Oblate Spheroid Waveguide - Consolidated Construction Thread

[gtforme00]

Recently it has been suggested that there is not enough concise information as to how to actually construct an oblate spheriod (OS) waveguide. While this information does exist in its entirety on this forum, it is very sparsely distributed amongst a large thread. The purpose of this thread is to extract this information from the comprehensive waveguide thread, and provide a concise reference to those who may be interested in creating their own OS waveguide.

If you do not know what an OS waveguide is, why you might want one, or have questions or objections about the usefulness of the OS profile, please refer to the above thread. This thread is for the discussion of construction techniques only (otherwise why bother creating a separate thread?)

Unless you have a lot of free time, available tools, inexpensive material, and sufficient talent, you will find that the price Dr. Geddes charges for his waveguides on his Summa webpage is actually quite reasonable. I can't imagine that designing, constructing, and finishing a pair of waveguides would take less than 8 hours of time and cost less than $100 in material. This does not include the foam which Dr. Geddes places in the waveguide, which is both expensive and difficult to find.


Step 1 - Size the Waveguide

It has been recommended by Dr. Geddes that the minimum effective size of a OS waveguide is 10" in diameter. Below is a quote which describes the benefits of different size waveguides.

Originally posted by gedlee


The waveguide size has more to do with how well it works than the crossover point. A 10" one does not control as well as the 12" which is not as good as the 15". Ideally it should be about 18" - I could make that size work perfectly. The 10" has a lot of flaws, but the size is attractive to most people.

The woofer size determines the frequency at which the two source patterns mate up. As you would expect the 10" woofer mates to the 90° output of the waveguide at a higher frequency than the 12" and 15". The 15" is about 800-900 Hz the 12 about 1 kHz and the 10 about 1500 Hz.

From my data, the ideal would be an 18" waveguide with a 12" woofer. This would elliminate all the flaws that I see in practice. But such a combination would not be an attactive speaker for a lot of reasons. the next best, and what I would build for myself, is a 12" woofer and a 15" waveguide. Simulations indicate this combination to be a very good match.

As of 07/18/2008 the only waveguides offered on Dr. Geddes site are the 10" and 12" models. As a result, the greatest advantage of making your own waveguide is the ability to make a 15" or 18" waveguide.

So pick the largest cabinet size you can get away with, and start from there. Next, you need to select your compression driver.


Step 2. Pick The Compression Driver

The compression driver you choose will affect the throat design of the waveguide. Dr. Geddes recommends a 1" throat diameter driver, due to the decreased high frequency extension of larger compression drivers.

Originally posted by gedlee

All compression drivers have the same situation. The larger the throat the lower its top frequency capability. A 1" driver can just hit 16 kHz, a 1.4" about 10 kHz and a 2" about 8 kHz. In my opinion only the 1" allows for a single HF device and each of the larger diameter throats will need another higher frequency device - a singularly bad idea. The only reason for a larger format driver is higher power (and macho bigger bigger bigger). If these are for home use then higher power is most certainly not required.

Different 1" compression drivers may have different throat exit angles, for example the B&C DE250 used in the Summa series has a 6.5 degree angle at the throat exit, which will ideally be matched at the throat entrance of the waveguide to reduce diffraction.

Dr. Geddes does not recommend thread on style compression drivers, preferring the bolt on flange type.

The pedigree of the compression driver is claimed to be much less important than the system design, and as long as the compression driver emits a plane wave at the throat (has a proper phase plug design), the sound quality between a $100 and $1000 compression driver has been statistically indifferent.

Originally posted by gedlee

We built Summas with TADs and B&C drivers - a 10:1 price difference. Both systems were optimize with different crossovers. We had 16 people audition them - blind, of course - and statistically there was no difference, marginally the B&C were prefered (but it was not statistically significant with only 16 people).

So could one actually go all the way down to a Radio Shack driver (if they made compression drivers) Maybe!! I've never tried.

I will say that I know that I could make a loudspeaker system using a $17 Chinese B&C copy work almost indistinguishable from the TADs. Thats pretty close to Radio Shack pricing. (I say almost because I haven't tried this and don't have the data, but based on my experience and driver tests, I would bet on the results of this test being a statistical wash.)

Dr. Geddes has expressed that the B&C DE250 is a good compromise of quality for cost, and chooses this compression driver in his design.


Step 3. Compute the Waveguide Profile

The OS waveguide is a catenary profile revolved about an axis. Although the mouth appears conical at a glance, it is actually slightly curved along the profile, tending to approach a conical form at the extents.

To avoid diffraction at the mouth exit, a fillet is used to round over the catenary profile to the baffle surface. The radius of this fillet should be as large as feasible. I cannot find the quote but I recall Dr. Geddes using a 2" radius on the roundover.

Originally posted by gedlee

The mouth treatment is not part of the OS equations. The mouth radius should be as large as practicable (a radius larger than 1/4 wavelength at the lowest frequency it is not necessary). So a rule of thumb is to target a radius of 1/4 lamda - which is hard to do.

Dr. Geddes has laid out his recommended design steps in the following quote:

Originally posted by gedlee

It turns out that OS waveguides are quite constrained by the math. Once you have a driver and then the coverage pattern, the rest is fixed. OS waveguides do not have a predicted "cutoff" so the low end tends to be dictated by the compression driver. The coverage angle can only be held down to where the mouth dimension is too small to control it. In the ESP15 (a Summa) this coverage narrows a bit at about 1000 - 2000 Hz. The waveguide is about 16 inches across. In the smaller ESP12 and ESP10 the waveguide is smaller with a notable raising of the lowest frequency of coverage. It is really important in these devices to have a sizable waveguide - too small is a lot of compromises.

So the design procedure is simple. Pick your Comp driver, and then your coverage pattern then the contour becomes:

y(x) = sqrt(throat radius^2 + x^2 Tan(coverage_angle)^2)

x is the distance along the axis. Note that at x = 0 the angle is zero and the radius is "throat_radius".

For extreme accuracy, which appears to be important, one wants the initial angle and radius of the wavegiude to match the exit angle and radius of the driver. This is tricky, but one who is competent at design can work out the correct numbers from the above equation. I did it in MathCAD. I can generate the contour for any radius and exit angle, (I would have to charge a fee for that). Its not an intractable problem however and trial and error on a piece of paper or in a spreadsheet can get you what you need.

Note that larger throats on the compression driver invariably lead to a lower frequency of falloff when the waveguide is true CD. Many drivers show good HF response out to 20 kHz on a plane wave or on a non-CD horn, but when put on a true CD device like an OS waveguide the response dies above 10-12 kHz for a 1.5 " driver and 9-10 kHz for a 2" driver. This is why I have only used 1" drivers. I have not found one that goes out far enough in any larger throat sizes.

There are several spreadsheets and at least one program that offer the ability to calculate the catenary profile.

Spreadsheets

Dennis H "catapult" - This spreadsheet uses the "goal seek" function of Excel to iteratively approach the correct throat angle and diameter. http://www.crestviewcable.com/~catapult/Geddes.xls

John Kreskovsky "John K..." - This spreadsheet agrees closely with the one above, but does not require the use of "goal seek" http://www.musicanddesign.com/codes/OS_wave_guide.zip

Patrick Bateman - hosted at Google Documents for those that need cross-platform compatibility http://spreadsheets.google.com/ccc?k...LDVHlIIw&hl=en A thread describing the construction technique of this contributor is at HT Guide

Programs

Hornresp - This program has the ability to calculate and simulate the acoustic response of an OS waveguide. http://mywebsite.bigpond.com/dmcbean/

Once you have created the profile, you have to decide how best to translate that profile into reality. There will likely be a different method for each different person that attempts this feat, but suggestions and experiences are welcome.

Good Luck,
David Malphurs

[Variac]

This thread is an excellent idea and you have already done a lot of consolidation!

[BHTX]

..x2. Thanks.

[nullspace]

Don't forget JoshK's spreadsheet calculator.

Regards,
John

[gtforme00]

Thanks for that link, I remembered that JoshK did a spreadsheet, but didn't see that he updated it to include the throat angle. An added benefit of JoshK's spreadsheet is that it allows you to input a round-over radius and it plots the resulting fillet to the flat baffle.

Thanks!
David

Post Script

I would be remiss if I didn't take a moment to thank Dr. Geddes for contributing the information he has to this forum, the main reason I felt compelled to create this thread was to relieve some of the pressure (unfairly) being placed on him to consolidate this information into a how-to guide which spells out every step of the process. DIY!

[LineSource]

In the DIY horn community the possibility of finding the Holy Grail is often the only way to justify spending time and money. From the attached information it appears that an 18" waveguide promises the Holy Grail in high efficiency 2-way design.

One possible Consolidated Construction Strawman could be an 18" waveguide crossed at 800Hz to a state of the art 15" woofer which could cover 25Hz to 18kHz.

The AE-Lambda team is promising a few new 15” woofers with 99db/watt underhung motors, and these seem worth watching. The attached graph for a production 93db/watt Lambda 15S in an 11 ft3 ported cabinet – Altec Nineteen volume

A plastics expert joining this thread would be very helpful!!

[jamikl]

Castable polyurethane may be of some interest here. I know that is is used in other hobbies/crafts and in the theatre effects area. Maybe there is a member reading this thread who can give some information at to suitability and techniques.
jamikl

[kstrain]

If this is to be the excellent reference it promises to be, it might help if a moderator could (with permission) fix the typo in the title (spheriod - spheroid).

Its already no. 3 on google under spheriod

Ken

Edit - I can't type either

[MaVo]

Here is a googled faq about castable polyurethane: http://www.elastech.com/service/polyfaq.htm

Another option could be epoxy: http://en.wikipedia.org/wiki/Epoxy used with various filler materials like fiberglass.

With all attempts at crafting a waveguide, it seems to be difficult to achieve a good accuracy while transferring the numbers into the physical object. I think Dr Geddes said that <1mm accuracy is of great importance. With professional equipment like a cnc lathe, this is no problem. But at home we need to find a good method of ensuring this accuracy. The object we craft is too complex to use a ruler and measuring by eye will be even worse.

@ gtforme00: Great work! I recently read my way through the same thread and i think you got the most important information. Now we need to figure out all the little details that will make the diy of a waveguide so complicated.

[MaVo]

@ Linesource: 18 inch waveguide indeed seems to be optimal, regarding the directivity matching of woofer and tweeter. Below is a hornresp sim of the directivity of a 18inch wg and a 15 inch woofer. They perfectly overlap around the crossover region.

We also see the problem which conical / os waveguides have, a slight narrowing followed by a widening of directivity around the lower cutoff of the wg. I dont know how Dr Geddes works around this problem.