cushions, was a model of transparency.

They're not pretty yet but to hear them, you'd never know each driver had only cost $26.

Tomorrow, when I have more light - and am well rested - I'm going to solder the TWFK's to see if they're in good working order, what they sound like on their own, and perhaps what they sound like together with the CI-22955. I can envision three, maybe four, driver configurations:

Bilavideo 1 - Single Driver - CI-22955
Bilavideo 2 - Dual Driver - TWFK
Bilavideo 3 - Triple Driver combo of CI-2295 and TWFK
Bilavideo 4 - Quad Driver combo of 2 TWFK's
Bilavideo 5 - Five Driver combo of 2 TWFK's and 1 CI-22955
Bilavideo 6 - Six Driver combo of 3 TWFK's
Bilavideo 7 - Seven Driver combo of 3 TWFK's and 1 CI-22955

This can go on to the point of absurdity. The "more drivers is better" impulse has to be balanced against another reality: More drivers take up more space. If I can get a single CI-22955 into an eartip, I can get it and as much as two TWFK's (or as many as 4 TWFK's without the CI-22955) before I've run out of space. After that, I'd have to do what everyone is doing now - put the drivers into a shell and substitute tubes for the intimacy of a direct-ear connection.

At this point, I'm already sold on the direct-ear connection. I think it's the future. It's more immediate and less likely to involve unnecessary resonance. When I listen to music, I'd much rather hear the driver than the tube connecting the driver. But that's me. Now that drivers have miniaturized to this level, there's no reason to keep following the model of the hearing aid, especially when that model was invented decades ago, when hearing aid drivers were enormous. As far as I'm concerned, less is more.

 

[saintalfonzo]

This is quite the undertaking and I look forward to seeing the results! This is becoming a huge subscription thread for good reason. Good luck; may we be buying your design in the future.

 

[Bilavideo]

Thanks, saintalfonzo!

 

[gilency]

One of the most amazing threads in head-fi right now. Subscribed! can't wait for the final product. how much will it sell for??

 

[Bilavideo]

Quote:

Originally Posted by gilency /img/forum/go_quote.gif One of the most amazing threads in head-fi right now. Subscribed! can't wait for the final product. how much will it sell for??

Thanks for the support! I'm hoping this thread will encourage creative minds to take control of what they put into their ears. The quality of today's IEMs is greater than ever, but there's still a huge gap in the information and options available to consumers of IEMs. In some ways, the world of IEMs is where the audio world was decades ago. Today, if you want to buy a loudspeaker, manufacturers routinely provide the speaker's power handling, voice coil diameter, impedence, frequency range, frequency response, magnet weight, SPL, Vas, Qms, Qes, Qts, Xmax and mounting depth. But if you buy an IEM, they'll tell you how many drivers there are, but not which ones. They'll tell you they use an integrated crossover but they won't tell you what parts they used or where it's set. In the absence of facts, we get wine-taster mumbo jumbo. Mysticism and rumor reign supreme.

But that can change. When it does, it will become more common for interested consumers to build their own units. "Custom" won't mean the shell fits your ear canal. It will mean that you didn't just get the system you wanted; you took an active role in deciding what that system looks and sounds like. It won't be a battle of the brands; it'll be an array of choices. People will come together, not so much to brag about their brand so much as to exchange ideas on what configurations best meet what needs.

As I explore this world for myself, I think it would be neat to see what choices could be made on the basis of certain drivers. I wondered, going in, whether you could design a single-driver system around the CI-22955, which is one of several drivers used by several big manufacturers in their triple-driver models. I was pleasantly surprised at how well the CI-22955 performed in that driver-in-the eartip design. Already, I would put my Bilavideo 1 (single driver in-the-eartip design) ahead of par with the $139 PFE, an IEM which many rave about as performing well beyond its price range. My Bilavideo 1 was so clear, so spacious and so bass rich, I think it would make many people happy. But it's a simple design. With a reasonable markup, I can see it selling for $75 to $100.

The biggest pain, so far, is soldering these little guys. They're so tiny, they're hard to work with, even with a hands free and a magnifying glass. I thought SOIC-8's were a pain! Today, after ruining one driver and almost ruining a second, I came to realize how something as simple as wire, if it's the wrong gauge, can pull these little guys apart.

But experimentation will give way to a community of DIY'ers, sharing ideas and pushing things to the next level. That community will still have people making products for one another, but the costs will be more reasonable. You'll be paying for someone's time, expertise and ingenuity, not for a corporate pyramid, advertising and a business model where you try to sell the cheapest crap you can for the highest possible price.

 

[kostalex]

 

[saintalfonzo]

Take your time, I'm sure your knowledge and perseverance will pay off. I will be anxious to see what you will come up with, as will many others. Just don't rush it.

 

[Bilavideo]

UPDATE

I got an email from Tom Hazlett at Knowles Acoustics. In response to my questions about IEM crossover, Tom generously provided me with several documents discussing design options, including crossover. It was a step in the right direction. The following are some suggestions from Knowles:

1. Back volume venting increases low-frequency amplitude.

2. The use of inductor coils, in small form factors, causes "saturation effects" and adds distortions to the signal path. Instead, resistors are used to attenuate HF while caps are used to attenuate LF.

3. HF filters typically involve two resistors and a cap. R1 and C1 form a bridge between the positive and negative wires to create a voltage divider. R2 is set between the bridge and the positive line of the source.

4. There's a formula for working out the specifics. {(R1 + Xc1)\\Driver}/{R2 + (R1 + Xc1)\\Driver} with Xc1 + 1/(2 x Pl X C1)

5. If R2 is equal to or greater than 0 ohms, there's no impact on the output vs. frequency response of the transducers (because the circuit provides "pure impedence compensation of the BA transducer(s) connected in parallel." Only the impedence vs. frequency response would be affected.

6. If R1 is equal to or greater than 0 ohms, the circuit becomes a low pass filter. The woofers is simply rolled off at high frequencies. The corner frequency is defined as Fc = 1/(2Pi x C1 X the driver). The impedence is limited to the value of R2; the second driver will be shorted and not function.

7. A low-frequency filter can be "capped" to shape the LF response of the woofer. The corner frequency is defined as (R1 + Xc). The attenuation of this path is capped by the value of R1 for high frequencies. With Xc1 = 1/(2 x Pi x f x C1), corner frequency can be calculated.

8. The magnitude of mid/high attenuation defines the peak ratio in the bass response and is a function of R2/R1.

9. Because the external ear is bypassed and the ear canal blocked, the natural response of an adult ear canal is about 3kHz. Earphones have to compensate for "the loss of canal resonance," making it difficult and undesirable to target a perfectly flat response.

10. Earphone testing is problematic because it requires the use of an acoustic coupler (to simulate the seal of an eartip in the ear canal). The choice of couplers affects frequency response. Test results above 9 kHz are often inaccurate and unreliable.

11. IEMs are more efficient than loudspeakers. Less than 1 mW is needed to achieve rated sound output; less than 3mW is needed for maximum output. Earbuds, on the other hand, may require 15mW or more to achieve maximum output.

12. Dampers (a/k/a acoustic resistors) attenuate HF and HF frequency peaks.

12. Because impedence rises with frequency, adding a resistor reduces LF but it also improves LF headroom.

13. "Earphone exit tubes are commonly sized at 2.2 mm ID to accept Knowles BF damper plugs." A larger ID maximizes HF.

14. The length of the exit tube affects location and amplitude of HF peaks. An increase moves the 2nd and 4rd peak to the left (lower in frequency) as well as its amplitude. A decrease moves them to the right (higher in frequency) and reduces their amplitude. It's a tradeoff between optimal bandwith and smoothing the frequency response.

15. ED series drivers typically need damping to smooth HF. Dampers are positioned at the end of the exit tube. BF plugs are available with various resistance values to smooth resonance peaks. TWFK drivers have extended HF response and "may benefit from smoothing of high frequency peaks." With no damper, there's a spike between 600 and 800 Hz, from 112 to 121 dB. A 330 ohm damper reduces this spike to 115 dB and sets its occurrence at between 500 and 600 Hz. A 680 ohm damper virtually levels it.

16. TWFK drivers have enough HF response that no increase in the exit tube diameter is desirable.

17. The recommended cap for the TWFK is .82uF. A lower value (such as .47uF) slightly increases bass (by about 1 dB) helps fill in the recessed mids between 400 Hz and 1kHz, but dramatically increases the HF spike at 1.5kHz (by about 3 dB). At the same time, it lowers HF output from about 4.5kHz and up. A smaller cap (1.5uF) has virtually no affect on bass but adds to the 1.5kHz spike (by 1-2 dB) and moves the 7kHz valley to 2.5 kHz. On the other hand, it provides the most ample HF response from 3kHz onward.

 

[Bilavideo]

IN MEMORIAM

Funeral services will be held today for two drivers who were killed in unrelated soldering accidents. CI-22955, $26, and TWFK, $52, both received fatal wounds after head-on collisions with Bill's bad soldering. Investigators on the scene have listed contributing causes to be the underwhelming magnification on Bill's hands-free and the enormity of Bil''s hot tip (Ladies, be warned). When the driver terminals slipped off, snapping tiny leads, the CI-22955 was rushed to surgery where its torso was opened up with a flat-head screwdriver for what turned into an impromptu anatomy lesson. Despite repeated attempts to revive him, CI-22955 died on the table. No attempt was made to save the even smaller TWFK, which frankly had it coming.

CI-22955 and TWFK are survived by their twins which, while meeting separately, married and live happily together within the left foamy tip of Bill's prototype. Their success together, as possibly the world's smallest three-way unit (the tip is the shell) can best be expressed in the maxim: If this tip's a rockin' don't come a knockin'.

In the aftermath of this terrible tragedy, Bill has ordered replacement drivers and has purchased a better soldering station and a larger magnification for his hands-free. Friends, don't let friends squint and solder.

 

[elliot42]

An unfortunate accident... they shall be remembered.