Hey guys...we need a little rallying here... - Page 20

Yes there is. Its the variation of path length from throat to mouth that mitigates the on axis dip. Geddes himself said it was the constant throat to mouth length in his axis symmetric designs that was the culprit for the on axis dip. We have seen data on the xt1086 to show it has less on axis dip, the QSC 152i has even less. Thus it isn't a stretch to infer that a SEOS profile should have less than elliptical but maybe a titch more than rectangular, but at the same time avoid the sharp HOM producing corners.

You are spreading out the frequency for which you get mouth reflection induced cancelation. Just like speaker design, horn/waveguide design is a delicate balance of tradeoffs. I enumerated why I think the SEOS *might* very well be beneficial. If it is too hard to produce vs. an EOS, I am fine with EOS, but if we are going to the trouble why not make the extra effort?

Does AVS have a wiki page? It might be easier to set up a wiki and track people's interest that way.

I said I'd be interested in a 15" wide EOS or SEOS waveguide and I still am. I have 5 QSC's already so I am in no need for more. I am more in Penn's camp of wanting to go a step further than what is cheap and easy to get. I'll put my money where my mouth is as long as the price is kept within reason (say <$500/pr shipped).

I don't want to keep Erich or Jack guessing. I don't know who was doing what at this point. I know Jack was able to produce the CAD quickly, but I still don't have an idea of cost including how we are going to get them from there to here (US). That was the issue in the diyaudio thread too.

What would be really awesome is to be able to produce a single prototype of a couple of different profiles and measure them. However, that seems to be too cumbersome, so we are having to make educated guesses.

I am ok with this taking a while if the results are worth the effort. I have plenty of irons in the fire already.

Josh, you have followed the group buy as long as I have. The other new issue is how people can pay. It seems paypal isnt a good option from the US Im very cautious with bank transfers, they offer zero protection!!

BTW, I don't consider any of this discussion a debate. I consider it doing our homework before pulling the trigger. If you want to produce an excellent product you have to do your R&D. Multiple prototypes is impractical so educated theorizing is in order.

Penn, yeah and the organizing of redistribution from a port to multiple locations seems to be a tough one too. I don't think it is going to be any different for this batch. I am greatly appreciative of Jack's help and I think he is able to produce great results from pics I've seen but there is still a few road blocks that keep it from being a reality thus far.

P.S.S. if it turns out that those who want the EOS/SEOS WG want a bigger one, say 18" wide, I am ok with that too. I am just compromising because we have to have some consensus.

Im in for that 18" batch. From Jzagaga email it sounds like he could make some in a week or two actually.

He asked me if I wanted some along with my OSWG-12 order.

Its funny you post that. This is how my company and other companies get things worked out. A bunch of engineers (Im the software guy) sit in a board room day after day debating left and right now to build an automated stacker system ....this discussion is no different then my day to day working life so I guess Im kind of use to how this sort of prototyping discussion works. Honestly the final product is never final until its working


yeah, recently someone in the US said they could be the point of contact. I would have went to a port in Jacksonville to pick them up and I know EricH has had contact with them for a while. We have spun our wheels for over a year on this stuff actually.

Not to throw in more confusion, but another already existing product, way outside of this box that we can use as a reference is the Beyma TPL-150H. It has a directivity of 80º x 30º, high sensitivity and low distortion (if the manuf curves are to be believed). The CTC shouldn't be an issue with this.

However, the rub seems to be that it is CD in the horizontal plane but increases directivity in the vertical (starts to beam). They are pretty expensive too. Still something to keep as a ruler if we get really wacky here.

So we have two options:

1. Jack in Poland makes them in batches and finds someone to distribute them in the US

2. Jack makes a single prototype and sells it someone in the US (Erich?) who can then make a mold and sell them to individuals

There are advantages to option 2 because it would allow for a consistent source of the WGs. That makes it more reasonable for someone like Zilch to come up with a design that non-designers can build from. If we have to deal with a pallet of WGs and a middleman in the US I doubt we'd get more than one order out of it.

YO, Penn!

I mentioned this earlier -- as consultant, Geddes early on designed an EOS waveguide for what I recall to be a Canadian manufacturer, which uses it in product. The reference appears in one of the diyAudio threads. Since you are familiar with much of that info, please track down that resource and let's get ahold of one or more of these for evaluation....

For now I think #2 is what we should do. We should let EricH figure out what the best appoarch is.

Did your PE contact give any idea of an ETA for these?
I'm in for a quad straight off.

Eric,

I was not suggesting this would happen in a week. I wrote "ready to drop a few thousand to get this thing going by the end of the week. Nothing about who, what, where, you know all the in-between stuff. " based on your comment the day before.



Your comments suggest you're ready to go this week, and your willingness to foot the bill for development is commendable. However you have not been forthcoming with details of the plan i.e. who is going to make the molds, where they'll be cast, what material they'll be made of, # of units per production run, etc.. These variables will effect cost/unit = # of potential customers, as well as long term accessibility among other things. With this knowledge we can determine the number of buyers we need to make the best product at the cheapest price, with the least amount of hassle.

Also what is your plan for the "prototype", is this going to be a simple CNC'd male buck that can be glassed for a one-off to test, or a full mold. If we go the 1st route it may be feasible to have both the EOS and SEOS proto'd. Thereby allowing measurements to determine the difference rather than making an educated guess.


Back to design decisions.

If you're looking to get this rolling ASAP as you wrote, I vote we stick with the 15", 1" throat, matched to DE250 or better, aluminum or cast CD mounting flange, matching elliptical outer flange = not rectangular.

Do you know Jzagaja's prototyping process. I'm just curious if he will end up making a mold to make a prototype. If so maybe purchasing the mold from him, and having them cast stateside will be the best solution. This would ensure the mold is proper since he has experience in this realm, and also possibly net him more $$ for his help.

http://www.dv2.fr/index.php/news/newsFull/idNews/25

Discussion begins at post# 1403 here:

http://www.diyaudio.com/forums/multi...ml#post1882184

I haven't seen the Geddes consulted EOS either. There is the Coda Audio G512 which looks like a EOS to me and the marketing blurb sounds like it comes straight from Geddes mouth:

The flange is still something Im not sure everyone agrees on.

Maybe someone can generate some quick pictures to help us decide. We should assume the SEOS is the design shape. I think we need the waveguide as close to the edge as possible so Im requesting the flange be a minimum thickness. I would even recommend that only the corners matter for flush mounting to a baffle (Each Corner has a screw hole, 4 in total). The roundover on the baffles (I use 1 1/4") can be an extension of the waveguide mouth atleast on the sides and top. It just seems that a continous curve is better.

coctostan just posted (post above) as I was posting this, that waveguide flange is what Im trying to discribe, very little flange on the sides.

In the end what ever design is done, Im thinking EricH can build a nice baffle around the design making it all one piece I can send him the prototype baffle with roundover for a rear mounted woofer too....all roundovers will be 1 1/4". If I find time I will be sending my QSC HPR-152i baffle to him so you can make a nice clean baffle, no screw holes, no PA looks at all.

Because of cost Autotech can make one prototype. Can we agree on type before Christmas?

These are the numbers so far that I think everyone agrees on.

- SEOS-15 90x45
- 1" bolt on
- throat geometry to mate to de250 driver

Yet another ultimate possibility JMLC-EOS-21.

I vote for an elliptical close trimmed flange (# 1/2) since I plan to use either a hardwood or veneered baffle, I don't want any extra plastic that the rectangle introduces. That and it would look better if somebody wanted to use it free standing.

The #4 pic is the SEOS (#5) profile stretched vertically since the ones Jack posted seem to have a much smaller transition radius and thus are much shorter than the original EOS that had a large radius

Jzagaja,

Since you have the most experience modeling, building and listening to different horns. What is your opinion for the best design given the following limiting parameters:

-Max 15" wide
-90 x 45/50 deg directivity
-1Khz or less XO while maintaining pattern control
-1" Throat

From what you've modeled which profile and transition radius will best accomplish the goals of even coverage, no on-axis null, no harsh vertical nulls or directivity transitions, least amount of HOM's.

EOS, SEOS, JMLC???? Flare Radius????

Thanks!

Zilch especially likes the #4 SEOS aspect ratio. It could be made with close-trmmed flange, as well, looks like....

Guys, just to lessen your worries, if all goes well, paying, reproducing, and getting them won't be much of a problem. This project is much farther along in those aspects. The only thing I needed to know was exactly what people wanted built.

Have no fear on actually getting the product once that's figured out, I've been working on that portion for a while now. I don't want to go into too many details about that until I'm 100% sure on a few things.

Let's not put too many irons in the fire, or have too many people trying to do the same thing. The production stuff in progress.......this will happen.

What we need is for the smartest WG people here to weigh in on SEOS or EOS, and whatever else needs to go into that design. The person is waiting to go once the word is given.

Zilch, can you go into what details you would like to see with that #4? What exact aspect ratio?

It's taller, for a potentially lower frequency vertical patten control, perhaps, without adversely impacting C/C distance when compared to the EOS designs, improving it, actually. I don't know how the numbers work with respect to integral mouth roundover radii, but it appears to afford some improvement there, as well.

Again, I don't mind the more rectangular look, prefer it, actually, especially if that confers performance advantages....

What's interesting about this EOS is that it employs a variable transition radius with having a large radius in the vertical tapering off to nearly nill at the edges. Question is whether that was just to squeeze it on the baffle, or whether it gives better control??

I'm with Zilch on this. No flange and fairly rectangular like what LB just posted.

There is a balance between mouth roundover and pattern control for a given width. It is my theory that mouth roundover, or lack thereof, will be the largest cause of diffraction with this WG. We already have a OS profile with a nice match to the DE250 exit. We don't want to completely forego the roundover or I think our OS efforts will be for naught.

That is where it is vital we have an idea of how low we want it to hold pattern. Penn, I don't think a 15" WG will give us 1khz unless it ends abruptly and then you introduce the bad stuff (honk, homs?). 1200hz is more realistic so then it is a question of how much roundover will that afford us?

Finally, I believe, and please correct me if I'm wrong, we could also narrow the horizontal pattern slightly thus giving us lower pattern control AND a larger roundover. Maybe 70-80deg?

LB-

The abrupt end to the horizontal on that Coda horn is the compromise I'm referring to. They xover that speaker at 1200hz and given its width there is no way they could give up space to a roundover and hold pattern that low. Diffraction took a backseat to form factor and directivity. For home hifi/HT use I think diffraction at the mouth is of higher priority and needs to be balanced with width and directivity.