Hey guys...we need a little rallying here... - Page 4

Check this out, if you are wanting to make an exact copy of the QSC horn, (Which I would only do if you are not planning on selling these, for ethical reasons.)

You can find 3D scanners that attach to CNC routers. It scans an object and makes a file that you can then use to create a duplicate. You could also make changes to the original.

Here's an example of one, although I think it only works with the ShopBot.

Look halfway down the page.

http://www.shopbottools.com/mProducts/accessories.htm

We should ask jzagaja on DIYAudio.com, the guy is making the best waveguides I have seen, I could maybe do the 3D modelling to start.

In this thread
http://www.diyaudio.com/forums/group...des-horns.html

To build an elliptical OS waveguide, maybe 15x10?? It would be a smaller version to his IWATA waveguide designs.

I read about those earlier in the year. You can also get 3D printers as well.

It's funny, last winter I thought up a machine that could do something like that, and at the time didn't even know 3D printers existed. I felt pretty smart for about a week or so, then found that someone already made 3D printers!


It really wouldn't be too hard to duplicate the QSC waveguide if someone wanted to spend the money doing it. But as of now it's been put on hold to see if PE is going to offer them up to the public. If they don't, then someone might try to improve on that design for the DIY crowd. If they do sell it, then it doesn't make much sense to duplicate an item that may only cost $12-$15 from PE. The average person could never make them for that cheap price.

What program file would that be saved in Penngray? As mentioned, I might know a local guy that could turn your file into the prototype. I'll be calling him on Monday. If he can't, maybe the company he works for can.

But its not maximizing the elimination of HOMs either. Its just not as good as the OS choice according all those waveguide discussions on Diyaudo.com


There are alreadly conical options too, not sure if any are bolt on since the only ones I know from dayton are screw on designs.

We would have to ask jzagaja what it is. I PMed him to get his opinion on doing at 15"x10" waveguide.

No disagreement in the max width of choice for the most DIYers and I guess as a business case we know the Abbey width sells much better then the Summa width but Im not concerned what size the majority wants. The majority honstly wants < 8" wide baffles...we have always been the minority to start with

We also already have the 10",14", 15" wide waveguide options from 18sound, Dayton and JBL. Im all for a QSC clone but we do not need 3D modelling to do that we can just copy the existing ones out there.

If we are going to spend time and resources on modelling something then lets model something that is superior to the rest, more or less in the 3 years of waveguide design activity on the many forums the holy grail is still what Geddes did, like it or not people strive to obtain that performance so Its beyond my logic why we are not modelling the OS design. You can not match/beat the best unless you are doing what the best are doing.

I know your goals are different, you are trying to please a larger crowd and I just want the best possible waveguide designs in my room.

I have been following this thread since it started.
I think that all of the ideas here are fantastic and I wholly agree with penngray here; if you are going to be starting from the design phase and from scratch, then why not aim to make the best possible WG.
Of course there are other options, such as the QSC clone etc. But why try to clone something when we all ready know how it works.
If we improve on the design, even just a little then that is progress, and that in the long run will help everyone.
I know I am very much interested in the final design here. Though I might not be much help I know that everyone including myself would benefit from this.

Max

From Jack on DIYaudio.com

I think we are getting ahead of ourselves here. We all want a reasonably priced quality WG. We all agree the QSC is/was a great solution and value. It is gone. We have four main routes with pluses and misnues:

1. Direct QSC clone. We know it works. No engineering or prototyping. The downside is that it is a direct ripoff and I personally don't think it is appropriate but to each their own.

2. Original design similar to the QSC with some improvements in the throat. It would be slightly easier to prototype with simple tools.

3. True elliptical OS using the same dimensions as the QSC. Should be superior to the QSC in "HOMs" but would require more engineering and likely more expensive prototyping.

4. Same as 3 but wider for lower pattern control. It won't fit in prefab boxes and becomes less attractive to the masses for concerns like WAF. For others it provides a middle ground (IMO sweet spot) between the 30" wide JBL and the 12" wide JBL. Plus it would allow bolt on 1" drivers as opposed to threaded.

Number 2 sounds like it would be the easiest and cheapest (aside from 1 which I'm personally counting out). An elliptical design IMO will need a CNC'd prototype which will cost money but could be a big hit in the DIY community since there is nothing like it AFAIK. If the small EOS is done the producer could gauge doing the large size based on its success (or vice versa).

Anyway we look at it we don't even know if an elliptical is feasible. We are talking about nothing until we figure out the feasibility.

Don't know why we're focused on the QSC for anything other than general prameters -- Parts Express is on it, and knows there is substantial demand for it in the DIY community, the purpose of this thread. I'm reasonably confident they'll come up with something.

The premise for "something else" here is elliptical oblate spheroid. Prolate spheroid is another option in Geddes parlance. I'm not arguing against any of that, merely debating the optimum size for such investment.

Geddes did do an elliptical for another company, Canadian, as I recall, some years ago; Penn may be able to track that down.

Also, talk to Wayne Parham about OS options -- he has expressed his own views on this in other forums....

I vote for the "Nothing Like It" choice

Either way, I will get a EOS.

FWIW, Im already assuming the QSC HPR-152i choice or very similar will exist in the future so do not mistake my posts as trying to talk anyone away from the QSC option so to me we can have #2 and #3. No reason to stop at one when we have some momentum and an active discussion.

Are you arguing for Dayton or yourself?

Me, I already have the QSC size, JBL size, Dayton round choice, etc. People can just buy the nice looking XT1086 if they want something maybe slightly better then the QSC.

Obviously Im pushing for something that has been in demand from day 1.

I think we both agree an EOS design being sold to DIYers would be awesome. We are just looking at the size requirements from different POVs. I think its important to XO below 1kHz, more important then any CTC issue.

I do agree with your position though, you are looking out for those who do not want > 16" baffles.

Don't tempt me, Penn.



You'd be giving it more credence if you had some experience measuring the consequences of ignoring CTC. The notches that develop are wider and deeper than the on-axis deal-breaking one inherent in Earl's "perfect" design which he himself desires to mitigate. Yes, lower frequency crossing gets at it, but there are associated tradeoffs and compromises; I'm not so sure the vocals are as strong an argument here as many suppose, either.

Earl claims not to care about the verticals. In substantial measure, that's because he does not presently have means to deal with them other than via padding the floor and ceiling. He plays fast and loose with lots of this in his own self-interest, and it's important to sort out what's actually what in the context of who's doing and saying it....

Zilch, I agree with you on the CTC nulls. They are a vital design consideration and the primary impetus for the pursuit of an elliptical shape over a circular shape. By my math, once you go to a 90x50ish horn, the CTC spacing takes care of itself so long as you keep the woofer tucked as close to the center of the horn as possible and take full advantage of the horn's horizontal pattern control (requires a CD that can play that low).

Something like the JBL 1" mega-PT is less than ideal IMO at 30" wide since there aren't any 1" CDs that can play low enough to take advantage of that width and you still have to mess with that extra horn height. The 1.5" version wouldn't have that problem with a 600-700hz xover and I might give it a go versus my QSC setup.

But yes, a wider WG would require a CD that can be crossed over lower and would eliminate most of the cheaper alternatives.

If we are confident that PE will come up with something like the QSC, then I think we are basically on a tangent about developing an elliptical OS WG.

Yes, until we know otherwise, in my view, but a WORTHY tangent, indeed....

I agree completely.

I do not have to give anything credence (You are the Expert DIYer not me ). There are already speakers being sold that have round waveguides, there are speakers being built that have larger waveguides. They have had great success to their builds.

Its not only Early but Paul W built an amazing speaker using a larger waveguide. You can also look to the designs from Duke @ audiokenesis or Emerald Physics deisgns all have CTC that you wouldn't like but why still do they have amazing sounding speakers?? I just say if they can do it, we can do it. Their products look a lot nicer too, that is important to me.

I have read enough to know that in the world of DIY there is more then one way to do any design and make it sound incredible.

I will say that the CTC on the JBL horns/Selenium CD may make for an easy crossover but it does not matter when it honks and sounds like crap to me so I have to look elsewhere and the HOM discussions have pointed to OS shapes as our best choice.

If you care about verticals then worry more about CTC then other issues. That is the great thing about having many choices. I love my ribbons and they drop like a rock 10deg off the vertical axis so vertical issues may not matter that much to some people.

You also remember I think everyone should just go active and the CTC can go away. Use a steeper slope and stuff fussing over it all

It'd have been good if you had actually built the crossover rather than listening to the entry-level E-Wave driver/waveguide combo without the requisite HF comp.


Delay has nothing to do with the magnitude of the vertical nulls; active merely provides for easier steering of the forward axis. Of course, you can adjust that to compensate for a narrow vertical lobe every time you stand up or lie down.


Pretend that's a slam-dunk answer also, if you like, but Earl hizself told you it's not so....

I'll mention again that Geddes is not doing 90° beamwidth. His included angle may be that, but the -6 dB pattern is somewhat narrower....

Yes, but notice in the individual polars that the SPL is "normalized." That's standard practice in conventional polar plots for determining the effective beamwidth. The off-axis may be down in SPL overall, but so long as the curve itself stays within the -6 dB limit, it qualifies.

His "new" directivity presentation is deceptive in several ways -- the designs are EQ'd for flattest performance at his intended design axis, somewhere in the range of -20°, where he has "set" the white-line cursor, generating the flatish FR across the top graph. To compensate for the fact that the overall SPL is down from true axial, the color scale is offset for this "normalization," and the response anomalies there obscured; the white cursor does not pass through them.

In the exhibit shown, "0" occurs at something more like -3 dB, whereas, in his recommended extreme toe-in deployment, it's "true" zero, including those anomalies, that is illuminating the opposite wall and generating the first contralateral reflection. He might legitimately argue that since the delay is long, this is not affecting imaging, but not so effectively, tonality or timbre.

Notwithstanding these qualifications, however, the black contour line represents -6 dB for Summa, and you can see that the beamwidth varies between 70° at ~1.5 kHz and ~85° at 7.5 kHz. Nowhere within the bandwidth of interest is it 90° "nominal."

Last time I looked, Summa as shown in his directivity paper more accurately illustrates the actual behavior of the product. There's nothing dishonest here; he's merely illustrating the performance in its best light, consistent with how he has designed it and recommends that it be used. There IS a bit of sleight of hand involved in the presentation, though, obscured by a gratuitous "No other manufacturer would DARE show their data in this level of detail" smoke screen....

Look at Abbey and Nathan; 60° nominal for Abbey and collapsing to as narrow as 20° in the top octave, and 70° for Nathan, with both of them exhibiting on-axis response anomalies well above "0" -- reddish does not even appear in the color key. It's all there, just not in stark relief....

http://www.gedlee.com/abbey.htm

http://www.gedlee.com/Nathan.htm

I didn't interpret it as anything more than contemplation either way, rather merely saw an opportunity to expound further on the subject, since I was looking at it for the first time in such detail myself. I had previously only noted that they were not 90°, and raised that question to him directly in another forum, in which he confirmed that they are not, which carries implications with respect to waveguide size, as we are discussing here....

If we spent more time with the actual directivity program, we could likely figure that out. I'm ignoring the red and black lines in the top plot right now, because I don't know either. The right-hand scale gives a clue.

The yellow one, I understand. It's the FR at the horizontal white-line cursor in the map below. In the program itself, you can move that around. He has set it at 18° for this screen shot, where, presumably, the FR is flattest, as that is what he wants to show for marketing purposes. The HF begins rolling off linearly at 8 kHz and is down 3 dB by 18 kHz, per the left-hand scale. That curve represents what the FR would measure at the 18° angle. The two blips just below 8 kHz may be found as the two small grey-outlined "0" dB contours in the map at those frequencies. Summa nowhere goes much above "0", but the other models do.

The iso-contours in the map represent 6 dB increments; the black one is at -6 dB below "0". The vertical dotted line in the top plot corresponds to the vertical white-line cursor in the map.

My color scale is different in CLIO, and I use substantially greater smoothing, but here's an EconoWave from a few months ago, Selenium D220Ti on the QSC 152i waveguide in a Dayton KD trapezoidal cab kit with 1.25" edge overhangs:





http://techtalk.parts-express.com/sh...07#post1658207

Pretty DUMB, violating all the rules, huh?

[Well, it gets better when I actually pay attention to them.... ]