[JohninCR]

Yep... that will do it.      How does the FR look on/off axis?

Within the WG dispersion angle the sound is quite balanced with less beaming of the fullranger.  ie The sound changes very little anywhere between +/- 40deg of axis.  Once you get outside of the plane of the waveguides, the response is still quite balanced, at of course a greatly reduced level.  That's where they are very different than any OB I've built or heard where the bass and treble disappear leaving only the midrange as you move off axis.  The closest to this kind of balance that I've heard with OB's were swept back winged line arrays, which had a broadened bass polar response due to the shift toward cardiod, but the treble still beamed. 

The narrow dispersion of the rear wave can make those direct reflections overpowering, so it will definitely require tuning for room and placement, however, I believe the effort will prove worthwhile.  I think Dr. Geddes is right about constant directivity resulting in more natural sound, since reflections of off axis response will have a balanced FR too.  I'm just trying to take his idea to the next level to include controlled dispersion in the bass too using dipole radiation, which I already find more natural sounding with all OB bass alignments.

[JohninCR]

Here are a couple of more recent attempts at dipole midrange waveguides that are lined with foam to absorb reflections and high order modes.  Tonality is great for both with a welcome increase in sensitivity due to the controlled directivity. 

I wouldn't recommend the 5 sided version, though it has an interesting look.  Something strange happens with the uneven expansion anywhere but directly on axis.  As a result the sound changes with even small position movements within the soundfield.  The depth wasn't sufficient anyway for directivity control to a low enough frequency.

The rectangular version works very well.  My only issue is that unfortunately the waveguide didn't eliminate some of the problems associated multiple drivers running full frequency as I had hoped when I borrowed some ideas from Tom Danley's patent for a multidriver horn.  Those little FE108's sure put out a big dynamic sound.  The way to go will probably be to replace the center driver with a tweeter and make them an MTM.  This attempt also came pretty close to what I was looking for in terms of controlled directivity, so I've already started on a smaller version for a single driver.  Hopefully it will retain the dynamic sound, because it's a really nice combination of the natural uncolored sound of OB's with the big spacious open soundstage, with the some of the dynamics of a front horn.  The foam isn't pretty, but they should look slick after I cover the foam with a smooth black material and do something about those side.

[DanTheMan]

Hey John, thanks for having the guts to try new things and the resolve and resources to test them.

[Russell Dawkins]

I agree with DanTheMan. I see so many innovative designs from you that I'm inclined to take you a little for granted, yet here you are, churning out one after the other! Your latest design makes me wonder about an old idea I had, but never built. Maybe you're the guy to try it!

What about a dipole line array where the side edges of the narrow baffle went into two columns or cylinders. It might be possible to make vertical cuts down one side of a pair of cylinders about 12 - 16" in diameter (sonotubes?) and have the baffle slot into it, or the baffle could be held to the cylinder by adhesive. Looking down on the finished device, you would see what amounts to symmetrical horn loading front and rear. To get the same effect with a single driver would suggest a donut shape around the driver. In each case I see the "horn" throat going right up close to the driver.

This grew out of wondering how to minimize edge diffraction. And liking dipoles.

I know this is a very un-Kosher horn and may well have colorations, but if I were the experimenter that JohninCR apparently is, it would be on my short list of things to try!

[JohninCR]

Thanks Russell and Dan,

I just feel there's a lot of unexplored territory outside of the box, so it's fun exploring how to come up constructions that make the waves do what I want them to do, instead of shaping them via electronic means.

Russell, if sonotube was available down here, I'd be at a different level....too many unbuilt ideas.  After reading some of Dr. Geddes stuff, I want to stay away from horn loading because of the colorations.  The array I did on page 1 of this thread is pretty close to what you are talking about.  It has conical waveguides front and rear, so it's dipole and I addressed edge diffraction with roundovers of nearly in 2" radius that don't show very well in the pic.  My mistake with it was allowing myself to be talked into using a throat, so it became a horn and too much damping was needed inside the WG. 

Here's how I'd transform your idea.  Geddes states that the most ideal waveguide shape is an oblate sphereoid, so you get 12" sonotube and cut it in half.  The cuts become the edges of the mouth.  Flatten it a bit to make it oblate, and splice 4" diameter tubing onto those edges to address diffraction.  One facing forward and one facing back gives you an improved version of what I did with plywood.  Some of that kerfed material might prove easiest for making the complex curves need if it can also be bent to the 2" radius, or simply posterboard on a shaped frame covered with fiberglass and resin for a bored surfboard or boat builder.

[JohninCR]

Now progress is proceeding at a pretty good clip.  Here's today's effort....no coloration even without foam, getting close to nailing the constant directivity, and easy to create high WAF in an interesting form.  The sharp edged terminus is obviously affecting imaging, but hopefully big roundovers remedies that.  If I can avoid foam, then it will be time to break out the good wood and figure out a complimentary woofer baffle.

Eric,
These are the 108's I got from you, so they may have found a permanent home already.

[konut]

Great work John. What determines the size ofthe guide? Is it driver size? What if they were half that size? How much guide do you really need?

[hurdy_gurdyman]

I thought this might be thought provoking.

http://new.photos.yahoo.com/the_hurdy_gurdyman/album/576460762386020019

It's an old Coronado radio from the 30's or 40's I bought locally just to get the field coil drivers. The cabinet appears to have some kind of wave guide on both the front and back. To think, engineers were thinking of this over 60 years ago.

Dave

[JohninCR]

Konut,

Larger would make things easier, and unfortunately smaller won't control dispersion to a low enough frequency.  Even waveguides for tweeters are fairly large.  A horn can be smaller, but they don't control dispersion very well.  I'm shooting for similar results as Dr. Geddes Summa, but in a dipole form.  His top section has a 15" mouth, and the roundovers make it 22" wide.  This allows him to cross to a 15" woofer in the 900hz range.  He uses a large woofer because of it's limited dispersion (beaming) at the higher end of its range.  The result is a very flat response on or off axis above 500hz.  I like his concept, but I believe there are significant gains resulting from being directional in the bass region too, and without a horn the size of a house, the only way to control bass dispersion is with open alignments.

The idea is to control dispersion in the bass region using dipole cancellation.  Then the beaming nature of a large driver should help control dispersion in the transition region to the waveguide, which controls dispersion from there on up.  The technical OB guys use small drivers and narrow baffles in their attempt to control directivity and achieve balanced off axis response, but take a look at the Summa's measured response here on page 11 http://www.gedlee.com/downloads/Summa.pdf .  I haven't seen anyone claim off axis response even remotely close to that.  I may be fooling myself that I can come close to that kind of response, but if the technical experts can't do it with typical dipoles, then waveguides seem like the best path to me.  Even with these rough test cabs I'm hearing a marked improvement over the same drivers on a flat baffle.

[JohninCR]

I thought this might be thought provoking.
http://new.photos.yahoo.com/the_hurdy_gurdyman/album/576460762386020019
It's an old Coronado radio from the 30's or 40's I bought locally just to get the field coil drivers. The cabinet appears to have some kind of wave guide on both the front and back. To think, engineers were thinking of this over 60 years ago.
Dave

Too bad waveguide theory wasn't around back then.  That shape would have created some horn effects, but not constant directivity.  Had they inverted those curves that design might still be in use today, because an oblate spheriod is supposed to be the best shape for eliminating high order modes (HOMs). Geddes says HOMs are a significant factor in horn coloration.  If those curves are hollow, it might be worth trying to turn each of them around.

[DanTheMan]

Dave, that looks like my kind of radio!  Mono, tubes, field coils, wave guides on both sides--maybe you should just fix that up and keep it original.  Does it work now?

[hurdy_gurdyman]

Dan,

The electronics was fried and would take a lot to fix. I took the speakers out for someone else and Darrel H now has the cabnet to play with. I understand he has it all cleaned up and looking nice. Going to put a 10" SI Coaxial and Augie in it. 

Dave

[mcgsxr]

It is my long time dream to use an old waterfall front/tombstone radio like that, as my sub enclosure, for the main floor system in my home.  I love the art deco style for that piece, and my wife hates large woofers.

Perfect match!

Now, if I could only find a reasonably priced carcass in decent shape, with the innards gone, or busted...!

[DanTheMan]

Darrel's really on to something there.  My Si kinda has that shape now, but without the arched top.  I've always wanted a juke box--that's a big reason why I went mono.  I wish I could build a cabinet that beautiful.  It will be interesting to hear how they sound in there.  I really feel like it's hard to screw up OBs.  It has been in my experience, but then again, every driver I've ever put in them was designed for it.

John, would it be beneficial to round over the area inbetween the 2 waveguides?  Intuitively it seems like it might, but with that just being an area where bass frequencies cancel out, it may not do a thing.  that is unless the wife likes it more

[JohninCR]

Dan,

It's hard to mess up the OB sound, especially using the 15's, but once you start adding horns or waveguides it becomes quite easy to mess up because in essence you are adding very early reflections.

Ideally, yes joining the front and back with a big radius roundover would be ideal, but only because the larger the roundover radius, the better for the waves exiting the terminus.  Once you get past about a 2" radius the benefits become negligble.  My most recent pair with the single 108s would require splicing in 3/4's of the circumference of about a 15" diameter pipe to pull it off, and I want to increase the current cab size as little as possible.  Plus, I like the look as is.  I'm only going to add roundovers because they are necessary.

The reason that they are necessary is because of the pressure change when the waves reach the ends of the panels.  If the change is abrupt, then that edge becomes a secondary source (some call it a reflection), but it is delayed in time, so it's like a very early reflection.  The same is true for all speakers, but it's even more critical for these waveguides, because since the wave has been constrained by the WG, the pressure is higher.  Before I was under the misconception that it related to the interaction of the front and rear waves,  but edge diffraction is all about the change in pressure at the edges.  The smoother the transition (ie larger roundovers), the less the edges can be a new source of sound.

I hope I explained that well, because I think it's an important issue virtually ignored in most speaker designs.

[DanTheMan]

John, have you noticed an increase in dynamics with WGs? 

[JohninCR]

John, have you noticed an increase in dynamics with WGs? 

Definitely.  What I've done is similar to putting a sub in the corner.  Of course these drivers aren't omni-directional, but most of the energy at the frequencies within the WG's range are focused within the 90deg dispersion angle that I'm using. instead of expanding further off axis.  I plugged the WG's into HornResp and it reflected very little horn loading, but it did show a bit more for the B200 version that is almost identical.  That is because the CSA of the WG starts at a size much closer to the Sd of the B200, however, the range where the horn loading occurs is below where the B200's response starts to fall off, so it is beneficial and helps flatten response.

I'll be able to quantify what's happening once I do measurements.  Right now I'm figuring out the easiest way to create the big roundovers I want.  I'm actually using the B200 version as temorary HT mains.  My room is narrow, so I have them toed in at 45deg, which puts me at about 20deg off axis.  The extreme toe in is so I can have them closer to the side walls, but with the WG prevents early reflections from the sides.  The efficiency increase is such that I reduced my typical volume setting from 50 to 43, despite listening from much further off axis.

If I can bring imaging into focus with roundovers plus maybe a little foam, then I don't see any negatives in this approach other than size, but lots of benefits.  I get better efficiency and dynamics, flatter response within the WG axis (ie less beaming), more balanced off-axis response (better quality reflections), more directivity (direct radiation is a higher % of response).

[DanTheMan]

Sounds like good work John.  I have been working on a front horn design for my SI, but I like your idea better.  You don't think that just a WG on the front would work out as well?  I'm afraid of rear reflections becoming a problem.  Of course w/o a WG on the back, your bass cancellation at the side null area wouldn't be perfect......what are you thoughts on one WG on the front?  I have a Pi MidHorn with an enclosureless driver I use as my electric guitar speaker--the clarity and dynamics are unreal!  My guitar falls mostly in the passband, so there aren't many negatives to the design.  I also have a sanshen(Okinawan banjo-like instrument) that plays incredible through the same speaker.  Anyway, just curious as to your thoughts.....Thanks a lot!

Dan

[JohninCR]

Dan,

I think you should determine exactly what it is that you're trying to address.  I'm after something specific in terms of polar response, so mine needs to be dipole.  A 15 kind of waveguides itself because it starts to beam much lower in frequency.  Plus you're talking big for a 15" driver.  Also, if you add a horn type shape only to the front, then you will shift the null at the sides forward and probably reduce bass response in front.  Someone did link a blown glass circular baffle that had a bit of a horn shape and was mounted to the front of the driver.  It was quite attractive, but the driver used was much smaller than your SI.

[DanTheMan]

I Just want uniform coverage over as much of my listening area as possible, increased clarity, increased dynamics, and greater efficiency.  The SI already does these things rather well, I just want to improve these qualities if it is possible without adding any deleterious effects(or at least minimal deleterious effects).  So I guess I'm not trying to address any issues, I just want more of a good thing.  I figure that having increased amplitude in the rear wave may not be ideal in all living spaces.  The horn I have in this configuration sounds really good, but has no bass--as I would expect from its design.  I was just wondering if you think a Front Loaded Wave Guide Open Back speaker could enhance the design of the SI?  e.g. would be worth exploring.  I know that you have a lot more knowledge than I do on both these subjects=OB and WG.  So your opinion would be of great value to me.  I'm going to carefully read Earl Geddes paperwork on WGs today.  BTW, size isn't a concern of mine--so long as it fits in my room, my SI is mono and my wife is understanding.

Anyway, How's the project coming?
Thanks Again!
Dan