We have all of our aluminum cones from the AV series spun by these people. http://www.superiormetalspinning.com/ Once we worked out the bugs and they realized the tolerances we needed to hold, things went fairly smoothly. For a round waveguide this can work quite well. You cut the block once and then can spin the alum easily from there. Back fill with polyurethane foam of appropriate density and you have a nice waveguide. The limitation is it is only possible for round waveguides.
For any kind of ellipse obviously this doesn't work. Tooling is expensive as Earl mentions. Injection molding would give the lowest per piece cost but tooling would be ridiculously expensive. You need t have both negative and positive molds made to incredibly tight tolerances. A deep drawn thermoformed part could be an option. This would require just the male part to the mold. This would be less expensive, but still several thousand $ for the tooling. Molds for fiberglass would work well as again you only need one side of the mold. The issue is always just initial cost. We have many things like this that could cost only a few thousand $ to get into production, but it's difficult to have the extra money to put out when it doesn't give an instant return on the investment.
John
For any kind of ellipse obviously this doesn't work. Tooling is expensive as Earl mentions. Injection molding would give the lowest per piece cost but tooling would be ridiculously expensive. You need t have both negative and positive molds made to incredibly tight tolerances. A deep drawn thermoformed part could be an option. This would require just the male part to the mold. This would be less expensive, but still several thousand $ for the tooling. Molds for fiberglass would work well as again you only need one side of the mold. The issue is always just initial cost. We have many things like this that could cost only a few thousand $ to get into production, but it's difficult to have the extra money to put out when it doesn't give an instant return on the investment.
John
gedlee said:
I appreciate the discussion on what it would actually costs but are these comments needed??? It reads like you are the only businessman on here and we are just foolish hobbiests
btw, I have run my own business succesfully for almost 20 years, thank you! I guess I run it very well because several thousand isnt much to risk on trying something....now John properly pointed out that this is all much more then a several thousand...you didnt explain it well at all....just keep posting rude replies. the real return...hobiest vs businessmen insult was not needed...you need to realize some of us are more succesful then you might think.
I guess we shall see over the next 5 years how good of a business man you are at selling your speakers...then maybe we can judge you too?
doug20 said:
I appreciate the discussion on what it would actually costs but are these comments needed??? It reads like you are the only businessman on here and we are just foolish hobbiests
btw, I have run my own business succesfully for almost 20 years, thank you! I guess I run it very well because several thousand isnt much to risk on trying something....now John properly pointed out that this is all much more then a several thousand...you didnt explain it well at all....just keep posting rude replies. the real return...hobiest vs businessmen insult was not needed...you need to realize some of us are more succesful then you might think.
I guess we shall see over the next 5 years how good of a business man you are at selling your speakers...then maybe we can judge you too?
Wow!! Nothing rude or insulting was intended and when I read the comment I don't see anything rude in it at all. All I meant was that AS businessmen people like John and I need to see a resonable return on our investments, which is not true of the hobbiest. If several thousand is nothing to you thats great, I'm envious! But its a lot of money to me and I don;t spend it lightly on a gamble. But the point here is that I DID!
I really don't understand your sensitivity to my comments. You act like they are personal slams to you, and as anyone will tell you, I don't do that. I was taken aback by the comment about several thousand being a small investment and thats all. I don't charge 10's of thousands for my speakers, my profit margins are quite reasonable and it takes a lot of sales to recoup several thousand dollars in investment.
thats fair....it did seem like an insult.
Calling me out as an hobbiest and you a businesman...meaning you somehow know better about business in general is an extreme reach (IMO). Combine that with "You do not seem to have a handle on the loudspeaker business" makes it fairly consistent to me.
I will take your word for it not being an insult. I would rather just keep on topic of what it truely takes and leave the color commentary for lessor forums.
I think the whole sevaral thousand discussion, is overblown. Several thousand is 3-5 thousand dollars to me and if that is what it takes to setup a CNC to build waveguides then it might be a very reasonable investment for some people. You indicated that its more of a cost per unit, I guess instead of a one time setup cost. All that was needed was clarification. Again, 5K for R&D when someone own a business is very normal in my world. I have seen $50k prototypes go nowhere in successful businesses.
Calling me out as an hobbiest and you a businesman...meaning you somehow know better about business in general is an extreme reach (IMO). Combine that with "You do not seem to have a handle on the loudspeaker business" makes it fairly consistent to me.
I will take your word for it not being an insult. I would rather just keep on topic of what it truely takes and leave the color commentary for lessor forums.
I think the whole sevaral thousand discussion, is overblown. Several thousand is 3-5 thousand dollars to me and if that is what it takes to setup a CNC to build waveguides then it might be a very reasonable investment for some people. You indicated that its more of a cost per unit, I guess instead of a one time setup cost. All that was needed was clarification. Again, 5K for R&D when someone own a business is very normal in my world. I have seen $50k prototypes go nowhere in successful businesses.
doug20 said:
You took the original statement wrong. I meant (and said) that you did not understand the "LOUDSPEAKER" business not "business in general". In the hi-end loudspeaker business the volumes are very very low (miniscule might be an appropriate word) and the markup, in my case, is not very high. Thus it becomes very difficult to justify an expense of several thousand dollars.
My first systems, the Summa, had tooling costs of about $5000, I never really did recoup those costs. I learned my lesson. The next system designs had tooling costs that are a fraction of that. I have recouped those costs. I am going to be very cautious about tooling costs in the future. Thats the point that I was making.
doug20 said:
I didn't notice any insult (intended or otherwise) from Earl. (..and occasionally I do "pipe-up" on the latter.
)He is just generally lamenting the fact that it's difficult to get a return on investment in this hobby/profession. Such posts are not there to either insult others or say you shouldn't DIY; and sure, it's a little depressing - BUT:
On the other hand he, (and confirmed by others), has provided practical information on where significant limitations will dramatically increase costs - to the pro or the diy'er.
Basically if you want something other than a conical shape - it's probably going to cost you a lot more. A lathed/spun conical waveguide however, can be cost effective. So if you are interested in that then (depending on material), your correct - it shouldn't cost many thousands of dollars. In fact as John mentioned - you can significantly reduce costs (for costly material) by: reducing the amount of material (..and "back-filling" with something cheaper). Alternatively, start with a cheaper material like plastic and have that "spun"; not having to worry about reducing material should save some on labor.
If you are into a non-conical shape then things can get a lot more complex (and costly as a result). Usually it's the time on the machine itself. However again, materials that cut more quickly will be cheaper. IF however the source is a full machine shop then they can often cut-away the excess with equipment more suited to that activity at a substantial savings (..in expensive machine time), regardless of material. THEN you can have the expensive machine to do the precision cutting much faster (..and less expensive).
Here is a place that can cater to just about anything:
http://www.emachineshop.com/
To keep costs low, my suggestion is:
Stick to plastics and a conical shape. Use Hornresponse to model your conical horn and then use emachine shop's software to "draft" it. See what the pricing is for various materials, and select the one that seems most appropriate for you.
Does anyone have any experience with emachineshop? I've seen the site several times, but I've been concerned about the process. I'm sure that they would give you what the drawing shows, but sometimes thats not the right thing. Working with a shop usually gets some of these problems worked out.
It took several passes on the elliptical drawings to get them right. There were three people looking at them and even then we missed some things. All this modern computer based stuff works great, once you figure it out, but boy, working out the bugs on someone elses invoicing can get very expensive.
It took several passes on the elliptical drawings to get them right. There were three people looking at them and even then we missed some things. All this modern computer based stuff works great, once you figure it out, but boy, working out the bugs on someone elses invoicing can get very expensive.
If you have a mathematically defined shape such as Earl's waveguide, then a hand made prototype and digitizing is less accurate than modeling it in CAD. The CAD model will be perfect. If you can't draw what you want, then you just find someone who can. The digitized model will only be as good as you make the model plus the error introduced by the digitizing method. The process from there is the same if you assume the digitizing process doesn't require any cleanup in CAD, which it likely will in my experience.
gedlee said:
Yup, good questions (with unfortunately no solid answers - I've not used them).
1. Will the free download program be sufficient to the task?
2. What are contract tolerances?
3. (Assuming you accept those tolerances) - If they fail to meet those tolerances, what's your recourse?
The good part about this is that it will only cost you some time to figure 1-3. Of course the bad part is that it will cost you your time.
I went looking for negative reviews from them before, and really haven't come up with any. Even "support" seems to be decent.. (in this "day and age") go figure?
if you have a free 3-axis cnc machine with enough Z just setting around then you might be able to entertian yourself with "Incremental Single Point Forming" - but it still seems to be more of a subject for active research rather than a commercially available prototyping technology
it appears to me after an hr or so of internet search that forming a elliptical mouth OS waveguide would be a fine candiate shape for the multipass method (to make the near vertical throat profile)
http://www.qmisolutions.com.au/article.asp?aid=223
some papers:
http://esaform2008.insa-lyon.fr/proceedings/MS17/p000MS17.pdf
and I thought most places were happy to put theses online but the abstract sounds good:
http://www.mek.dtu.dk/Publikationer/Publikationer-2008.aspx?lg=showcommon&id=232136
it appears to me after an hr or so of internet search that forming a elliptical mouth OS waveguide would be a fine candiate shape for the multipass method (to make the near vertical throat profile)
http://www.qmisolutions.com.au/article.asp?aid=223
some papers:
http://esaform2008.insa-lyon.fr/proceedings/MS17/p000MS17.pdf
and I thought most places were happy to put theses online but the abstract sounds good:
http://www.mek.dtu.dk/Publikationer/Publikationer-2008.aspx?lg=showcommon&id=232136
ScottG said:
I didn't notice any insult (intended or otherwise) from Earl. (..and occasionally I do "pipe-up" on the latter.
He is just generally lamenting the fact that it's difficult to get a return on investment in this hobby/profession. Such posts are not there to either insult others or say you shouldn't DIY; and sure, it's a little depressing - BUT:
On the other hand he, (and confirmed by others), has provided practical information on where significant limitations will dramatically increase costs - to the pro or the diy'er....
Thats fine, Geddes already posted it wasnt his intent...it still reads as such to me but that is just my opinion and its not really a big deal either and I accept replies that its not.
Im just extremely interested in the discussion on how someone would go about fabricating waveguides, Im not even interested in buiding a business, I have one that makes a better profit in another industry because like you said ROI in the audio business isnt that great for most.
The spun aluminum links/discussion is very interesting IMO.
Doug
If you are dead set on doing this yourself then here is what you want to do.
Make a mold in wood by turning it on a lathe. Then use fiberglass on this mold to make your parts. This is as easy and cheap as it will get in one-offs. Its not cost effective for multiples however.
You might even find a fiberglass house who will do the whole job for you as glass is probably pretty common in FLA. I'm guessing that a pair done this way would be about $700. A lot less if you can turn the mold yourself.
The metal spinning would require the same mold, but has another step required that is not so easy (the metal skin is not strong enough for a direct attachment of the CD to it). This would make it less cost effective on one-offs than the glass approach. The metal has a great appearance which is IMO its only asset.
Good luck.
If you are dead set on doing this yourself then here is what you want to do.
Make a mold in wood by turning it on a lathe. Then use fiberglass on this mold to make your parts. This is as easy and cheap as it will get in one-offs. Its not cost effective for multiples however.
You might even find a fiberglass house who will do the whole job for you as glass is probably pretty common in FLA. I'm guessing that a pair done this way would be about $700. A lot less if you can turn the mold yourself.
The metal spinning would require the same mold, but has another step required that is not so easy (the metal skin is not strong enough for a direct attachment of the CD to it). This would make it less cost effective on one-offs than the glass approach. The metal has a great appearance which is IMO its only asset.
Good luck.
I had a bit of free time today so I knocked out a 90x40 elliptical waveguide model:
This would sit on top of a woofer cabinet with the same curvature. I have a few more projects to finish and then I'll start carving this out.
An externally hosted image should be here but it was not working when we last tested it.
This would sit on top of a woofer cabinet with the same curvature. I have a few more projects to finish and then I'll start carving this out.
I am wondering if part of the problem for Dr. Geddes is seeing the product as "high end" and therefore probably limiting the number of customers he might have. If the gamble was taken to shift the pricing down to a middle of the road market then just maybe the sales volume may be there to allow for the costing of moulds to have the waveguides more cheaply manufactured.
I am not insulting the attittude but just wondering if a change in perspective might make a difference. I am also aware, before it is pointed out, that it is not my money at risk.
jamikl
I am not insulting the attittude but just wondering if a change in perspective might make a difference. I am also aware, before it is pointed out, that it is not my money at risk.
jamikl
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.