Coaxial planar dipole waveguide...

JohnR · « **on:** 2 Nov 2011, 03:53 pm »

How's that for buzzwords?!

See pictures below for two planar tweeters surrounded by 8 planar mid drivers. This is an arrangement inspired by the talk about coax "point source" drivers. You will never get a point source with a large planar driver, of course, but since I had been wrestling for some time with the idea of an array of tweeters above an array of mids, I decided to try this type of arrangement out instead. I also was inspired by a discussion with Paul Spencer on his DIY Synergy Horn. The end result is what I was able to come up with that gave the smoothest on- and off-axis response.

Anyway, here are pictures of the prototype. Front:

Rear:

scorpion · « **Reply #1 on:** 2 Nov 2011, 04:03 pm »

Knots of fun !

What's the image like ?

/Erling

JohnR · « **Reply #2 on:** 2 Nov 2011, 04:07 pm »

Hi Erling, knots indeed! Or at least twists and turns

Can't say anything about imaging at this point as I have only made one.

Pez · « **Reply #3 on:** 2 Nov 2011, 04:11 pm »

AJinFLA · « **Reply #4 on:** 2 Nov 2011, 05:30 pm »

Awesome. That's right up my alley John.
Is that a BG neo3 in the middle, surrounded by..??

Don't be afraid to use cardboard!

cheers,

AJ

studiotech · « **Reply #5 on:** 2 Nov 2011, 10:55 pm »

Looks like some Monsoon planar elements....could be really good.

I was gonna try the same thing with the Neo10 are around the Raal, but got lazy and didn't want to spend any more on drivers.

Greg

tesseract · « **Reply #6 on:** 3 Nov 2011, 01:26 am »

Quote from: AJinFLA on 2 Nov 2011, 05:30 pm

Awesome. That's right up my alley John.
Is that a BG neo3 in the middle, surrounded by..??

Bigger BG's?

JohnR · « **Reply #7 on:** 3 Nov 2011, 01:41 am »

Heh, AJ and Greg are right, Neo3 PDR in the middle and Monsoon mids around.

Here's the on-axis response of the mids, before and after EQ. By themselves, the mids produce output down to below 300 Hz, although I don't know yet how much they can do at those frequencies.

Here's the horizontal polars at 0, 15, 30, 45, 60, with a 1.5 khz crossover. The interesting thing about this is that the off-axis response is very sensitive to the angles. There are actually three: the angle of the tweeter "waveguide", the angle of the outer mid panels, and the angle of the baffle. They basically transition from about 90 degrees to very shallow.

I was trying to get the 45 degree off-axis to be smooth, so that worked out OK. I did have at one point a more even spacing of the 15 and 30 degree lines, but not this time. Not sure what I did - every time I move it the response changes. Anyway, there have to be compromises with something like this I figure.

I don't think those tweeters are going to like being driven down to 1.5 kHz, so I'll probably need to move the crossover up. After that - maybe time to add a woofer.

JDUBS · « **Reply #8 on:** 3 Nov 2011, 01:48 am »

This is awesome, JohnR! I have Unity horns, so I can really appreciate this approach.

Very cool stuff, indeed.

-Jim

JohnR · « **Reply #9 on:** 3 Nov 2011, 02:04 am »

Thanks Jim

Paul and I talked about how to use them in a Unity type of configuration and I tried a configuration with the mids at 90 degrees as they would be in a Unity. But the tweeter needs to be able to go below 1kHz for it to work, so I gradually opened up the angle and ended up here.

JohnR · « **Reply #10 on:** 3 Nov 2011, 04:00 am »

This is what happens with the crossover moved up to 3 khz. There is some serious lobing off axis caused I assume by the width (and angle?) of the mid panels. Hm. Well, I guess large planars always do this kind of thing anyway... right? Overall the dispersion is a bit less than 90 degrees.

Curves are 0, 15, 30, 45, 60 degrees, with 90 degrees (black) also thrown in.

2ms gate this time.

studiotech · « **Reply #11 on:** 4 Nov 2011, 10:26 pm »

I fight with the same issues. I would prefer a lower crossover point btw the neo10 and raal for even horizontal dispersion, but the souind of the Raal lower in its range does not suit me, so I keep it high and live with the off axis nulls. Hey, my spot is in the middle and my first reflection points are well treated. In a more live room, I think having a smooth power response is more important.

What about crunching the mid panels even closer and sort of mounting the neo3 over top of them? My BG Z62 center channel has the neo3 perched directly on top of the neo10.

Greg

Tyson · « **Reply #12 on:** 4 Nov 2011, 10:29 pm »

What about a line array of the Neo 10's with the tweeter in the middle of them? Sorta like what VMPS does with their RM40's.

studiotech · « **Reply #13 on:** 6 Nov 2011, 03:28 am »

Quote from: Tyson on 4 Nov 2011, 10:29 pm

What about a line array of the Neo 10's with the tweeter in the middle of them? Sorta like what VMPS does with their RM40's.

This can be a troubling design to pull off well because you transition from a line source in mid frequencies to a point source for highs. Those drivers are the Monsoon planars out of the old computer speaker systems, which is actually what VMPS uses(in a modified form), not the Neo10.

Greg

studiotech · « **Reply #14 on:** 6 Nov 2011, 03:32 am »

John, I was thinking that you could also try to make some Raal style foam or felt cheeks down the outer sides of the planars to reduce their effective width at higher frequencies. Kinda of like your own giant PDR version of BG planars.

JohnR · « **Reply #15 on:** 7 Nov 2011, 09:28 am »

Thanks for the comments and ideas!

Quote from: studiotech on 6 Nov 2011, 03:32 am

John, I was thinking that you could also try to make some Raal style foam or felt cheeks down the outer sides of the planars to reduce their effective width at higher frequencies. Kinda of like your own giant PDR version of BG planars.

That's what the "swagman's roll" on either side is in the picture above. This had the effect of reducing output at 45 deg in the 3-5 kHz range. I couldn't get it mounted right next to the tweeter although it may be worth a try at some point.

Quote from: Tyson on 4 Nov 2011, 10:29 pm

What about a line array of the Neo 10's with the tweeter in the middle of them? Sorta like what VMPS does with their RM40's.

I've avoided that because of the line source vs point source thing that Greg mentions. However, that was mainly with respect to the woofer. By themselves the Monsoon panels are good up to over 15k I think ( I will have to find a measurement) although they are very directional of course. Either way it will be worth an experiment.

At the moment I'm playing with the woofer arrangement to go underneath it but haven't had much time to work on it.

JohnR · « **Reply #16 on:** 10 Nov 2011, 11:17 am »

Quick pic of the woofer module prototype, to go underneath the planars. Nothing specific to report audibly/measurements-wise at this point, although it looks promising.

THWO · « **Reply #17 on:** 10 Nov 2011, 12:34 pm »

Nice project, John, keep going.

If radiation off axis is an important issue to you: How about tilting both outside vertical rows of the mids a little bit to its outsides, may be just by some degrees? Might this improve the radiation off axis a bit? May be the rear radiation will become a bit "turbulent", but I don´t know whether that would be an audible disadvantage at all.

Just a naive idea...

Regards,
Till

JohnR · « **Reply #18 on:** 10 Nov 2011, 04:27 pm »

Hi Till, thanks

They need to be tilted "inwards" to control dispersion below 1 kHz. Tilting outwards gave too broad dispersion below and too narrow above. I know the latter seems counter intuitive but I'm fairly sure from memory that's what happened - there were a lot of measurements though!

Bob in St. Louis · « **Reply #19 on:** 11 Nov 2011, 02:25 pm »

Neat stuff John!
Next time you're in town, bring them by. I'd love to hear them.

Bob