Planar Speaker Asylum

Welcome! Need support, you got it. Or share your ideas and experiences.

Return to Planar Speaker Asylum


Message Sort: Post Order or Asylum Reverse Threaded

Midrange/tweeter line array project

50.65.41.21

Posted on February 13, 2017 at 13:26:36
Posts: 2
Location: Alberta
Joined: February 13, 2017
I am thinking about undertaking a line array project. I have a set of Infinity IRS Beta's where the L-Emim's are no longer functional, and replacements are no longer available. I want to build a pair of panels to replace the originals using the Bohlender Graebener Neo10 Planar Transducer and the Aurum Cantus G1 Ribbon Tweeter in a line array configuration. Can I please get some feedback as to whether or not these are a good choice of transducers and could someone point me in the right direction to do some research on crossover design and line array configuration. I am also trying to simplify the system by going with a 3-way configuration instead of the original 5-way Beta design...and similar to the IRS V's. I also understand the cost involved in this project and have taken that into consideration.

 

Hide full thread outline!
    ...
RE: Midrange/tweeter line array project, posted on February 13, 2017 at 20:40:49
Satie
Audiophile

Posts: 5426
Joined: July 6, 2002
Well, there is something different you can do with this and use the VMPS RM40 and V60 as a guideline. The low frequencies are additive with the number of added drivers i a line but the higher frequencies do not add up significantly.
Neo10 drivers have rapidly falling dispersion past 2khz. so are not an easy match to a true ribbon tweeter even in a tall line array with multiple ribbon drivers. As they will sound rough and metalic when driven too low, even when FR plots show that you are well within their operation range. To make it happen you would need either a high slope high pass or DSP to cut off the ribbon fast enough so that it does not distort at high load.

Using the VMPS construct as a guide, I would suggest using a tapered driver line array: BBMMTMMBB, B=4 neo10 - 2 on top two at the bottom, then M=4 Neo8PDR - 2 below the top Neo10 and 2 above the bottom Neo10. T= one or two true ribbon drivers such as the Aurum Cantus. The ribbons would be crossed over to be supertweeters just above 10khz, The Neo8 and Neo10 would be fed the same signal. Try to run it where the composite driver midrange does not use a low pass filter and the supertweeter is put in with a first order high pass somewhere above its acoustic crossover - say 15khz.

There is very little discussion of this construction anywhere. If it were not for some very serious speakers that employ this geometry I would not have thought it practical to construct DIY. The tricky part is getting the driver offset of the ribbon supertweeter right. The "traditional" DIY ribbon line array tweeter as companion to a midrange line array is simply very expensive and most of the expensive ribbons are doing next to nothing while their max output can add up to the 125db range and more. Which you would never tolerate even for a transient burst.

BTW, someone I know uses Neo8 drivers for the EMIMs of his RS2B. He thinks they work marginally better than the originals.

This composite line array should allow you a midrange high pass around 150-200hz

 

RE: Midrange/tweeter line array project, posted on February 14, 2017 at 13:50:50
Posts: 2
Location: Alberta
Joined: February 13, 2017



Thank you Satie for all of the great information and advice. I am still in the process of educating myself regarding the technical aspects of speaker design and the science behind it...and I am sure this project may appear way over my head for a novice. I'll figure it out eventually. It appears I may have to do a little R&D work of my own to see what works as, like you mentioned, there is little available technical support out there. The VMPS reference was a great help as well and I'm still exploring what information I can find on those speakers. Just for the heck of it, I quickly put together a version of a design for the panel I would like to build...I hope the JPG attached to this thread.
If you have anything further to add, I'm all ears...literally :)

 

RE: Midrange/tweeter line array project, posted on February 14, 2017 at 14:45:53
Satie
Audiophile

Posts: 5426
Joined: July 6, 2002
This is not exactly what I had in mind but it is an interesting alternative. In this case, the side by side location of the Neo10 vs the mid/tweeter line would mean that you are back to a 4 way as the Neo10 would need a separate low pass and the Neo8PDR would need a high pass so that you don't have too large an overlap with the drivers placed horizontally at ear level. In this case, the crossover can be in the 1khz range and a 1st order can work.

The downside to this setup is that neither of the line arrays is quite as tall as you would like. If you want the extra extension of a line of Neo10 then you would want 6 of them per side and the mid/tweeter line would need an additional pair of Neo8PDR so that you don't get too much of a residual resonance peak and you have a sufficiently tall line source for that portion.

This arrangement you came up with will provide better lower mid impact and dynamics and allow you to cross over to the woofer section at 100hz. But you will lose a bit of coherence and breadth of the stereo field coverage by splitting the horizontal sources for the midrange in a separate line from those of the midbass Neo10. You also would have more drivers to buy (if you follow my suggestions to realizing your idea).

In the tapered driver array I suggested, you have a tall line source with the center at a tall ear level - and it would be a two way mid/tweeter section. Where it would lack in performance comparing with your idea is that you would have to retain a 150'ish hz crossover and not have quite as much lower mid slam. I still think that would be a better and cheaper alternative.

 

RE: Midrange/tweeter line array project, posted on February 16, 2017 at 00:45:51
Roger Gustavsson
Audiophile

Posts: 2051
Location: Huskvarna
Joined: February 12, 2010

I prefer as simple as possible. Why not an Array of Neo 10 plus Neo 3?

 

RE: Midrange/tweeter line array project, posted on February 16, 2017 at 10:26:54
Satie
Audiophile

Posts: 5426
Joined: July 6, 2002
That is because he wants a true ribbon for its better detail. Those are more expensive than the Neo3 PDRs. The idea for the BBMMTMMBB setup is to retain the coherence and larger stereo field of a single line source for the entirety of the spatially significant output range. The Neo10 would provide further extension towards the bass, allowing a lower HP freq.. That setup would allow the use of far fewer of the expensive ribbons than a separate tweeter array.
If you run a separate tweeter line with true ribbons then the upside is that you can lower the crossover freq. by an octave or more from where you would for a single tweeter. Thus you can gain better detail down into the midrange.

 

RE: Midrange/tweeter line array project, posted on February 16, 2017 at 12:54:15
josh358
Industry Professional

Posts: 12327
Joined: February 9, 2010
You'll run into trouble with this layout because you'll have varying dispersion patterns.

In general, you want either a line source or a point source. The size of the driver should be smaller than the highest frequency wavelength it's reproducing and drivers in different frequency ranges should be less than half a wavelength apart. Both vertical and horizontal dispersion are a concern. This means it isn't a good idea to double up tweeters as you've shown. Ideally, a line source stretches all the way from the floor to the ceiling (there is however a caveat to this in the case of tweeters), and a point source is infinitely small.

Of course, none of this can be achieved in practice so it has to be approximated. I think Roger's layout does a good job of that. You could do even better by going four way -- a line of Neo 10's, a line of Neo 8 PDR's, and then the Aurum Cantus (or if you're willing to bi amp the less efficient Magnepan line source ribbon tweeter which I think are now $750 each).

You could also try to imitate the Beta layout with BG's -- it was a wonderful speaker though I was bothered by the fact that the frequency bands seemed to emanate from different heights. That would include the minimal open enclosure that lets the backwave through for good imaging. However, a good deal of experimentation might be needed to tune the response. A line array is easier, particularly if you can triamp and adjust the crossovers rapidly to speed optimization. It's also potentially better though more expensive, because of the radiation pattern and the large driver area which gives you higher SPL, lower distortion, and better efficiency.

I like Satie's BB-MM-T-MM-BB D'Appolito arrangement though I'm not sure I'd double up on the midranges owing to vertical dispersion issues.

My vote would be a four-way line source approach, that should be even better than the Beta and really all but a few commercially available speakers of any description.

 

Page processed in 0.018 seconds.