Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Fenris Wulf]

I built a 3-way system (subwoofer, 2-way mains with horn tweeters, active crossover) and I'm not happy with the sound of the tweeters. They use constant directivity horns with the proper 6 dB/octave treble boost. I'm planning to add a column of direct-radiator supertweeters and use a passive crossover so they're fed from the same amp as the horn tweeters.

I'd like to use a single capacitor as a 6 dB/octave high-pass on the supertweeters, and nothing on the horn tweeters. Then I can use an L-pad to dial in the right level on the supertweeters, so the system as a whole has flat treble response without CD compensation. Will this work? Will the horn tweeters lose any treble with this arrangement? Am I going to run into phase shift issues, or do I just reverse the polarity on the supertweeters?

I might put the L-pad on the horn tweeters instead, because they're more efficient.

[Nicolas Poisson]

Hi,

Theoretically, this should work regarding the rough flatness of the response. Sure there will be phase rotation due to the capacitors, but anyway the wavelengths are so small that there will be phase troubles whatever the tweeters you use, just because of the distance between the devices. This might be the main concern indeed.

However I have doubts on using different speakers on the same channel: there is no way to adapt the response to the acoustic of the room. Managing several identical sources is already a complex problem; it is useless to make the problem even more complex using different devices.

Why do you need more tweeters? What type of room do you want to put all the stuff in? For what application? Home theater? Live music venue? Dance club? Why choosing additional tweeters different from those you already have? How do you plan to manage the interactions between multiple sources?

[Fenris Wulf]

It's for live music, both outdoors and indoors. The horn tweeters are lacking in the top octave and I want to extend the response with direct radiator tweeters. The L-pad will let me balance the tweeters and supertweeters for different acoustics.

The drivers for each frequency range are arranged as a line source. It'll be difficult to align the supertweeters with the horn tweeters, since the horn tweeters are recessed about six inches. I may have to put the supertweeters above the horn tweeters, instead of side-by-side.

I'm concerned about cancellation in the midrange, where the high-pass filter will have the most phase shift. Is there some way to correct it with an all-pass filter on the horn tweeters?

[Nicolas Poisson]

The problem is that wavelengths are very small, typically 10 to 1 inch for mid and highs. A displacement greater than 1/4 of the wavelength will create phase issues. There is no way to correct this with a delay since it completely depends of frequency. Even gluing the supertweeters to the existing tweeters will create problems in the highs. Phase issues tend to lower the average level and often intelligibility drops. It is impossible to create additive waves if the speaker were not designed in this purpose. What you propose sounds like enthusiast DIY solution to me, but I am afraid that it will not deliver a good sound. I suggest you turn toward simpler and more proven solutions.

I do not really understand what you mean by "lacking an octave". It can mean so many things!

If you mean that the top octave ranging from 10 to 20 kHz is not present according to the spec sheet, I would firstly ask: does this matter when you hear the system? Or did you convinced yourself this was lacking just looking at the frequency response graph or reading audiophile forums? Indeed, what do you call “supertweeters”? The idiot-phile stuff ranging up to 100kHz? Or standard PA speakers capable of reaching 20kHz?
Remember that a 20 years old person with good hearing can hardly hear 20kHz, and many high end sound reinforcement systems do not go higher than 15kHz. Also remember that a flat response is often not a good idea in sound reinforcement. The fact that the response is the same for every seat is way more important.

If you really hear something is lacking, it can have many origins and the solution may be much simpler than adding stuff.

Why not EQ'ing?

Or maybe the system is not correctly set up, the curve of the line array is not correct and you already have huge speaker interactions in the highs. Then the solution would be to modify the array curve.

Or simply the speakers are not correctly oriented toward the audience and most "rows" are out of the tweeters directivity.

If the current array is correctly set up but its poor quality prevent from any hope of getting better sound. Then the best solution is to change the complete system.

Or...

I could state hundreds of other hypothesis... what you describe is too vague. In order for us to help you, you need to express more precisely:
- What is the problem? How did you detect it? Does it concern the whole audience or just the front rows? Or the sides?
- What hardware do you currently use? Brand, model, number of speakers, etc.
- How is it set up? Hang? On stage? By a professional? Just put there?
- What is the typical venue? How many seats? Dimensions?


Regards

[Ivan Beaver]

As Nicolas has already said-getting an "alignment" is going to be really hard.

Let's break this down to 2 basic problems.  The first is how are you going to get the tweeter array BY ITSELF not to interfere?  Since the wavelengths are so small-getting the devices close enough to each other is a MAJOR problem.

Then Step 2-once you have figured out how to do step ONE.  How do you plan on getting the "array" close enough to the regular HF driver?  Again the whole spacing vs wavelength issue.

There is a big difference between doing it right-and just making some noise.  Making a "good noise" is not so easy-especially up high.

The fewer devices you can get by with-the better it is going to be-with ANY product.

You "might" be able to get it kinda close at one listening position.  But at a different position-it will be off.

My suggestion is to use a better quality HF driver and be done with it.

[Fenris Wulf]

There's nothing wrong with either the design or the HF drivers. I did my homework when I built the system. The horn tweeters are line source and have a wide horizontal dispersion. However, I just don't like the sound of the horns in the top octave.

You're right about the difficulty of aligning the horn tweeters and supertweeters. Putting the supertweeters above the horn tweeters might work.

It looks like I'll need a conventional two-way crossover to avoid phase problems. I'll have to experiment with the crossover slope and frequency and see what I can get away with.

Part of the problem is the crappy EQ in the dbx DriveRack that I'm using as a crossover. It's not really suited for CD compensation. I have a bunch of stereo console modules with 1st-order shelf EQ, I'll have to rack one and use it on the output.

[Mac Kerr]

There's nothing wrong with either the design or the HF drivers. I did my homework when I built the system. The horn tweeters are line source and have a wide horizontal dispersion. However, I just don't like the sound of the horns in the top octave.

If you don't like the sound of the horn, then there is a problem with the design. You are considering fixing that problem by adding some "supertweeters". A better fix would be to try a different horn and/or horn driver.

You're right about the difficulty of aligning the horn tweeters and supertweeters. Putting the supertweeters above the horn tweeters might work.

Putting the supertweeters above the horn won't change the problems inherent in time aligning them to the horn. You might be able to get the bottom one in phase, but as you move away from the horn with more supertweeters the relationship to the horn will change, and they will move farther out of phase.

It looks like I'll need a conventional two-way crossover to avoid phase problems. I'll have to experiment with the crossover slope and frequency and see what I can get away with.

Part of the problem is the crappy EQ in the dbx DriveRack that I'm using as a crossover. It's not really suited for CD compensation. I have a bunch of stereo console modules with 1st-order shelf EQ, I'll have to rack one and use it on the output.

For aligning the components of your multiway system the dbx DriveRack will be much more useful than a single high frequency shelving filter. In addition to being able to apply multiband eq to each component individually it will have delay to let you adjust the time relationship between components to get the phase as close to in phase at crossover as possible.

Mac

[Stu McDoniel]

I built a 3-way system (subwoofer, 2-way mains with horn tweeters, active crossover) and I'm not happy with the sound of the tweeters. They use constant directivity horns with the proper 6 dB/octave treble boost. I'm planning to add a column of direct-radiator supertweeters and use a passive crossover so they're fed from the same amp as the horn tweeters.

I'd like to use a single capacitor as a 6 dB/octave high-pass on the supertweeters, and nothing on the horn tweeters. Then I can use an L-pad to dial in the right level on the supertweeters, so the system as a whole has flat treble response without CD compensation. Will this work? Will the horn tweeters lose any treble with this arrangement? Am I going to run into phase shift issues, or do I just reverse the polarity on the supertweeters?

I might put the L-pad on the horn tweeters instead, because they're more efficient.

Not enough information here. Are you using compression drivers on CD horns?
1" exit? 1.4" exit? 2" exit?  What exactly are you using?  What you have described above as your solution is basically incorrect to do it right.  Honestly if you went with
a 1" exit waveguide with a nice compression driver your issue would be resolved
and you would not be dealing with the initial design problem you have now.

[dick rees]

If you don't like the sound of the horn, then there is a problem with the design. You are considering fixing that problem by adding some "supertweeters". A better fix would be to try a different horn and/or horn driver.

Mac....

OP states that he doesn't like the sound of the horn "in the top octave".

My quick and easy fix is to dump most or all of everything above 10K........

[Nicolas Poisson]

Fernris, if we have no details we cannot help. When we ask for details, you just answer that you did the work correctly so this cannot be the right line of inquiry.
Until know, you gave no indication of what your current system is. Is it DIY? Or a commercial product?
Also the highs do not please you. But it seems that you did not investigate what precisely does not please you. How can you hope to solve the problem if you did not identify the root cause? And you give so few indications that it is impossible for us to help determining the possible root causes.
I still even no know whether the line array is the right solution for you. That is why I was asking for the typical venue, number of persons attending the shows, etc. I do not know if you use a "micro" line-array for 150 people or eight 2x12" boxes/side for 5000 people...

Splitting the highs in two bands, one for current tweeters and the other one for supertweeters will solve the problems of interactions between these two series of speakers. But it will not solve in any way the interaction between the supertweeters themselves. Or maybe you thought adding only one supertweeter per side? Then again, what is the size of your array? What "supertweeter" were you thinking to?