MARCH 2004
APRIL 2004
MAY 2004



Sponsored by




Become a Donor
 to Project May








February 16 -
Mr. Widget

The purpose of this entry is to distill the ideas from Project May team discussions regarding the project requirements. In particular, it focuses on concepts that would make best use of the donated components that we will be using in the Project May speakers in conjunction with the design goals put forth by Don McRitchie. The Electro-Acoustic Design and Industrial (Aesthetic) Design will undoubtedly affect each other with the Electro-Acoustic Design taking the initial lead. This does not mean that general design solutions,  materials, fabrication techniques, and the like can not be discussed  before the conclusion of rough prototype construction and testing.

Currently there are three Electro-Acoustic Designs that are being  considered. All three feature the same components. Twin 1500AL woofers  and a 435Be compression driver on a custom horn that incorporates the  045Be UHF compression driver. (The final external design and material of  these horns is yet to be determined.) I will be machining them using  JBL's geometry from the K2-S9800. This will allow us the flexibility of  adjusting the external surfaces for ease of assembly within the woofer  module or modules and making modifications for aesthetic purposes as  dictated by the industrial design.

The three Electro-Acoustic Design directions are as follows:

The first choice is an MTM design with both woofers working in parallel  up to the crossover point. We will need to play with the woofer spacing  to try to minimize the inevitable comb filter effects that might create
undesirable sonic anomalies. Most successful MTM designs use 6.5" or smaller woofers crossed over above 2KHz. This places the two woofers very close together. The large woofer diameter of the 1500AL as well as  the significantly lower 750-800Hz crossover frequency require that our woofers be spaced further apart. There are numerous successful examples of tower designs with widely spaced woofers, but it will certainly give the Electro-Acoustic Design team a bit of a challenge.

The second choice is a tower layout that is visually identical to the MTM designs but electrically one woofer is used only as a helper woofer below 100Hz or so in a fashion similar to the JBL 4435.

The third choice is electrically the same as the last choice but with both woofers placed vertically near each other below the HF units.



February 16 -
Mr. Widget

The following two drawings show the practical range of woofer spacing distances. Using MTM theory we would like the two woofers to be 36" apart based on out crossover frequency. If this proves less than satisfactory we can close the gap but due to the UHF driver the closest we can push the woofers is 28.25" center to center. Both drawings make the assumption that we would like to maintain the center of the mid horn near Don's seated ear level of approximately 40" from the floor.


February 16 -
Mr. Widget

This drawing shows two potential top views for the system. The mid horn will be about 17.5" wide. Based on Don's requirements for minimal diffraction effects the baffle is extended beyond the horn either with 45 degree chamfers or large radii curves.

All of these drawings are to be considered as basic schematics. These are not designs, rather they are "package drawings" showing the locations or proposed locations of the basic components. Additionally each woofer will need a 4.5 to 5 cu. ft. enclosure. With this information I hope that the members of the team will bring forth creative design solutions toward a final design for a pair of speakers that are acoustically as near ideal as possible, reasonable to produce, aesthetically beautiful, and able to be shipped to Winnipeg.

I hope this is enough information to get the Design and Fabrication Team started and aid the Electro-Acoustic Design team in determining some of their dimensions for the initial prototype.


February 18 -

The following are simulations of an MTM response for various Project May configurations:

MTM model, 2 15" woofers, 36" center-to-center spacing, vertical response graph

15" MTM model, 28.25" center-to-center spacing, vertical response graph

MTM plot, 28.25 inch spacing, 800Hz LR 4th order low pass xover, summed response on design axis of 39 inches.

Vertical off axis plot of drivers and crossover mentioned in plot directly above.


February 18 -

And here are some TMM response simulations for various Project May configurations:

15" TMM design, woofer centers 11" and 26.25" below midrange horn, drivers wired in parallel, vertical response graph.

TMM using 2 15" drivers. This graph shows the summed response with a 2nd order Butterworth crossover at 150hz with the polarity on the bottom driver reversed. The top driver has standard polarity and has a low pass 4th order LR network at 800hz.

Here is the vertical off axis plot for the drivers and crossover in the above image.


February 19 -

The following is the list of equipment that I am using for the prototype simulations and measurements.


SoundEasy v8.0 http://www.interdomain.net.au/~bodzio/ $225, updates are $25
There is a link to a valuable user guide towards the bottom of the home page written by John Kreskovsky.
User group at Yahoo http://groups.yahoo.com/group/SoundEasyuserslist/

LspCAD Standard, build 5.25 http://www.ijdata.com/ $129, updates have been free, but the designer is talking about a major upgrade soon, price unknown. LspCAD pro has additional features for $495, the most useful being the crossover emulator. Here is a design walkthrough http://www.ijdata.com/ugly_duckling.pdf

Both programs are very powerful, but LspCAD still seems to have a faster learning curve versus SoundEasy. For the features provided, SoundEasy is competitive with LspCAD Pro, and is currently a better bargain for the price. Both programs should provide comparable results for the final crossover.

One of the most important features of any good CAD program is the optimizer, which allows the program to adjust component values to meet a specific acoustic response target. This is something you just cannot do quickly with cookbook formulas and a hand calculator.

UniBox is a very useful box design shareware progam written by Kristian Ougaard. I highly recommend this free download. Get this for yourself and save a lot of time answering box tuning questions http://www.danbbs.dk/~ko/ubmodel.htm


Computer: Home made Windows XP system with Celeron 1700 and 256MB Ram with an ASUS motherboard.

Sound Cards: Creative SoundBlaster 5.1, soon to be replaced by card below.
M-Audio Delta 410, $113-200 on clearance, 4 input, 8 output sound card, recently discontinued by the manufacturer. I chose this card because it has been shown to work well with both of my CAD programs. M-Audio still offers more expensive cards like the Delta 1010LT and Delta 66 with PCI interface but they are in the $400 price range. The Delta 410 can still be found at some audio stores and on ebay. Since it is a discontinued model, the company may stop upgrading the program files.

If you have a powerful laptop, and can afford it, an alternative that would allow building a completely portable measurement system would be the Firewire 410 for $400. I would like to try this option but do not have this kind of discretionary funds.

Preamp: Bamberg Lab Microphone Pre-amp Probe, $110 fully assembled and tested. This is a microphone preamp with a power supply for the Panasonic mic capsules and it includes probes for testing purposes http://www.bamberglab.com/BESL_MPP.htm

Microphones: Panasonic capsule microphone with calibration file provided by Kim Giardin $32
Capsules similar to these are used in the Liberty instruments Mitey Mikes.
Contact kmgrdn@luminet.net
Uncalibrated capsules can be purchased from http://www.digikey.com
There are many other sources for preamplifiers and microphones, the best place to start your search is always JPO’s links list http://members.chello.se/jpo/

Amplifiers: Adcom GFA-555 and GFA-2535. Discontinued Adcom models, these were purchased on ebay.

Assorted connection cables, probe wires, and batteries were purchased at my local Radio Shack.


February 23 -

I have the cabinets set up with detachable front baffles but none of the new equipment I ordered has arrived yet.

The Delta 410 is now back ordered until 3/8. The test jig and a test jig/mic preamp, were shipped on Friday and should arrive this week.

I posted a model of a 3-way crossover using the S9800 components. Someone mentioned that we could have system sensitivity as high as 99dB/1Watt/1M, but it is clear that this isn't going to happen. The system sensitivity is determined by the lowest sensitivity driver, in this case it appears to be the 045Be supertweeter.

All I can do right now are some amplitude measures but I'm concerned about the weight of the woofers after talking to Steve. He said the 1500AL's weigh close to 40lbs each, so I think I may need to laminate another 3/4" sheet onto my front baffles to support all the weight. I certainly don't want one of Don's woofers to tear itself out of the cabinet and bend the frame, or worse. Is there a trick to mounting woofers without marring the frame with screw marks? I'm using 2 1/2 screws and t-nuts.

I need to clear the memory card on my digital camera and start taking some pictures as well so we can have documentation at every stage. I'm really impressed with the Kreg pocket hole jig. Using the screws and glue in one step, rather than biscuits and clamps to glue every joint is saving me immense amounts of time as I can glue multiple joint at one time now rather than only 2 panels at a time.

I will try to build a baffle to put between the 2 bass cabinets today or tomorrow so that I can try out various horns in the gap between them. I will make it big enough to accommodate my 2380 horn, 11" x 18", so that I can try out some 2" compression drivers as well as my 1" drivers. It looks like I'm going to have amplitude measurements before I get the TS measurements on these drivers. I really want to wait until my jig arrives before starting on TS measures. I'll try to post some pics and measurements as soon as possible to stir up some discussion.

The woofers are going to arrive Tuesday morning at my office so things are going to start happening fast now. I sent a request for funds to buy MDF last week, to an address Ian sent me, and haven't received an acknowledgement of my request. Do you know anything about how the fund drive is going or should I contact Mike Baker? The cabinets I have will work for testing a mono speaker MTM but not so well for the TMM array. Since my wife is still unemployed I can't afford to spend another $100 on MDF.


February 26 -
Mr. Widget

The mid horn prototype is being manufactured using a CNC machine. The one we have is a 3 axis servo controlled router. What this means is that when we program it, we can control x, y, and z simultaneously. The result is we can sculpt any shape as long as there aren't any under cuts. (The motor and cutter only goes straight up and down. It won't twist or cut from under the part.) The tools (mill bits) we use come in a variety of shapes and sizes, but the most commonly used ones are either flat bottomed (Flat End Mill) or rounded (Ball End Mill). We use the flat ones to cut out a shape or to create a flat surface and we use the ball end mills to machine a curved surface.

Here is a front view of the horn. The red line indicates where we will have to make the horn in two halves and glue them together. In the 2397 horns we cut the top and bottom as separate parts and put them together with the side walls. Since there are curved surfaces that blend the horn top and bottom surface with the sides (not shown in this early model) we will most likely want to machine those surfaces with the sides at one time.

This is a top view of one half. The blue area is the part that we can make in any shape desirable for ease of manufacture and assembly as well as blending into the baffle. Note the right side is different from the left showing one possible solution.

Here is a wire frame of the lower half of the horn. The green line shows where the surface blend between the face and side walls will be.

A side view.