[FYJX]

Hey, can't wait till you get that together. GOOD LUCK! and HAVE FUN! Hope i can get a listen to these haha.

^^^ i want to see the innards of the TF10 =) and the SE530

 

[pdupiano]

Wait a sec, Bilavideo, didn't you tell us that there are in fact no real tweeters when it comes to iems? So how did you come to the conclusion that the caps in multi-driver units are more for protection rather than crossover features?

Additionally I wonder if the custom silicone mold in the UE super fi 5 is specifically designed for a particular angle. If I remember correctly, when comparing the SE530 and the W3' as well as the UM3x, each canal had a different angle and length. I think we can conclude that the length of the tube matters (as well as the tube material used) but I'd also guess that the angle of the drivers may impact the overall sound.

 

[meraj.salek]

Pdupiano, I've been thinking along the same lines...
I thought all these drivers were designed to be hooked up directly. The levels these are to operate at is no where near that of loudspeaker tweeters...hmmm

Also, wouldn't driver mis-match and interaction have an effect on the sound? While the drivers are probably too close to the eardrum for the effects to be noticed, isn't it still possible?!

 

[Bilavideo]

Quote:

Originally Posted by pdupiano /img/forum/go_quote.gif Wait a sec, Bilavideo, didn't you tell us that there are in fact no real tweeters when it comes to iems? So how did you come to the conclusion that the caps in multi-driver units are more for protection rather than crossover features?

I misspoke.

First, I need to clarify what I mean by "tweeters." When I said there are no "real tweeters" out there, I meant that all the drivers - at least from KA - are "wide-range drivers." There are no drivers that only output HF. Therefore, there's no need for a "protective" crossover. You're not going to blow the driver six ways from Sunday if it's designed to reproduce a "wide range" of frequencies. And, as meraj.selak has pointed out, we're talking about extremely small amounts of power, nothing like the hundreds of watts that pump through a typical woofer.

That said, while KA doesn't make any drivers that are only tweeters, it does make a few drivers with "extended HF": The FK and the TWFK (which is a tweeter/woofer FK). In marketing the FK, KA says, "The wide bandwidth version pushes the boundaries of high-frequency performance. As part of the TWFK receiver, the WBFK offers the widest bandwith response of any receiver." If you compare the specs on the FK http://www.knowles.com/search/prods_...-23451-000.pdf and the TWFK http://www.knowles.com/search/prods_...-30017-000.pdf, the TWFK has better bass than the FK, but only by a few dB (97 vs. 94 dB at 200 Hz).

Which one produces the best HF extension depends on what you want. The FK spikes at 2.5 kHz where the TWFK's spike is closer to 3 kHz. What's more, that first FK spike hits 108 dB whereas the TWFK's first spike peaks at 103 dB. So, do you want the higher HF or the louder one closer to the midrange? Both slough to a trough after that. For the FK, it's to 95 dB at 4.5 kHz while the TWFK troughs to 93 dB a little closer to 5 kHz. Both hit a second and final peak at 6 kHz, but the FK hits 100 dB whereas the TWFK hits 97. According to the graphs, both drivers are still outputing around 75dB at 10 kHz

So, there are at least two KA drivers that have "HF extension" worth calling a tweeter (The TWFK has genuine "sizzle") but because they're both wide-range drivers, the crossover must not be there for protective reasons So, it's there for the sound, right?

This is where I got my wires crossed. In loudspeaker design, crossover networks have high-pass filters for the tweeters, low-pass filters for the woofers, and band-pass filters (establishing a floor and ceiling frequency) for the midrange (assuming it's a three-way crossover). In a basic First Order Butterworth, caps are used for the high-pass and inductors are used for the low-pass. When I contacted KA about this, their rep said inductors aren't used in earphone crossovers because of the hum. It's obviously not an issue in loudspeaker designs, where the woofer is sitting across the room and the woofer's output is so much greater, but in an earphone, nobody wants to deal with the distortion. Fair enough, but without inductors, I wondered how manufacturers would create a decent low-pass. KA spoke of using caps and resistors. Resistors increase impedence, using lowering volume across the spectrum. You could use a resistor to balance impedence and reign in a wild driver from drowning out its neighbor, but I still didn't see how a resistor would create a low-pass filter.

Westone and UE (on the UE10 and UE11) seem to be doing a better job than most. In so many designs, it appears that the "crossover" is only be applying to the "tweeter." This had me scratching my head until it hit me that caps are typical high-pass filters. There wasn't any magic here. The cap was doing what caps generally do in crossover designs: cutting out the bass. Sure enough, the specs on the TWFK came from a driver that was tested in tandem with a .82mF cap. But that confused me all the more since the TWFK is supposed to be a "dual" driver with a woofer and tweeter. Why would you want to cap a woofer? Wouldn't that be counterproductive? In the meantime, both "tweeters" and drivers used as "woofers" are mechanically filtered with cloth filters of varying ohms of resistance, raising questions about the effectiveness of a crossover that has to be assisted by a cloth filter.

One possible reason to cap a "tweeter" is to send more power to the "woofer." Electrical "flow" is a bit like water "flow" in the sense that blocking a path over here can redirect the flow somewhere else. This is why developers get hit with exactions, forcing them to compensate for the extra "water runoff" when they throw down large amounts of asphalt and concrete. LF signals blocked from going to the "tweeter" end up diverted to the "woofer." When the "woofer" is really a wide-range driver with a sock shoved in its mouth, the more power you can give it, the better.

If you look at the performance specs for the two most obvious "woofers" - the CI-22955 http://www.knowles.com/search/prods_...-22955-000.pdf and the SR Series http://www.knowles.com/search/prods_...438NWS-000.pdf - neither looks much like a woofer. The CI's first peak is at 900 Hz while the SR peaks at 2.5 kHz (Its peak of 120dB is 6 dB higher than its deepest bass). But with a big enough sock in its mouth, each driver has potential. At 100 Hz, the SR hits 112 dB. Not bad. At 200 Hz, the CI hits 114 dB. If you're pulling 110 dB or better through 500 Hz, just imagine putting both of these in the same earphone. There might be a need to use resistors to adjust the output balance between the drivers, but the result could be a very euphonic blend.

Given the limited use of the crossover to sculp the sound, I assumed that its real purpose was protective. As I look at it now, its real function IS to tailor the sound, but in an indirect way. With a dual driver like the TWFK, it seems insane to filter out the LF since the "dual" driver's woofer would need it. But a capped tweeter will send more power to the other driver. If that other driver happens to be the "woofer" of the two, then the cap will act (in some ways) like a resistor, diverting LF power to the other driver. It will make the "woofer" louder, at least with respect to the LF, which is an indirect way of achieving a result normally involving an inductor attached to the woofer.

How much of this is real and how much of it is just for show is anybody's guess. I'll have to experiment some more with the drivers to know for myself.

Quote:

Additionally I wonder if the custom silicone mold in the UE super fi 5 is specifically designed for a particular angle. If I remember correctly, when comparing the SE530 and the W3' as well as the UM3x, each canal had a different angle and length. I think we can conclude that the length of the tube matters (as well as the tube material used) but I'd also guess that the angle of the drivers may impact the overall sound.

I was looking at that, too. The silicone looks like a holster aiming the two drivers' outlets at the same sound port. The reason for this is functional. As I discovered when piggybacking the CI with a TWFK, there's a gap between the divers' outlets. What UE has done here reminds me of a bartender pouring two bottles into the same container. To do that sort of thing, it helps to go at it from 45-degree angles.

 

[meraj.salek]

So is it correct to assume that a 2-way IEM is basically 2 same drivers, 1 with a cap in series to cut out bass frequencies and the other with a filter in front to dampen the high frequencies?

What variable is tweaked as the driver count goes up? Different filters?

Trying to wrap my head around this

 

[FYJX]

Quote:

Originally Posted by meraj.salek /img/forum/go_quote.gif So is it correct to assume that a 2-way IEM is basically 2 same drivers, 1 with a cap in series to cut out bass frequencies and the other with a filter in front to dampen the high frequencies? What variable is tweaked as the driver count goes up? Different filters? Trying to wrap my head around this

For Ultimate Ears, basically that is the case. However, for some other company's like Westone and Klipsch, there is actually a crossover which splits the frequency range rather then dampen the signal to make each driver a dedicated freq driver.

 

[Bilavideo]

Quote:

Originally Posted by meraj.salek /img/forum/go_quote.gif So is it correct to assume that a 2-way IEM is basically 2 same drivers, 1 with a cap in series to cut out bass frequencies and the other with a filter in front to dampen the high frequencies? What variable is tweaked as the driver count goes up? Different filters? Trying to wrap my head around this

In a single-driver setup, the only real difference between offerings is the choice of wide-range drivers and the cloth filter used to attenuate or smooth a wild HF. Within the frequency range of a driver, there's a center point, a center frequency. We seem to sense when it skews high or low. Filtering is used to attenuate runaway HF. Resistors are used to deal with bloated LF. According to KA, spiking impedence cuts bass and adds headroom.

Single-driver designs can sound amazingly good, but you're not going to get that last bit of sizzle from a single driver. The CI series, for example, sounded pretty nice when I popped a couple of those drivers into an earpiece (a poor man's IEM). The CI has a lot of output so, with the right fit, you get a very dynamic sound. I did the same thing with the BK-26824, just because it was the cheapest BA I could find ($38/pair) and I wanted to see how it would perform. I was surprised at the quality, which struck me as middier but still quite nice for the price. But when I compared either driver to the TWFK, there was a definite sizzle gap. Properly filtered, even the cheapest BA will give you a nice, balanced, sound - but in a single-driver design, you'll miss the top end.

To get a two-way design, you can either pop in two single drivers or one dual driver. The easiest answer is the $52 TWFK, which is +5 at 200 Hz and 5+5.5 at 5.5-6.5 kHz. A person could easily just go with the TWFK (Pop it right into a tip and be done), but with two caveats: First, although this earphone is fairly flat for much of its run, there's a sizable spike aound 3 kHz, a spike of +11. To my ears, this shifted the center frequency forward, making me want to filter it, which would alter its HF extension (It was tested undamped). The second caveat is that it was capped during testing (.85uF). This would have cut the bass, maybe to highlight the tweeter side. When I tried it unfiltered and uncapped, it had a surprising amount of bass but was a bit harsh on the high end. When I used the white filter KA recommends, it smoothed the performance. I think a person could go with this, uncapped, and have the most basic of duals by popping it into a tip. But you'd still want $3 worth of white filters (You can get a package of 10 from Microsonic for $13).

Another way to run a two-way is with two separate drivers. Which two is up to you. Here are some possibilities:

1. You could take two identical drivers and simply cap one and filter the other. The capped driver would be your tweeter. The filtered driver would be your woofer.

2. You could match a CI or an SR with an FK or WBFK. You can get a pair of SRs for $28, a pair of CIs for $52. Both will need a strong filter to HF spikes which are louder than their bass (I recommend the reds). Depending on the FK, you could get different degrees of HF extension.

 

[Bilavideo]

LITTLE SHOP OF TWEETERS

DON'T FREAK OUT - THIS CHART IS A WORK IN PROGRESS
I'M COMPARING FK/WBFK OFFERINGS AND THIS IS THE EASIEST WAY TO LAY OUT THE CHOICES

Let me state my preferences up front. Here's what I look for in a dedicated tweeter:

1) HF Extension: How far does this driver go before it peters out? Even if the lower treble is more important than the top end, wider extension adds presence and sizzle.
2) Frequency Centerpoint: Ideally, I'd like that first spike to be as far from the midrange as possible. I can get midrange from another driver. All things being equal, an early spike portends weaker HF extension.
3) A Milder Spike: It's easy to place a higher value on a stronger spike because it represents louder highs, but a spike is an imbalance. A little goes a long way. I want HF representation, not distorted highs that go through my head like a sharp needle.

Given the value and coolness factor of the TWFK, any FK or WBFK driver worth considering needs to at least have the same quality of HF extension as the TWFK dual, which has the following data:
http://www.knowles.com/search/prods_...-30017-000.pdf
HF Extension: 73dB at 10kHz
1st Peak: 2.9kHz (103dB)
2nd Peak: 6kHz (97dB)

FK-23451 ($38 - Digikey/$40 - Mouser nonstocked) (630 ohm)(107dB)
http://www.knowles.com/search/prods_...-23451-000.pdf
HF Extension: 80dB at 9kHz
1st Peak - 2.5kHz (108dB)
2nd Peak - 6kHz (100dB)

FK-23989 ($45.54 - Mouser, min. 12 nonstocked)(145 ohm)(95dB)
http://www.knowles.com/search/prods_...-23989-000.pdf
HF Extension: 70dB at 9kHz
1st Peak - 2.7kHz (102dB)
2nd Peak - 6.3kHz (87dB)

FK-26260 ($40.19 - Mouser, min. 13 nonstocked)
http://www.knowles.com/search/prods_...-26260-000.pdf
HF Extension: 80dB at 9kHz
1st Peak - 2.5 kHz (105dB)
2nd Peak -5.7 kHz (97dB)

FK-6260 ($33/Digikey, min. 28, non-stock) (135 ohm)
no datasheet

FK-23466($37 - Digikey, min. 25, non-stock/$45.54 - Mouser, min. 12 nonstocked) (630 ohm)(107dB)
http://www.knowles.com/search/prods_...-23466-000.pdf
HF Extension: 80dB at 9kHz
1st Peak - 2.5kHz (108dB)
2nd Peak - 6kHz (100dB)

FK-26433 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26433-000.pdf
HF Extension: 80dB at just shy of 9kHz
1st Peak - 2.5kHz (107dB)
2nd Peak - 5.7kHz (97dB)

FK-26746 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26746-000.pdf
HF Extension: 70dB at 9kHz
1st Peak - 2.7kHz (103dB)
2nd Peak - 6.4 kHz (87dB)

FK-26768 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26768-000.pdf
HF Extension: 77dB at 10kHz
1st Peak - 2.5kHz (109dB)
2nd Peak -6.4kHz (97dB)

FK-26777 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26777-000.pdf
HF Extension: 80dB at 8kHz
1st Peak - 2.4kHz (108dB)
2nd Peak - 6kHz (92dB)

FK-26816 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26816-000.pdf
HF Extension: 70dB at 9kHz
1st Peak - 2.7kHz (100dB)
2nd Peak - 5.5kHz (85dB)

FK-26837 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26837-000.pdf
HF Extension: 70dB at 9kHz
1st Peak - 2.7kHz (103dB)
2nd Peak -6.4kHz (87dB)

FK-26888 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26888-000.pdf
HF Extension: 70dB at 8kHz
1st Peak - 2.9kHz (101dB)
2nd Peak - 6kHz (88dB)

FK-26889 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26889-000.pdf
HF Extension: 70dB at 8kHz
1st Peak - 2.8kHz (101dB)
2nd Peak -6kHz (88dB)

FK-30020 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-30020-000.pdf
HF Extension: 70dB at 9kHz
1st Peak - 2.7kHz (102dB)
2nd Peak -6.2kHz (87dB)

FK-30030 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-30030-000.pdf
HF Extension: 70dB at 9kHz
1st Peak - 2.7kHz (103dB)
2nd Peak -6.2kHz (87dB)

FK-30040 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-30040-000.pdf
HF Extension: 70dB at 8kHz
1st Peak - 2.8kHz (99dB)
2nd Peak -6.5kHz (81dB)

WBFK-23990 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-23990-000.pdf
HF Extension: 77dB at 10kHz
1st Peak - 4kHz (104dB)
2nd Peak -7kHz (97dB)

WBFK-30000 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-30000-000.pdf
HF Extension: 73dB at 10kHz
1st Peak - 3.3kHz (103dB)
2nd Peak -6.4kHz (97dB)

WBFK-30019 ($40.19 - Mouser)
http://www.knowles.com/search/prods_...-30019-000.pdf
HF Extension: 77dB at 10kHz
1st Peak - 4.1kHz (103dB)
2nd Peak -7.1kHz (96dB)

WBFK-30042 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-30042-000.pdf
HF Extension: 73dB at 10kHz
1st Peak - 3.4kHz (104dB)
2nd Peak -6.6kHz (95dB)

WBFK-30095 ($40.19 - Mouser)
http://www.knowles.com/search/prods_...-30095-000.pdf
HF Extension: 80dB at 8kHz
1st Peak - 3.4kHz (106dB)
2nd Peak -5.4kHz (103dB)

 

[Bilavideo]

[size=medium]TOP TWEETERS[/size]

BEST EXTENSION

1. WBFK-30019 ($40.19 - Mouser)
http://www.knowles.com/search/prods_...-30019-000.pdf
HF Extension: 77dB at 10kHz
1st Peak - 4.1kHz (103dB)
2nd Peak -7.1kHz (96dB)

1. WBFK-23990 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-23990-000.pdf
HF Extension: 77dB at 10kHz
1st Peak - 4kHz (104dB)
2nd Peak -7kHz (97dB)

1. FK-26768 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26768-000.pdf
HF Extension: 77dB at 10kHz
1st Peak - 2.5kHz (109dB)
2nd Peak -6.4kHz (97dB)[/COLOR][/B]

4. WBFK-30000 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-30000-000.pdf
HF Extension: 73dB at 10kHz
1st Peak - 3.3kHz (103dB)
2nd Peak -6.4kHz (97dB)

5. WBFK-30042 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-30042-000.pdf
HF Extension: 73dB at 10kHz
1st Peak - 3.4kHz (104dB)
2nd Peak -6.6kHz (95dB)

BEST HF CENTERPOINT

1. WBFK-30019 ($40.19 - Mouser)
http://www.knowles.com/search/prods_...-30019-000.pdf
HF Extension: 77dB at 10kHz
1st Peak - 4.1kHz (103dB)
2nd Peak -7.1kHz (96dB)

2. WBFK-23990 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-23990-000.pdf
HF Extension: 77dB at 10kHz
1st Peak - 4kHz (104dB)
2nd Peak -7kHz (97dB)

3. WBFK-30042 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-30042-000.pdf
HF Extension: 73dB at 10kHz
1st Peak - 3.4kHz (104dB)
2nd Peak -6.6kHz (95dB)

4. WBFK-30095 ($40.19 - Mouser)
http://www.knowles.com/search/prods_...-30095-000.pdf
HF Extension: 80dB at 8kHz
1st Peak - 3.4kHz (106dB)
2nd Peak -5.4kHz (103dB)

5. WBFK-30000 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-30000-000.pdf
HF Extension: 73dB at 10kHz
1st Peak - 3.3kHz (103dB)
2nd Peak -6.4kHz (97dB)

GENTLEST SPIKE

1. FK-26816 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26816-000.pdf
HF Extension: 70dB at 9kHz
1st Peak - 2.7kHz (100dB)
2nd Peak - 5.5kHz (85dB)

2. FK-26888 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26888-000.pdf
HF Extension: 70dB at 8kHz
1st Peak - 2.9kHz (101dB)
2nd Peak - 6kHz (88dB)

3. FK-26889 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26889-000.pdf
HF Extension: 70dB at 8kHz
1st Peak - 2.8kHz (101dB)
2nd Peak -6kHz (88dB)

4. FK-30040 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-30040-000.pdf
HF Extension: 70dB at 8kHz
1st Peak - 2.8kHz (99dB)
2nd Peak -6.5kHz (81dB)

5. FK-30020 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-30020-000.pdf
HF Extension: 70dB at 9kHz
1st Peak - 2.7kHz (102dB)
2nd Peak -6.2kHz (87dB)

6. FK-26746 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26746-000.pdf
HF Extension: 70dB at 9kHz
1st Peak - 2.7kHz (103dB)
2nd Peak - 6.4 kHz (87dB)

7. FK-26837 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-26837-000.pdf
HF Extension: 70dB at 9kHz
1st Peak - 2.7kHz (103dB)
2nd Peak -6.4kHz (87dB)

8. WBFK-30000 ($45.54 - Mouser, min. 12 nonstocked)
http://www.knowles.com/search/prods_...-30000-000.pdf
HF Extension: 73dB at 10kHz
1st Peak - 3.3kHz (103dB)
2nd Peak -6.4kHz (97dB)

9. WBFK-30019 ($40.19 - Mouser)
http://www.knowles.com/search/prods_...-30019-000.pdf
HF Extension: 77dB at 10kHz
1st Peak - 4.1kHz (103dB)
2nd Peak -7.1kHz (96dB)

GENTLEST SPIKE AMONG SMALL-QUANTITY DRIVERS

WBFK-30019 ($40.19 - Mouser)

HF Extension: 77dB at 10kHz
1st Peak - 4.1kHz (103dB)
2nd Peak -7.1kHz (96dB)

WBFK-30095 ($40.19 - Mouser)
http://www.knowles.com/search/prods_...-30095-000.pdf
HF Extension: 80dB at 8kHz
1st Peak - 3.4kHz (106dB)
2nd Peak -5.4kHz (103dB)

FK-23451 ($38 - Digikey/$40 - Mouser nonstocked) (630 ohm)(107dB)
http://www.knowles.com/search/prods_...-23451-000.pdf
HF Extension: 80dB at 9kHz
1st Peak - 2.5kHz (108dB)
2nd Peak - 6kHz (100dB)

 

[Bilavideo]

[size=medium]THE WBFK VS. THE TWFK[/size]

Decisions, decisions. Do you go with the dual or the best single?

TWFK ($52 - Mouser)
http://www.knowles.com/search/prods_...-30017-000.pdf
HF Extension: 73dB at 10kHz
1st Peak: 2.9kHz (103dB)
2nd Peak: 6kHz (97dB)

FREQUENCY/MAX

100Hz - 96dB
200Hz - 97dB
300Hz - 97dB
400Hz - 96dB
500Hz - 96dB
600Hz - 96dB
700Hz - 96dB
800Hz - 96dB
900Hz - 95dB
1kHz - 95dB
2kHz - 96dB
3kHz -103dB
4kHz - 98dB
5kHz - 92dB
6kHz - 97dB
7kHz - 90dB
8kHz - 83dB
9kHz - 77dB
10kHz - 73dB

WBFK-30019 ($40.19 - Mouser)
http://www.knowles.com/search/prods_...-30019-000.pdf
HF Extension: 77dB at 10kHz
1st Peak - 4.1kHz (103dB)
2nd Peak -7.1kHz (96dB)
100 - 88dB
200 - 90dB
300 - 90dB
400 - 91dB
500 - 91dB
600 - 91dB
700 - 91dB
800 - 91dB
900 - 91dB
1kHz - 91dB
2kHz - 92dB
3kHz - 95dB
4kHz - 103dB
5kHz - 95dB
6kHz - 93dB
7kHz - 96dB
8kHz - 90dB
9kHz - 80dB
10kHz - 76dB

WHAT I THINK

The TWFK clearly has better bass, which is to be expected since it is a dual woofer/tweeter. As woofers go, it's not promising anything heavy. What you're really getting is a relatively smooth, flat, performance right up through 2kHz. For heavier bass, you'll need a dedicated woofer/subwoofer. The 6dB spike at 3kHz levels to a +2dB spike by 4kHz and troughs to 92dB (-4) by 5kHz before hitting the second spike, a mild 97dB (+1) at 6kHz before the roll-off. From 6kHz, it's an approxiately 7dB drop per kHz, still pumping 73dB (-23dB) at 10kHz.

As a single wide-range driver, the WBFK has diminished bass (by about 8dB at 100Hz narrowing to about 5dB by 500Hz). Like the TWFK, it provides a relatively flat and smooth response through the midrange, spiking 8dB (2dB more than the TWFK) but at 4kHz (1kHz higher than the TWFK). It's in the HF that the WBFK offers an interesting alternative to the TWFK. Consider the following comparison.

FREQ - TWFK - WBFK
1kHz - 95dB - 91dB (-4)
2kHz - 96dB - 92dB (-4)
3kHz -103dB - 95dB (-8)
4kHz - 98dB - 103dB (+5)
5kHz - 92dB - 95dB (+3)
6kHz - 97dB - 93dB (+4)
7kHz - 90dB - 96dB (+6)
8kHz - 83dB - 90dB (+7)
9kHz - 77dB - 80dB (+3)
10kHz - 73dB - 76dB (+3)

Both "tweeters" provide excellent frequency response throughout, with the TWFK providing more bass and midrange and louder treble up through 4kHz. But by 5kHz, the WBFK clearly provides more HF and ends up with 3dB more sizzle at 10kHz. If you're looking for the IEM version of a "supertweeter" (recognizing the hyperbole in the statement), the WBFK-30019 is the treble freak's tweeter of choice.

 

[cetoole]

Quote:

Originally Posted by Bilavideo /img/forum/go_quote.gif This is where I got my wires crossed. In loudspeaker design, crossover networks have high-pass filters for the tweeters, low-pass filters for the woofers, and band-pass filters (establishing a floor and ceiling frequency) for the midrange (assuming it's a three-way crossover). In a basic First Order Butterworth, caps are used for the high-pass and inductors are used for the low-pass. When I contacted KA about this, their rep said inductors aren't used in earphone crossovers because of the hum. It's obviously not an issue in loudspeaker designs, where the woofer is sitting across the room and the woofer's output is so much greater, but in an earphone, nobody wants to deal with the distortion. Fair enough, but without inductors, I wondered how manufacturers would create a decent low-pass. KA spoke of using caps and resistors. Resistors increase impedence, using lowering volume across the spectrum. You could use a resistor to balance impedence and reign in a wild driver from drowning out its neighbor, but I still didn't see how a resistor would create a low-pass filter.

RC lowpass filter, series resistor, cap to ground. Pretty simple, works well. Just think about it as an impedance divider. The resistor has constant (practically speaking) impedance WRT frequency, while the capacitor has decreasing impedance with frequency. At low frequencies, nearly all of the signal is allowed to pass, but as you go up, more and more is shunted to ground.

 

[Bilavideo]

Quote:

Originally Posted by cetoole /img/forum/go_quote.gif RC lowpass filter, series resistor, cap to ground. Pretty simple, works well. Just think about it as an impedance divider. The resistor has constant (practically speaking) impedance WRT frequency, while the capacitor has decreasing impedance with frequency. At low frequencies, nearly all of the signal is allowed to pass, but as you go up, more and more is shunted to ground.

Cetoole, you are the man!

Hey, everybody, check this out. http://hyperphysics.phy-astr.gsu.edu...c/filcap2.html

http://www.ecircuitcenter.com/Circui.../lpfilter1.htm

http://www.falstad.com/circuit/e-filt-lopass.html

http://www.electronics-tutorials.ws/.../filter_2.html

 

[Bilavideo]

Okay, I've got a question now. Let's say we have an SR series driver, with a DC resistance of 25 ohms and impedences of 36.5 ohms at 500 Hz and 55.5 ohms at 1 kHz. How would we get the values for the resistor and capacitor in an RC low-pass filter if we wanted the cutoff frequency to be 200 Hz?

Is there anyone willing and able to walk us through this?

 

[Bilavideo]

I just got a ton of great advice from cetoole. When I get a chance I'll post some of excerpts. This guy is amazing!

 

[Bilavideo]

I got an extensive PM from Cetoole about designing crossover circuits to tailor the cutoff. It was so extensive that when I excerpted parts of the e-mail, I think I garbled some of his ideas. (Sorry about that.)

In an effort to perhaps simplify the point for a layman like myself, Cetoole summarized his advice as follows:

"Basically the way I think you would design the crossover would be just to determine sensitivities, then calculate for the correct insertion loss to balance the nominal outputs of the different drivers, and then calculate the filter capacitor."