Speaker I was my first full speaker project. Previously I had rebuilt a pair of speakers that I was borrowing from my parents (a tweeter was going bad). Originally, I bought a pair of inexpensive satellite speakers made by JBL and was going to add a subwoofer. As you will see, there is no “just going to add a subwoofer.”

The current state of Speaker I.  Triangular woofer cabinet with Vifa 10″. The triangular shape has no parallel walls, so resonances should be minimal. The cabinet is internally laberenthed via partial internal baffles. This helps brace the cabinet and controls the sound’s flow.

The satellites use a Peerless 5″ soft paper mid-range and 3/4″ Vifa tweeter. The cabinet is a Woodstyle product. The crossover is 3rd order Butterworth at 3 KHz.

I was going to New Mexico to work for the summer. Concrete blocks were too heavy to bring, so I bought some woodstyle cabinets. With driver’s remounted, I left the crossover hanging out the back thru the reflex-port. When I returned from my summer job, I reworked the crossover. What a difference! Vocals that I did not understand before were now clear as day! Later, the satellites made a trip to DCM. Testing there showed a fairly flat response (as had my own testing). The only major issues were

1. A dip at the crossover frequency of about 2 dB which gives a slightly recessed, distant sound quality, especially on vocals.  (The cross-over was 3rd order electrical Butterworth, not 3rd order acoustic Butterworth.)
2. 360 degrees of phase shift, that is, one full wavelength of time delay, on the tweeter.
3. Increased distortion around 200 Hz caused by the surround of the mid-range.

Just consider that the total speaker cost was about $225 for the satellites (including $100 for the woodstyle cabinets) and $200 for the subwoofer modules and sat/sub crossover. I spent a fair amount of time on the revised crossover design. That time spent made much more of the $23 mid-range and $13 tweeter than I think the average speaker gets from drivers that cost twice as much with a cook-book crossover design.

The original subwoofer cabinet was a five-sided; a pentagon! Don’t ask about all the hours of work it took to make this shape. Originally a Madisound 10″ woofer was used. Later, testing showed that the woofer’s efficiency was not 89 dB as claimed, but closer to 85 dB.  So the cabinets were made smaller and I replaced the Madisound 10″ drivers with a Vifa 10″ model. With an efficieny rating of 90 dB, the bass level matched the satellites. The white PVC port was also added, which extended the bass response.

One of these speakers is not like the others… it has a concrete brick for a cabinet!  You know, the 8″ x 16″ variety. Tweeter in the top chamber, mid-range in the bottom chamber. The block is covered with a faux wood paper laminate. The front baffle was covered in black felt.  This was the satellite cabinet before the Woodstyle. I’ll give you a hint… the most important cow is in the middle of the herd 😉

Although I made initial tweeter evaluations on large panels, the final speaker design decisions have to take into account the cabinet shape. Why? In the movie Fire Fox Clint Eastwood had to think in Russian. As a speaker builder you have to learn to think in terms of wavelength . One of the audio “rules of thumb” is any object within one wavelength will have an affect on the corresponding frequency range. (However, in my experience, the frequency corresponding to half of a wavelength is more likely to be affected.) The frequency corresponding to a wavelength is computed via

frequency = (wavelength) x (speed of sound)The speed of sound in air is about 345 meters / second; a typical cabinet front is 0.15 meters to 0.5 meters in width. This translates into a frequency of 2.3 kHz to 700 Hz. That is, the edge of a cabinet is acoustically “close” to the drivers and it affects the way sound radiates from a loudspeaker in the critical mid-range.

Prototype #1 Description

The design of the this prototype was motivated by the feasable shape that a center channel is allowed. The cabinet pictures with drivers mounted show that this design is pretty similar to many speakers on the market. The cabinet exterior dimensions are 8″ width, 18.5″ height, and 11.5″ depth. The construction material is 3/4″ particleboard (bookshelfs) and is doubled up in the front for the driver baffle. The left and right baffle edges are routed to 1/2″ radius corners. Internal bracing every 4″ is attatched to the sides, top, and back.

Front View	Side View

Tweeter positioning tests

These tests were performed without the double (1.5″) front baffle, or any of the internal bracing. The first test was with the tweeter positioned in the center of the baffle.
Immediately noticable is the dip between 2.5 and 3.5 kHz.

The same measurement with a little longer time window:

Now, move the tweeter so that the ratio between between the left and right sides is 2 to 1.
However, the definition of the distance to the left and right sides needs some clarification.
First, the edges of the cabinet are rounded, so the “edge” of the cabinet is where? Is it the point on the side where the curve ends? The middle of the corner? The beginning of the corner? Second, should the distance from the edge be measured from the center of the tweeter, or from the edge of the tweeter? I just wish I could remember what I did for this test!

The results of the measurement show that the center of the baffle is not the best location
for a tweeter. The dips in the 2.5 kHz to 3.5 kHz region are now almost gone.

The most prominent anomyly in the response is at 3 kHz. A waterfall plot of this impulse response shows that this frequency has a substantial time span. The wavelength at 3 kHz is 11.5 cm, which does not corrspond to any physical structure, nor does half of this wavelength. But 11.5 cm is for a wave traveling in *air*, not in a solid material. Thus, my conclusion was the peak at 3 kHz was most likely due to a cabinet resonance. Ergo, bracing of the cabinet, even when made out of 3/4″ particle board, was going to be necessary.

Final Raw Driver Measurements

Can’t remember the measurement distance. The mid-ranges are wired in parallel for this test.  No data on output signal level. With a time measurement window of 7.3 ms, the response is affected by reflections from within the room (the squiggling on the responses).

Both response curves show a two major dips in the 2 – 10 Khz frequency range. Further testing is being conducted to locate the source (diffraction, the speaker stand, driver problems, etc.). One area of suspicion is the mid-range; the tweeter positioning tests were made without the mid-ranges present. Consider the table of length-to-frequency conversions:

Distance	Measurement	Full Wavelength Frequency	Half Wavelength Frequency
Mid-Range Cone Depth	1″	13.5 Khz	7.3 kHz
Mid-Range Cone Width	3.5″	3.9 Khz	1.9 Khz
Distance from Tweeter to Mid-range Surround	2″	6.8 kHz	3.4 kHz
Distance from Tweeter to Right Edge	2-3″	4.5-6.8 kHz	2.3-3.4 kHz
Distance from Tweeter to Left Edge	4-5″	2.7-3.4 kHz	1.4-1.7 kHz

From the dips in the tweeter response, the possible culprits would be the recessed nature of the mid-ranges, and the distance from the tweeter to the mid-range edge. The latter suggests that recessing the mid-range deep enough to “hide” its surround from the tweeter waveform might be desirable.

---

Copyright © 1998,1999 John Lipp

GRAPH COMING SOON

This project has many possible applications, and as many possible low frequency designs. Two possible designs are illustrated in the graph above (a Bode plot). The yellow line corresponds to a 7.3 liter sealed box tuned to a Butterworth alignment. However, the low frequency cutoff is only 110 Hz, which is not low enough for use as a THX center channel (80 Hz is the spec). If connected to a sub-woofer, the cross-over would have to take into account the natural high-pass function, or be set to a high frequency (at least 400 Hz). The blue line is a 15 liter vented cabinet with the port tuned to 63 Hz (a port 3″ in diameter about 7″ long). The low frequency cutoff is now less than 70 Hz!

Another design (that I don’t know how to simulate the response of), is a transmission line design. One quarter of a wavelenth at 55 Hz is 1.56 meters. Folded, the transmission line will be reasonably compact. The cone area is 80 cm², thus, the line’s volume (for two drivers) as 25 liters. The final design will be significanlty larger due to the internal bracing used to form the acoustic labyrinth of a transmission line. But even if doubled, the total volume is less than 2 cubic feet. Most subwoofers are quite a bit larger.

A factor that weighs in favor of the tranmission line is the desire to make the front baffle very wide to minimize baffle diffraction. At the same time, the baffle needs to be tall enough to put the drivers about ear level. The minimum height for the latter requirement is about 42″. Using the golden mean, the cross section of the transmission line is 2.75″ by 4.5″. Maximimizing the baffle area means 2.75″ is the internal depth of the cabinet volume.
Using 3/4″ MDF the depth is deeper by 2.25″ for the double thickness front baffle and
single thickness rear wall. Each side of the 4.5″ wide line has half of a wall width (remember, the line is going to be folded). The exterior walls add an additional wall thickness to the total baffle width. My current plan is to use 3/4″ MDF or 3/4″ baltic birch / apple-ply for internal walls, 3/4″ MDF for the outer walls. The calculation is

2 x 61.5″ x (3/8″ + 4.5″ + 3/8″) / 42″ + 3/4″ = 16″Some room for the crossover components also needs to be reserved, so the baffle will be about 18″ wide.

However, a monolith is still not the final shape. (Yep, the dimensions are pretty close to 1 by 4 by 9!) An uneven baffle width will further reduce the affects of edge diffraction, possibly more so than attempting to make it wider.

---

Testing Methodology

Shown above is the rather crude testing methodology that I began with. The white board in the first two pictures is a 40″ x 60″ foam board that can be found at any office supply store. It is about 0.2″ thick, so not particularly strong. I glued an additional piece of foam board on the backside to stiffen up the area around where I mounted the drivers. The last photo shows a double layer, 1.5″ thick “skinny” test baffle that I built, the stand for it, and two more 40″ x 60″ foam boards. The two white circles on the test baffle cover up the mounting holes
for the Vifa and Morel tweeters. All the tweeters were flush mounted for the tests, as was the mid-range.

The frequency and phase responses shown below have to be taken with a grain of salt. Later testing showed that the foam board is at least partially acoustically transparent in the mid-range frequencies, and possibly has a mid-range resonance as well (see The Tweeters, Part 2).
By the time I made this determination, I had already returned the Focal tweeter, so I have not been able to verify which response artifacts are “real” and which are “foam board.” However, discussions with fellow speaker builders verified the Focal’s severe resonance problem above 10 Khz as real. Between that, the unit’s high price, and the poor arival condition, I incurred no emotional loss returning the Focals.

Distortion measurements were all made at a distance of less than an inch to maximize the number of valid data bits. Since the card’s sample rate is just over 40 Khz, 3rd order harmonic distortion measurements above 7 Khz were not possible, nor 2nd order harmonic distortions above 10 Khz.

Vifa D27SG-05 Tests

Frequency Response

 

Waterfall CSD

There is some activity associated with the response dip around 7kHz. This indicates some type of possible resonance behavior. As mentioned in the intro, this could be an artifact of the foam board.

THD

Signal level of -19.5 dB.

Impedance

Signal level of -39.0 dB measured while on prototype cabinet 1. This is very smooth and does not indicate any resonance behavior as measured while on the foam board. This smoothness is possibly the result of ferrofluid’s mechanical damping.

---

Focal TC90KB

Frequency Response

Notice the dip at 2kHz, and the crazyness above 10 Khz. The dip at 2 kHz is present in Focal’s own response data, as is the wierdness above 10 kHz.

Waterfall CSD

Wow, look at the activity above 10Khz! This is pretty bad ringing.

THD

Signal level of -19.5 dB. Notice the Focal unit’s THD is evenly distributed between the 2nd and 3rd harmonics (plot 2 of 2). Both the Vifa and Morel units’ distortions are dominated by 2nd order distortion products. This characteristic is maintained at lower signal levels.

Impedance

Sorry, no impedance plot for the Focal. However, other data I recorded shows impedance anomylies at 2.5 kHz, 5 kHz, and several above 10 kHz.

---

Morel MDT-30-S

Frequency response and phase plots.

Notice the small dips at 4.4 kHz and 12 Khz.

Waterfall CSD

The peaks/dips seem fairly well damped. However, the z-axis is only 0 to -20 dB whereas on the other drivers the range is 0 to -25 dB. 🙁

Total Harmonic Distortion

Signal level is -19.5 dB.

Impedance

Signal level is -39.0 dB. Note the anomyly around 12-15 kHz. There is also an anomyly around 5 kHz, but it is much more subtle. This indicates a mechanical resonance in the system which can store energy. At higher test signal levels the anomylies are more pronounced.

---

These are the results of my March 1998 search. Scan-Speak and Peerless are absent; no shielded drivers are available. Dynaudio and Morel place the magnet inside of the driver basket. That placement gives a tight magnet field that is close to, but I don’t believe as good as, bucking magnet shielding. Eton makes a 4″ shielded mid (at $98), but nothing larger.

Seas also makes some shielded co-axial drivers; a mid-range with the tweeter located
where the dustcap is normally found. The mid’s cone acts as a horn on the tweeter and modulates the tweeter’s output via the mid’s cone motion. The latter is a non-linear distortion. Why then would anyone use a co-axial driver? The big reason is the co-location of the mid-range and the tweeter. The time delay between them is minimized, and almost invariant with changes in the listening axis. The modulation distortion is minimized by limiting the mid-bass and bass content sent to the mid. A 400 Hz cross-over to another mid-range or mid-bass unit has been commercial used to accomplish this task. Since the project goal is an MTM, not an WTMW (W for woofer), no co-axials were considered.

Param	Vifa	Vifa	Vifa	Focal	Seas	Seas
Model	M17SG-09	P17SJ-00	???	5N411LB	P14RC/TV (H626)	P17RE/TV (H690)
Price [1998]	$33.50	$34.90	$46.70	$84.50	$42.20	$52.50
Size	5″	6.5″	6.5″	5″	5″	6.5″
Material	Paper	Paper	Poly	Neoflex	Poly	Poly
Fs [Hz]	54	34	41	45	40	34
Qts	0.35	0.34	0.35	0.35	0.21	0.27
Efficiency [dB/2.83v]	88	89	87	87.5	89.5	88.5
Vas [l]	12	53	33	14	18.9	30.5
Xmax [mm]	2	3	4	3.25	3	3
Power [w]	35	50	70	60	60	80

These prices were “current” of 03/31/98 from Madisound. Xmax ratings are peak one way.

The Vifa D27SG-05 does not have “tinsel leads” or a similar technology. I interpret this to mean a crossover frequency below 3 kHz is not likely to lead to long term tweeter reliability.  A crossover frequency above 3kHz limits the mid-range selection to 5″ or smaller drivers.

This is not a cost-is-no-object design, which yields the Vifa M13SG-09 and Seas P14RC/TV as the two best candidates. The P14RC/TV manufacturer response curves shows some response anomylies in the 1-2 kHz region where the human ear is most sensitive. My conclusion: get the Vifa M13SG-09.

The measurements of the Vifa M13SG-09 were good enough that no other units have been examined.

Vifa M13SG-09

The raw driver. Several points are worth noticing. The translucent plastic tabs placed in the mounting screw holes keep the surround from getting crushed during shipment. Very Nice! The magnets (one is a bucking magnet, placed such as to cancel out the magnetic field of the other) are covered with a steel cup.Although the cone is paper, it appears quite shiny. This is from a plastic coating, which gives the mid a “wet” look. The dustcap is well damped, but is not soft like the Vifa poly mid.

The first frequency response test was performed on a small test box that was also used for calculating driver Vas. The measurement distance was 46 cm and the time window is about 18.6 milliseconds.The small dimensions of the test cabinet make it difficult to evaluate which of the response anomilies can be traced to the driver. The manufacturer’s data sheet shows a broad 1-2 dB dip at 2 kHz and a narrow 3-5 dB dip at 6.5 kHz.

The LAud results of mid-range B tested for its Thiele-Small parameters.

The table below summarizes the Thiele-Small testing I have done so far on the dozen mid-ranges that I ordered. (Yep, a dozen!) The results are before any break-in of the suspensions. The test signal level was -30 dB.

	Spec	A	B	C	D	E	F	G	H	I	J	K	I
Re [ohm]	5.6	5.6	5.5	5.6	5.45	5.5	5.6	5.5	5.5	5.6	5.6	5.6	5.6
fs [Hz]	54	54.45	57.71	57.25	55.31	56.63	57.96	56.19	54.22	52.87	53.82	56.95	58.26
Qts	0.35	0.370	0.383	0.379	0.356	0.351	0.349	0.371	0.351	0.368	0.352	0.352	0.373
Vas [l]	12	11.5	10.2	N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A

 

---

Copyright © 1998 John Lipp

The selection of high-quality, sheilded tweeters in March of 1998 was darn sparce. Only one unit from each Seas, Focal and Morel. Vifa was the only player offering 3 choices, the D27ASG-05 (aluminum dome), the D26SG-05 (textile dome), and the Vifa D27SG-05 (silk dome).  Unfortunately, none of Vifa’s shielded tweeters include Vifa’s “tinsel leads.”  The “tinsel leads” must be more flexible, more resilant to long tweeter dome excursions.  All of the non-shielded tweeters that carry this feature rate 100 watts of power handling with a 12 dB/octave crossover at 3 kHz.  Similar tweeters without these special leads are also rated at 100 watts, but with a 12 dB/octave crossover at 4 kHz. (It is also worth noting that all 3 shielded tweeters have lower power handling ratings than similar, non-shielded units. I can only speculate the cause; higher expected tweeter power levels in a home theatre environment?) The aluminum and textile dome units have resonant frequencies about 1.5 Khz, so the choice Vifa tweeter was the silk dome unit with a resonance of 1 kHz.

So far I have been able to evaluate the Vifa, Focal, and Morel units. At some point I may check out the Seas unit. It is rather expensive at $64, but looks to be a top-of-the-line tweeter with every bell and whistle that Seas has to offer (even silver voice-coil wire). My final selection is the Vifa, with the Morel being held for further testing (modification of the back chamber for lower resonance).

 

Model	Vifa D27SG-05	Focal TC90KB	Morel MDT-30-S
Pic
Material	Silk	Kevlar	Textile
Faceplate	Plastic	Plastic	Metal
Ferrofluid	Yes	No	Yes [?]
Resonance Frequency (measured)	1000 Hz (1090 Hz)	950 Hz	700 Hz (837 Hz)
Moving Mass	0.3 g	?	0.44 g
BL product	2.7 NA	?	3.5 NA [?]
Qts (measured)	? (0.86)	0.90	0.57 [?] (0.64)
Qes (measured)	? (1.24)	1.21	? (1.02)
Qms (measured)	? (2.83)	3.47	? (1.71)
Re	4.6	5.9	5.2 [?]
Efficiency	92 dB	92 dB	90 dB
Power Handling	80W	75W	200W
Recomended Crossover	12dB/Oct @ 4k	12dB/Oct min	?
VC Height	1.6 mm	2.2 mm	2.7 mm [?]
Gap Height	2.0 mm	3 mm	2.5 mm [?]
Price	$22.95	$59	$60

 

Notice the white stuff around the edges of the Focal tweeter.  It is from the styrofoam packaging where the surround was touching during transportation. I could find nothing to remove the styrofoam contamination without also harming the foam surround. Needless to say, I was pretty disappointed.

 

The idea behind this project was to fulfill many speaker project needs:

• My satellite speakers in Speaker I are showing thier age. Another crossover redesign might help, but it won’t overcome the shortcomings of the midrange unit, the Peerless KO-50G. I bought the mid-ranges over 10 years ago on a college student’s budget. I’m sure their design originated many years before that.
• I wanted to give my brother Doug a fun wedding gift. His satellite speakers are similar to mine, but without my latest x-over design. Unfortunately, I wasn’t even able to finish the first prototype cabinet in time for the wedding. On the good side, now that he knows about the project, he can give some input on the desired cabinet shape.
• A colleage at work needs (Ok, wants) a center channel to match up to his Ariels. It is going to sit above his TV, so magnet sheilding is a must. Unfortunately, the drivers used in the Ariels are not magnetically sheilded, so a straight forward adaptation is not possible.

These needs culminate into the following initial project design requirements:

• Vifa or better drivers. Anything less won’t match well to the Ariels.
• All the drivers are to be sheilded. (This really limits the choices!)
• Bass response down to at least 100 Hz, 80 Hz preferred (the standard for a THX center channel).
• The desired system efficiency is in the range of 94-96 dB in the mid-range. This is required for my brother’s system to properly integrate with the bass units whose efficiency is 96 dB. My bass units are only 89-90 dB efficient, so I see an electronic cross-over in my future.
• An MTM (mid-range / tweeter / mid-range) setup. This helps satisfy the need for efficiency, increases the bass capability, keeps driver diameters smaller, and helps even out vertical dispersion.

Being my third major speaker project, I want to break new ground (for me) in the design of Speaker III. Thus, I have the following additional, self-imposed design goals:

• Time-delay on the tweeter to align the driver responses.
• As near a constant group delay cross-over design as possible. This is in addition to the usual requirement for flat frequency response.
• A baffle shape selected with attention to the issue of cabinet diffraction loss.
• The Ariels are a transmission line. Although too big for a center channel, they might do for the main satellites for my brother or I.