Prepared for: Transducer Theory
Prepared by: Kenneth Stahl, Designer
February 15, 2011
Proposal number: 000-0001
2109A Woodmar Drive Houghton, MI 49931
906-231-2939 [email protected]
To build a set of Hi-Fi studio quality mixing monitors. Keeping in mind the lifestyle of a recording engineer size matters, trying to keep them transportable is important. Also it will be important to produce High-SPL’s.
* Develop an understanding for Sound Pressure Levels and what will be the average listening level.
* Examine the design trade-offs that exist in loudspeaker design
* Developing a standard for low frequency response.
* Compare Box-Types.
* Develop a system to minimize diffraction
* Examine Various Woofers (Simulate).
* Examine Various Tweeters.
* Determine Box Size and Makeup.
* Dealing With Standing Waves (Bracing).
* Build Ideal Crossover and examine summations.
* Accounting for a budget and making economically sound choices.
Design Proposal 1
The last chain in a recording studio is no doubt the listen medium. A biased listen medium is detrimental to the success of a project, studio, engineer, and to the music itself.
Unfortunately most reasonably priced monitor systems are not suitable for actual studio use. With mass production and profits as a motivation it is no real surprise that this is the case. The Rocky Mountain Way loudspeaker system is an economical alternative to what is currently on the market and provides unbiased sound reproduction.
First off defining the SPL capability is the first part of the lengthy design process.
SPL’s or Sound pressure levels are is defined as the physical intensity of sound relative to a standard reference value. Also know as the measurement of how loud a system is and is measured in deciBels (dB). In the planning stages of developing a loudspeaker system, it is necessary to determine what the desired output or how loud? With the design objective in mind, building studio monitors, an average of measurements was first observed at a comfortable mixing level. After which the next step was to push the volume until the level was both too quiet and too loud. The averaging process for each level was taken by three measurements throughout the day and can be seen with the following results:
Design Proposal 2
COMFORTABLE SPL: 89.33 DB
WHAT IS TOO LOUD: 96.66 DB
WHAT IS TOO QUIET: 72 DB
OBSERVED SPL (DB)
COMFORTABLE LEVEL TO LOUD TO QUIET
KENNY’S SPL TESTS (DB)
Comfortable Level To Loud
1-19 @ 4:00pm
1-20 @ 10:00am
1-19 @ 11:00PM
A loosely coined listening environment would to listen from 1 meter from the “Rocky
Mountain Way” system because they will likely be on a desk at that distance from the intended listener. Therefore the above tests were performed at roughly 1 meter from a loudspeaker system, listening to a high-definition recording of Donald Fagen’s “Morph the Cat”.
The results show that the comfortable listening average is 89.33 dB.
Design Proposal 3
When accomplishing High-SPL’s in a loudspeaker design it is important to understand the relationship between an amplifiers output power and the relative SPL change. According to the Speaker/Watt relativity concept for each doubling of power you add 3 dBw SPL’s
loud speaker system, this will come in handy when looking at the desired SPL in the later stages of the design process. In the following example amplifier power is shown in relation to the relative SPL change.
Power vs. SPL dBw
0 10 20
Plummer, Christopher Personal Interview. 11 January 2011
Design Proposal 4
In loudspeaker design there are three major areas of interest, bass response, size and SPL capability. Identifying higher importance to one of the areas will result in the other two being less appealing. Taking these principals and applying them to a pie-chart of a fixed size it becomes apparent. By shifting each piece of the pie to a larger section you would be indicating that particular design criteria is more important and the the other two sections would suffer
. Below there are four pie charts indicating the various difference between these three categories. Note: In this example when the size is smaller than the percent of the pie is larger.
2 Murphy, John. Introduction to Loudspeaker Design. City: True Audio, 1998.
Design Proposal 5
Considering the design objectives of a loud, flat frequency in the music spectrum
(20Hz-20kHz), and small enclosures a trade off is necessary. This situation is described as the “Impossible Speaker” because of the physics of how a loudspeaker replicates sound
Taking that into consideration, monitoring systems for studio applications two of the the controllable parameters are most important bass response and high SPL. It is a give and take but the “Kenny’s Desired Loudspeaker chart” it is an example of the compromise necessary to meet the design objectives.
This is where the experimenting process became much more subjective to genre specific music. Therefore a test was developed using three tracks and listening too them on a large well balanced system that was fully capable of replicating the entire range of the audio spectrum 20Hz-20kHz. By using a 48dB roll off plugin to take away the low frequencies until the test was used to determine at what frequency bass response was unaccept-
Murphy, John. Introduction to Loudspeaker Design
Design Proposal 6
able. The three songs in the test were “Morph The Cat” - Donald Fagen, “Girly Worm” -
Mike Viola, and “Come Together” - Beatles.The results were similar for each song only varying slightly and can be seen in the following chart.
USING LOGIC’S BUILT IN EQ AND A GOOD LISTENING ENVIROMENT
MORPH THE CAT GIRLY WORM COME TOGETHER
GREAT LOW END (BOTTOM OF
GREAT LOW END (TOP OF
GOOD LOW END (TOP OF RANGE)
NOT ACCEPTABLE LOW END (TOP
FREQUENCY VS. PERCEIVED SOUND
Great Low End (Bottom of range) Great Low End (Top of range)
Morph The Cat
Good Low End (Top of Range)
Not Acceptable Low End (Top of Range)
Design Proposal 7
Looking at those results it can be observed that if this were an ideal world a designed loudspeaker would go down to 20Hz. Since it is not, satisfactory low frequency response of this loudspeaker system would need reproduce signals nominally to 55Hz. To fully satisfy that requirement, the design should consist of F3 frequency of around 50Hz. F3
Frequency is the point at which the SPL is 3dB off nominal
For mixing music frequency response should be as large as possible. It is impossible to know exactly what a song is going to use so it is best to get the largest range. For example a bass guitars fundamental on the low e-string is 41Hz (Green). A shaker or drum cymbal can easily exceed 10-20kHz. It is extremely important to have the same SPL for both instruments to make quality decisions when recording/mixing/editing a song.
Options for box type that were greatly considered for this loudspeaker system included closed and vented box. A closed box is a fully sealed box that lets minimum air escape the enclosure. Where as vented enclosures have a vent (a pipe) that opens and allows air to escape and resonates at a certain frequencies. Reading through guides, examples, and speaker print offs leads a to a clarifies a thought, “A very clear advantage for
Plummer, Christopher Personal Interview
Design Proposal 8
. Sealed boxes have a slighter roll-off characteristic after
F3 at a rate of -12 dB/octave. Where as a ported system cut-off is a steep -24 dB/octave.
That means that often lower frequencies can be heard using a sealed design even if the ported F3 is lower
. Both of these points are arguments for a sealed box however there many equally as valid arguments for ported boxes. Below is a a pros and cons list that examines both cabinet types.
Noakes, Cameron. "Diy Audio Corner." Enclosure Dilemma: Ported vs Sealed. Available from http://diyaudiocorner.tripod.com/dilemma.htm
. Internet; accessed 20 January 2011.
6 Noakes, Cameron. "Diy Audio Corner." Enclosure Dilemma: Ported vs Sealed.
7 Graph from -
Noakes, Cameron. "Diy Audio Corner." Enclosure Dilemma: Ported vs Sealed.
Design Proposal 9
Focusing on the mechanics of the box and speaker interaction there are some important notes in the above chart. The transient response of a closed box will be better than the ported box. “Q” is a measurement of resonant magnification in a box. More importantly it is the measurement of electrical, mechanical, and pneumatic circuits used to represent resonance control factors
. A sealed box loudspeaker system has lower Q levels than a ported box and even has some transient perfect points in which the transient frequency is
0 or critically damped. Which can help to increase the detail of sound coming out of the speaker
. One study showed by average a good combination of a closed box system would be a F3 around 50Hz, a Qtc from 1.2-2.0, and a size less than 2ft 3 . This system demonstrates strong bass
Because this loudspeaker design has the intended use of mixing musical recordings better transient response,Punchy bass rather than oomph-y bass, and a the lower level roll off in the extreme low end is more important to achieving that goal. In conclusion a sealed/ closed box is going to be used in this speaker design.
Dickason, Vance. Loudspeaker Design Cookbook. Peterborough, New Hampshire: Old Colonial Sound Lab, 2006.
Dickason, Vance. Loudspeaker Design Cookbook.
Dickason, Vance. Loudspeaker Design Cookbook.
Design Proposal 10
Sound diffraction is when the sound wave bounces off a hard surface and develops a second point source. This can cause major drops in specific frequencies for example in an experiment found at the author (Ludwig) describes a test in which there was 5dB drop at 7kHz when they mounted the speaker on the edge of a baffle. To combat this issue he mounted the speaker flush with the baffle and it completely removed the 5db drop. “It was quite clear,” he writes, “mount all drivers flush with the front panel”
. Other suggestions included baffle size, foam on the outside of the cabinet, beveled edge, and driver location.
Tests of each of the suggestions above were done by Vance Dickason and the results of which would indicate that each parameter has somewhat of a “say” in the amount of diffraction.
Ludwig Sr., Arthur. "Silcom." Loudspeaker Construction. Available from http://www.silcom.com/~aludwig/Loudspeaker_construction.html
. Internet; accessed 20 January 2011.
Design Proposal 11
To combat diffraction a combination of techniques will be used. 3/4” Black-Hole, acoustical foam, will be a placed on the frontside of the baffle and the drivers will be mounted to the flush to the Black-Hole. Using the damping foam will dampen the waves that would usually bounce off the cabinet. Being largely untested, for the practical implication of how you would mount the driver units flush to a dampening material, the results seem logical. Using these two techniques should greatly reduce any unwanted effects from diffraction.
* Seas Prestige H1411 (as seen below)
* Morel SCW 636
* Morel EW 638
* Morel CAW638
* Fostex FW167
* HiVi D6.8
Woofers are the large drivers that generally reproduce only mid and low frequency sounds. The search for woofers included only 6-
10” drivers, with flat frequency re-
Design Proposal 12
sponse and minimal break up effects. When the selection process began there were Two important things that were considered for driver selection, cost and low frequency extension. It was shown before that low-frequency extension is a product of both the driver and the box so below are charts of the above listed drivers in different box types.
22.39 Meters not
Design Proposal 13
10 0.255 Meters
These charts show the differences between a QB3 vented box and closed box F3
Design Proposal 14
(low frequency extension). From the charts 4 drivers stand out as having the highest performance and produce the results that would most satisfy the original design objectives.
They are as follows:
H Alpha (F
1. SEAS PRESTIGE H1411
0.9889 1.0065 0.9776
42 56.6 22.39
2. SCAN-SPEAK 18W/8535-00
1.0578 1.3552 1.1153 28.5 56.6 0.049
3. MOREL CAW938
Q ts to high and not on chart
4. SCAN-SPEAK 18W/4531G00
1.1305 1.7964 1.2702
38 28.3 0.166
F3 AND VOLUME:
CLOSED BOX F3= 44HZ, 56.6 LITERS
VENTED BOX F3=28.5HZ, 56.6 LITERS
CLOSED BOX F3=42, 56.6 LITERS
VENTED BOX F3=38, 28.3 LITERS
Examining the Seas H1411 10” woofer, there are many reasons to like this driver.
Most notable thing about this driver is that it gets positive results from in in a closed
Design Proposal 15
speaker enclosure. www.madisound.com
suggests a sealed box enclosure thats size is
1.5 cubic feet with 50% filling, which will result in a F3 about 48Hz. However when put in a
2 cubic foot box the simulated frequency response is more flat an the F3 is reduced to
45Hz. Not only can this driver deliver the low end that is desired, it also has a nice roll off in the higher frequencies which is a natural crossover. Which will help when it comes to designing the crossover this will become important. The H1411 meets the standards that are laid out in the design objectives. Below is the simulated frequency response of the H1411:
Design Proposal 16
The Scan Speak 18W/8535-00 yields impressive results in a vented box with a 2 cubic foot volume. With an F3 around 28Hz it appropriate to say it is able to reproduce the low frequencies. However that is not without trade offs as mentioned in the box type section. The next simulation shows the results of the simulation of the Scanspeak 18W/8535-
00 in a vented enclosure. *Note* - The excursion of this driver is pretty high.
Design Proposal 17
The Morel CAW938 is also a very nice driver. It has a flat frequency response in a closed box with an F3 of 42 Hz. The CA938 has a interesting magnet design that places the magnet on the inside of of the speaker frame. However something that this decreases is the amount of movement that the driver can have before it hits its excursion level and then damages itself. As can be seen in its simulation the excursion level skyrockets at
The driver that will be chosen for this design is the Seas H1411 10” Woofer. This driver meets all the expectations laid out in the design requirements and it is has positive review for being a good sounding woofer. The stats reinforce the positive reviews on the
Design Proposal 18
web. It will also work well in a 2-way closed box design and will be easy to incorporate a tweeter into the system.
* Morel CAT 308
* ScanSpeak D2904
* Seas E0047
* Seas Prestige 27TDFC
* Morel ET338
* Fostex FT48D
* Audax TW034
* Audax TW025A26
* Sease 27TBC/G
* Audax TW025A28
Tweeters have much less to do with the box as they have there own there own enclosure. The parameters that were taken into consideration when looking at tweeters were their resonant frequency, power handling, and the cost. In the crossover section it will be seen that the lower resonant frequency of a tweeter the easier it will be to work with. That being said it is one of the highest factors to take into account when weeding out the
Design Proposal 19
tweeters. Another factor that were taken into consideration when trying find the right tweeters was professional recommendations.
Below is a chart that compiles the information of all the tweeters that were examined for this loudspeaker design.
The are three frontrunners the seas 27TDFC the Scan Speak D2904, and the Seas
E0047. A recommendation that was found at www.madisound.com
explains that the
27TDFC is a great combination to be used with the Seas H1411 woofer. After a more in
Design Proposal 20
depth look this tweeter has a reasonably flat frequency response from 1Khz and up which can be seen in the following graph:
)*(#+,-#./,0%#1!/!2#3#,4%%3 "#$%!& '(!&
The Scan Speak D2904 is also a very good option, it has a
similar frequency response to the
27TDFC. As seen to the left the only difference is that this tweeter is has a metal dome which results in
a peak above audible frequencies.
[,RG&,!I)LJ<,]RE,'RX',TEYY,bV##GHFPH#',%RS#GHc NJKK '` (!
The T29MF001, “MAGNUM” is a 25mm magnesium dome tweeter with a patented
Neodymium based magnet system. It is the tweeter of choice for those who seek extremely precise and realistic sound reproduction combined with a relatively low crossover frequency.
A unique HEXADYM patented magnet system based on 6 radially magnetized NdFeB magnet blocks. Effi cient ventilation and damping of every potentially resonant cavity behind the dome, surround and voice coil. Moderate magnetic stray fi elds present no problems in AV installations.
A generously underhung voice coil ( + and - 0.5mm ) ensures low distortion even with low crossover frequencies.
An optimally shaped magnesium dome membrane which behaves like a piston throughout the audible frequency range and shows a controlled break up above it.
A homogenous, linear surround manufactured by SEAS from SONOMAX, a soft polymer material of high climatic stability.
Flexible lead-out wires which ensure a good connection between voice coil and terminals.
This arrangement also helps to prevent lead breakage due to the large excursions encountered when low crossover frequencies are used.
Low viscosity magnetic fl uid which provides excellent cooling while maintaining a low resonance frequency.
6,0 mm machined aluminium front plate with a moderate horn loading characteristic which ensures linear frequency response, and a stiff and stable connection to the cabinet.
A substantial injection-moulded rear chamber made from zink eliminates unwanted chamber wall resonances and conducts heat away from the magnet system.
The Seas E0047 is less appealing for studio speakers because its not fully flat frequency response. Granted the boost at
3KHz is only 2-3dB off of the average. These tweeters do though have a good reputation for sounding good.
100 1 000 10 000
The frequency responses above show measured free fi eld sound pressure in 0, 30, and 60 degrees, mounted in a 0.6m by
0.8m baffl e. Input 2.83 V
, microphone distance 0.5m, normalized to SPL 1m. The impedance is measured without baffl e using a 2V sine signal.
Recommended Frequency Range
2000 - 25000 Hz
Voice Coil Resistance
Voice Coil Inductance
Effective Piston Area
Magnetic Gap Flux Density
8 cm 2
27TDFC is the tweeter that will be in this loudspeaker system. A decision not made lightly, the choice was made both on price but also Madisound’s recommendation. Other softdome tweeters were in the top 10, however due to higher resonant frequencies they were eliminated.
Design Proposal 22
Physical Space of the woofer will greatly determine the size of the box. Using a 10” speaker is going to mean a larger box than using a 6” speaker. Taking that into consideration the H1411 woofer measuring 10.5” with flange will require at least 1-2” on each side for proper mounting. As explained in the woofer section the best comprise of results for the woofer is to build a box that is 2 cubic feet. Taking into consideration the golden ratio of speaker building the internal dimensions of 2 cubic foot box should be 24.46” x 15.12” x 9.38”. However trying to reduce the hight of the speaker and to provide for extra volume lost for parts inside the cabinet, the dimensions would need to be modified. Therefore the size outside of the box size will be 24 1/2” x 15 1/4” x 14” and 23” x 13 3/4” x 12 1/2” inside the 2 layers of wood.
Using two layers of wood will be essential for the completion of this design for a number of reasons. First of which is to try to reduce the cabinet resonances using only one board. By having two types of wood it will increase the cabinets rigidity and increase the mass of the walls. This will effectively lower the wall vibrations which helps to eliminate the unwanted cabinet effects. In the North Creek Cabinet Handbook they recommend using 3/
4” MDF and 3/4” Baltic Birch Plywood. It states that this the Plywood composite panel will
Design Proposal 23
be 4 times as stiff as one layer of MDF and in combination they will greatly reduce panel resonances.
North , Creek. Cabinet Handbook. Old Forge New York: North Creek, 1992.
Design Proposal 24
Bracing is as it sounds taking wood and adding braces between the sidewalls. This is a technique used to eliminate standing waves inside a cabinet and to reduce the cabinet
The implementation will include at the very least one brace with an approximate hight of 3-4 inches. The position of the brace will be above the slightly above the woofer as to not introduce reflections out the front of the woofer. It will be fastened into a brace and glued that is glued into the sidewall. It will also include an assortment of 1/2”-1” holes to minimize the amount of cabinet volume lost by adding the brace.
In a multiple driver loudspeaker systems crossovers are almost necessary. A cross-
Tweeters are small drivers that produce high frequency signals (think violin). W/O crossovers the full spectrum of sound would be transmitted to the tweeter and would cause the tweeter to fail. Similarly low-frequency woofers cannot reproduce high frequency sound sources but rather low
North , Creek. Cabinet Handbook.
14 Murphy, John. Introduction to Loudspeaker Design. City: True Audio, 1998.
Design Proposal 25
sound sources like drums. A multi-speaker loudspeaker system cannot exist effectively without a crossover network.
The magnitude at which the crossover reduces undesirable frequencies is called its
“order”. The rolloff of is measured in dB per octave and the following chart shows the difference between a 1st-4th order crossover.
C T A V E
With an active crossover, the amplifier is connected directly to the driver, and the only thing between them is the loudspeaker cable.
capacitors and inductors to filter out the undesirable frequencies. These components take away power from the speaker and reduce the efficiency of the speaker. The simple fact is
Rod, Elliott. "Active Vs. Passive Crossovers." January 11, 2004.http://sound.westhost.com/biamp-vs-passive.htm
Design Proposal 26
that active crossovers are just simple and allow the user to change parameters much more effectively.
Why Active Crossovers are preferred in loudspeaker systems:
2.Distortions are handled in one band preventing unwanted distortion to the wrong driver.
3.Correction of undesirable bumps in frequency can be quickly changed
4.No change in impedance seen by the amp (direct connection to the driver)
5.Higher order filters can be used without loss of system efficiency.
6.In passive loudspeakers preforming at high SPL voice-coil heating will change the impedance of the drivers. This effect will cause a corresponding change in the crossover frequency of passive systems and is not in an active system.
A passive Crossover uses electronic components such as capacitors and inductors to divert undesired frequencies that is delivered to a speaker. The biggest downside to a passive crossover network is the introduced stress. Not only are they difficult to make minuet changes, they also reduce the efficiency of the driver network. Inductance (capacitance
Newell, Philip, and Keith Holland. Loudspeakers. Oxford: Focal Press, 2006, Kindle Edition.
17 Newell, Philip, and Keith Holland. Loudspeakers. (1872-76)
Design Proposal 27
too) has a resistive component that will reduce the amount of power that is delivered to the amplifier.
Benefits to using a Passive Crossover:
1. In the box (no need for external gear)
2. Lower Initial Costs.
Therefore to optimize a speaker system it is almost necessary to go with an active system. With the design goal to create a set of high quality mixing monitors it is a nobrainer. An active system using a DBX Driverack will be used to create the best sounding and most efficient set of speakers.
Design Proposal 28
Using the H1411 Woofer and the 27TDFC for the tweeter the crossover on the tweeter will need to be low to accommodate for the roll of the woofer. It is important to know that you want to be at least 20dB down at the Fs of the tweeter because it will damage the tweeter if you are not.
So with this combination the best results are seen when you are using a 2nd order crossover @ 2Khz for the tweeter and a 2nd order crossover @
1khz for the woofer. The results can be seen below *note that the sum is the result of the combination of the drivers.
Plummer, Christopher Personal Interview
Design Proposal 29
500 750 1000 1500 2000 3000 4000
As is shown in the chart, this combination will create relatively flat frequency response throughout the frequency spectrum. Also the resonant frequency of the 27TDFC is 550 at which the tweeter is 30dB down. Thus preventing any damage to the tweeter.
Design Proposal 30
Expenditure by Category
Total $ !
Madisound eBay eBay eBay
Seas Prestige H1411
Crown XLS 402
DBX DriveRack PA
Parts and Pieces
Parts and Pieces
Design Proposal 31
27TDFC Manufactures Spec Sheet
H1411 Manufactures Spec Sheet
DBX Driverack PA Manual
Design Proposal 32
Murphy, John. Introduction to Loudspeaker Design. True Audio, 1998.
Rod, Elliott. "Active Vs. Passive Crossovers." January 11,
2004.http://sound.westhost.com/biamp-vs-passive.htm (accessed 01-29-2011)
* Newell, Philip, and Keith Holland. Loudspeakers. Oxford: Focal Press, 2006, Kindle Edition.
* “Audio Speaker Enclosures Construction tips,” last modified January, 2011, http://www.kbapps.com/construction.html
* John Albright, April 25, 2003 (12:05 p.m.), comment on Lancestrom, “volts per meter VS watt per meter question,” The klipsch forums, January, 2011, http://community.klipsch.com/forums/p/25374/210918.aspx
* Katz, Bob. "Level Practices (Part 2) (Includes the K-System)." Digital Domain. 01 September 2000
* Plummer, Christopher Personal Interview. 11 January 2011
* Green, Grant. "Frequency." Contrabass. Available from http://www.contrabass.com/pages/frequency.html
. Internet; accessed 20 January 2011.
* Noakes, Cameron. "Diy Audio Corner." Enclosure Dilemma: Ported vs Sealed. Available from http://diyaudiocorner.tripod.com/dilemma.htm
. Internet; accessed 20 January 2011.
* Dickason, Vance. Loudspeaker Design Cookbook. Peterborough, New Hampshire: Old Colonial
Sound Lab, 2006.
* Ludwig Sr., Arthur. "Silcom." Loudspeaker Construction. Available from http://www.silcom.com/~aludwig/Loudspeaker_construction.html
. Internet; accessed 20 January 2011.
* North , Creek. Cabinet Handbook. Old Forge New York: North Creek, 1992.
Design Proposal 33
* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project