The risk with tarketing the commodity market is that there is a very hefty upfront investment in the production tooling to get the cost down. This is a very high risk market since most customers in this category is not in pursue of high fidelity. If someone is willing to put down an order of 10,000 or more for annually for the next 5 years, I'm sure Dr. Geddes would be happy to get things rolling.
I know what you are saying but I wonder again whether anywhere near this number is necessary. How many wave guides have been sold to date?
If the waveguides are as good as they are reported to be, and I have no doubt they are, then there should be a much larger demand through the forums and in the manufacturing - pro field if the cost was more amenable to most people surely.
In my opinion the whole way the discussion of the waveguides, how they perform, how they are costed, is really counter productive to selling them in larger numbers.
jamikl
If the waveguides are as good as they are reported to be, and I have no doubt they are, then there should be a much larger demand through the forums and in the manufacturing - pro field if the cost was more amenable to most people surely.
In my opinion the whole way the discussion of the waveguides, how they perform, how they are costed, is really counter productive to selling them in larger numbers.
jamikl
jamikl said:
soongsc said:
Earl's technology presently has issues which preclude it from garnering acceptance to a broader customer base.
There are also other available technologies incorporating his fundamental teachings which are not burdened with these problems, but nobody has figured out how to successfully market them yet, or even determined that it would be a worthwhile pursuit, for that matter....
This forum plus any other forum really represent a very small market. You have significant amount of duplication of people in these forums as well. It's not a matter of whether the design is good or not, it's a matter that people are not looking for something better in audio.jamikl said:
Look at how JBL enters the low price market and it's not hard to figure out how they view this market sector. As long as there are growing variety of things to buy, each sector will have more difficulty in addressing price/performance issues.ZilchLab said:
jamikl said:
Jamikl
The positioning of a business model is a very difficult thing. My designs could never be "main stream" due to the size. This factor alone seriously limits the market appeal. Smaller doesn't work because the high directivity requires a larger size to work properly.
IF the volume were there then yes, the costs for tooling etc. might be reasonable, but from what I see the volume is not there. This necessitates a low volume market approach and a high piece cost and end user price. Its a classic chicken and egg thing. Soongsc has it quite correct that a major market penetration approach would be very costly and very risky, which is not in the cards right now.
And in the end, the product to the consumer would end up being about the same price as it is now. Thats because, while the higher volumes could easily cut the production costs in half, the marketing and distribution overhead would double it back up to where it is now - nothing changes for the consumer, except, of course, in a business model like that kits would not be availble as they make no sense in a volume business.
fb said:Anybody have any experience modelling waveguides and getting them CNC milled into wood? I'm thinking multiple sheets of plywood stuck together for the depth.
I've seen the oblate spheroidal equation, how would one go about putting this through a CAD program? I've got sketchup....
How about a non-axisymmetric profile? Like the 18sound XT1086 but bigger to match bigger woofer...
Ta
Here's how I do it:
#1 - Draw the curves in 2D. I use Xara Xtreme, but any 2D program would work, even Open Office.
#2 - Then I slice the part into segments. For instance, a 12" deep waveguide is sliced into 12 pieces that are each 1" thick.
#3 - I draw the pieces in 2D, and print them.
#4 - I staple them to foam, or glue them to wood, and put them back together.
See this post here for an example:
http://www.diymobileaudio.com/forum...creating-perfect-soundstage-3.html#post758009
doug20 said:
Now days, the CNC machines accept AutoCAD files. It is done using an import/conversion process, but functionally, you can pass the AudioCAD (or Rhino or whatever) DXF file into the CNC software. You usually have to do some massaging, like path the tools, determine cutting sides (flip operations) and maybe set the cutting heads, stuff like that. But it's a whole lot easier to do than it was 20 years ago.
I would probably do a device like John's as two separate blocks that were glued toghether and finished. I've done several as four sections. This makes extra details important, because you have to make room for hold-downs that are later cut off. You have to hold the block in place in some way, and it is smart to path the cutting heads where they do the most good. But by separating the horn into sections, you can cut it on just about any good CNC router table.
I've worked pretty closely with my CNC shop, and built many horns over the years. I've built folded basshorns, midhorns and tweeter horns. So I'm pretty familiar with what's involved. But I would still look to the advice of someone that works with CNC machines every day. All these tiny little details, like how to hold the block in place while cutting it, those are the details that are easiest to overlook but probably the most important in the end.
To answer a few questions, there is no bump or dimple on top of the waveguide. There is a flat spot on the exterior, or a portion where it only curves in one direction, so that is a different distance from the edge of the waveguide, and the large radius that blends the waveguide into the outside of the cabinet changes shape there a bit but it's still tangent to everything and smooth. I could make the radius smaller there to avoid that, but I chose not to. Here's a cross section of the horn on top of a cabinet:
All the CNC machine controllers I've worked with (which is not a ton, but more than one) require G code input or were conversational controls, meaning you input geometry on the machine itself. G-code is a big text file containing mostly positions that the machine moves to with additional commands to start and stop the spindle, etc. To go from a CAD file such as I have created to G-code, you need a CAM program (computer aided machining) or if it's a simple shape, you can write the G code by hand. A CAM program takes a geometry file from a CAD programs and figures out the path the cutter should take to produce that shape based on some (or a lot of) user input. To make a 3D shape like this, you need a more sophisticated CAM program than making a simple baffle, for example. I've never seen a machine controller that you can just load a cad file directly into, but that doesn't mean they're not out there. But for something like this, that is not the normal method.
An externally hosted image should be here but it was not working when we last tested it.
All the CNC machine controllers I've worked with (which is not a ton, but more than one) require G code input or were conversational controls, meaning you input geometry on the machine itself. G-code is a big text file containing mostly positions that the machine moves to with additional commands to start and stop the spindle, etc. To go from a CAD file such as I have created to G-code, you need a CAM program (computer aided machining) or if it's a simple shape, you can write the G code by hand. A CAM program takes a geometry file from a CAD programs and figures out the path the cutter should take to produce that shape based on some (or a lot of) user input. To make a 3D shape like this, you need a more sophisticated CAM program than making a simple baffle, for example. I've never seen a machine controller that you can just load a cad file directly into, but that doesn't mean they're not out there. But for something like this, that is not the normal method.
A simple CAM program will not work with 3D parts like this. Also, I've never seen a machine that comes with CAM software, but once again I haven't seen them all. Most high end CAM software is a personal decision, much like a CAD program. I spent a lot of time evaluating software before I bought mine.
John Sheerin said:
The two tables my guys work with, one a Shopbot and the other a Thermwood, both have conversion software that goes from DXF to G-Code. It will do a "best fit" for pathing but you sometimes (usually) should modify this, so there is some manual intervention/optimization required. I believe the interface/conversion software is bundled with the Shopbot, but I think the Thermwood requires a third party conversion package.
From looking at their webpage, the Shopbots have the option of adding Vetric CAM software when you buy the machine. That's what does the conversion of cad file to g-code. That's one of the software packages that I looked at, but it was not sophisticated to do what I wanted. Then there's a machine controller that runs the machine using the G code. I'm not sure what Shopbot uses for that.
For CAD I'm using Pro/E for complicated stuff (like elliptical waveguides), and AutoCAD for simple stuff. I'm using Visual Mill for CAM. The machine controller isn't really important as long as your CAM program has a post processor for it, but I use Mach3.
I am planning to make CNC waveguides (and horns, and cabinets, and whatever else I feel like). I've been planning on it for about the last year, which is about how long it took me to research, design, and build my own CNC router. I've been doing aluminum bits for my motorcycle with it for the last few months, but I'm switching the machine back to wood at the moment. I'm just finishing up my dust collector before I start so I don't make a (bigger) mess of my garage. My first horn is going to be a 320Hz LeCleac'h for a JBL 2435 which I already have programmed (see below). This elliptical waveguide will be after a midbass horn which will hopefully give me time to get it finished and programmed, but I'll probably do the woofer cabinet first and set the LeCleac'h horn on top of that for a while until I get some other horns designed.
2435 horn:
Some recent projects:
I am planning to make CNC waveguides (and horns, and cabinets, and whatever else I feel like). I've been planning on it for about the last year, which is about how long it took me to research, design, and build my own CNC router. I've been doing aluminum bits for my motorcycle with it for the last few months, but I'm switching the machine back to wood at the moment. I'm just finishing up my dust collector before I start so I don't make a (bigger) mess of my garage. My first horn is going to be a 320Hz LeCleac'h for a JBL 2435 which I already have programmed (see below). This elliptical waveguide will be after a midbass horn which will hopefully give me time to get it finished and programmed, but I'll probably do the woofer cabinet first and set the LeCleac'h horn on top of that for a while until I get some other horns designed.
2435 horn:
An externally hosted image should be here but it was not working when we last tested it.
Some recent projects:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.