[Paul Hynes]

Opnlybafld,

Good man. It is fortuitous that you have an amplifier that can drive the resulting load. If you are using new wire give the wire a week or so to burn in before appraisal and please post your findings so we can all benefit from your experience.

Browntrout,

I am keen to ensure people do not discount ideas because theory suggests that the idea will not work. In some instances the theory is incomplete. If we all end up designing by textbooks there will be no new innovation.

Danny,

I am always keen to learn so please educate me as to why you have made the following statement :-

Regardless of how many you put in a line, you will not increase the output levels in the top octave. You will still have the same output as a single unit. So if one unit will output 80db with a 1 watt input (as seen in the measured responses) and you are getting 106db in the lower ranges then you already have a variance of 36db.

I am sorry to say that the above statement does not make much sense to me as I am not sure where you get the 80 db and 106 db figures from, although I have to say that I have had a couple of glasses of wine with my evening meal.

Wiring in series parallel will not change any comb filter behaviour. However the disruptive effects of series parallel wiring on the signal integrity will make it very difficult to hear the effects of any comb filter behaviour.

I have not measured the frequency response of the line array. I do not possess the required equipment and, as I am not marketing the design, I am certainly not going to invest thousands of pounds in test equipment just to see what the arrays are doing in areas that are not relevant to the listening position. I think you are missing the point here. With well-recorded program there is a credible image of real musicians, playing real instruments, with a real dynamic range in an acoustic space. Under these circumstances I do not really care how the array measures. I do not use other audio products as a reference, only live acoustic music, and these line arrays tick the boxes in this area. If there are any response anomalies of the magnitude you are suggesting, they are certainly not detracting from the musical performance and certainly not obvious in my listening position. Therefore I must suggest that if the peaks and troughs that you insist are in the frequency response, of a line source array that you have no experience of, are factual, they have little importance in the real world.

I am a little concerned that you keep plucking statements out of thin air, without qualifying the circumstances that allow you to arrive at these conclusions. It would be more helpful if you could give us some information about how you have arrived at these conclusions. After all, you do appear to be playing the devils advocate. As I have said before I am open to education. Let us know what the physics are.

Please also explain why the rear-firing tweeter is required to compensate for a top end that goes away and also where the top end goes away to.

Having a response within +/- 1 to 2 db would be nice as long as you do not kill the life in the performance with too much manipulation using electronic compensation whether passive or active.

Browntrout,

I have to agree with you from my long experience in the audio field for nearly 40 years. Text books, measurements and simulations do not tell the whole story and a lot potential progress is halted by an inability of the mainstream to think laterally. It is the lateral thinkers that are not constrained by currently accepted practice that often move things forward.

Rudolf,

No offence but simulations are not my scene.

Opnlybafld,

I am very pleased that you have taken the time to try my parallel wiring suggestion and found some positive results from it. Praise is due to you for doing it so quickly. It sounds like you are a keen experimenter. Keep posting your experimental results with this wiring arrangement so others can benefit for your experiences.

Regards
Paul

[opnly bafld]

Paul,

I think Danny was referring to the Visaton published graph and the spl between @3k to 10k vs. spl at 20k.
I would guess most of the regulars around here can't hear much over 15k (if that) anyway. 

I totally agree with the flatness of the frequency response, I'm all for it except.....what do I give up?
Pick your poison. Some things I can live with and others..........well.

Lin

[Danny Richie]

I am always keen to learn so please educate me as to why you have made the following statement :-

Regardless of how many you put in a line, you will not increase the output levels in the top octave. You will still have the same output as a single unit. So if one unit will output 80db with a 1 watt input (as seen in the measured responses) and you are getting 106db in the lower ranges then you already have a variance of 36db.

I am sorry to say that the above statement does not make much sense to me as I am not sure where you get the 80 db and 106 db figures from, although I have to say that I have had a couple of glasses of wine with my evening meal.

Okay, I'll try to explain.

Go to the first post that I made and follow the link. Read what I did there and check out the graphs.

I'll re-post this one here:



The lines that are Red, Orange, Yellow, and Grey are the vertical off axis measurements of a single speaker.

Purple, Blue, Green, and lighter Blue are of the two speakers stacked one on top of the other as a MTTM arrangement.

Now note the Purple line.

Both tweeters are playing and the mic is dead center between the two of them but I got less output than with a single tweeter. It is about 6db down at 20kHz. If I moved the mic up a millimeter at a time (or down) I might get that 20kHz range to come up a few db's and if I am real lucky, then I could get it close to the output of a single tweeter.

Even if I could get perfectly dead center between them so that they are not out of phase, I am still slightly off axis from each of them. The higher you go in frequency the more narrow the dispersion. And if it makes this much difference with a 1" dome tweeter, then imagine how much the off axis energy would fall off with a 8" full range woofer.

Any time a driver is producing a wavelength that is shorter than the width of the diaphragm then it will be beaming that range. Move off axis and it drops off dramatically. This is a given with any driver and any size. This is why larger drivers have horrible off axis response compared to smaller ones.

So with a line of 8" full range drivers lined up, and even if you get dead center on one of them, there will be no off axis energy from the others at 20kHz to add anything to the overall output level, and if there was still any energy left there it would certainly not be in phase so it still would add no gain.

Check out these line sources that I designed:

http://www.av123.com//index.php?option=com_virtuemart&page=shop.browse&category_id=36&Itemid=37&vmcchk=1&Itemid=37

And the LS-6 has already won several industry awards.

One has a line of 6 tweeters and the other a line of 9 tweeters. The sensitivity of a single tweeter is 91db. Note also that each speaker only has a total sensitivity of 91db despite the long line of them. They simply do not couple and add output in the top octave.

Note these line source speakers that I designed many years ago:

http://www.epiphanyaudio.com/EALEGENDARYV3.htm

They won an Editors Choice award and a Golden Ear award...

Specifically check out this one: http://www.epiphanyaudio.com/20-21.html

21 Neo 3 tweeters and each one with a sensitivity of 96db. Note the total sensitivity of the whole speaker is only 96db.

So since the B200 only outputs 80db at 1 watt/1 meter (according to the measured response) then adding more of them in a line will never increase that, but you will get gain in the lower end.

So if you are getting 106db of output in the lower ranges (where they couple, and that is normal) then you have a 36db disparity.

Make sense?

And what you may be noticing different about the all series wiring verses all parallel wiring is likely do to the change in inductance. Putting them all in series raises the inductance quite a bit (8 times as much) and as you know a high inductance is like putting an inductor in the signal path. It rolls off the highs.

Running them all in parallel lowers the inductance to 1/8th of a single driver.

So while this is working better for you, because they are all running full range, It's not going to be the same as a line of woofers that are only playing to 1kHz or so.

[AK]

I think adding a horn in front of each driver would be a good solution.
and it's not too complicated to build if rectangular horns are used.

[Danny Richie]

I think adding a horn in front of each driver would be a good solution.
and it's not too complicated to build if rectangular horns are used.

I am sorry, but that would spread the acoustic centers even further apart, make the line even longer, and compound the problem.

[Paul Hynes]

Lin,

I guess a couple of glasses of wine and a late night don?t make for lucid thinking.

I agree that the output at 20 KHz will not be magically augmented, by using 8 drive units in parallel. As I am part of your sub 15 KHz hearing club it is not a big issue to me.

Some will require the output maintained to 20 KHz, and it may be possible to integrate a super-tweeter into the design to achieve this, without damaging the musical integrity.

Regards
Paul

[AK]

Danny, if horn loads each driver to say 1khz, how can it create problems above that frequency?

[FullRangeMan]

Iam not a expert like many of you guys. But I feel this test Mr. Danny made are valid only to this loudspeaker type/brand, seems it have a xover and tweeter, it is not a FR as the B200. Certainly this test is a good example of what can happen.
Over the years I learn the treble extension is not important, but the musical integrity of the fullrange is paramount. I have a sole FR sealed box with a 100mm cone, the driver is the Beyma 5MP60/N specified 50hz to 12khz and I do not fell lack of treble. (I do some tuning on this Beyma).
If I said silly please excuse-me.
Gustavo

[HT cOz]

Guys I must be a total amateur because I can't even get my sub matched to my speakers without a mic, graph paper, and test tones.  Even then I struggle.

[Danny Richie]

Danny, if horn loads each driver to say 1khz, how can it create problems above that frequency?

Because the diameter of the horn will make the distance between each driver greater and cause the cancellation of the comb filtering effects to be even lower in frequency.

But I feel this test Mr. Danny made are valid only to this loudspeaker type/brand, seems it have a xover and tweeter, it is not a FR as the B200. Certainly this test is a good example of what can happen.

It is the same result with full range drivers.

When I get time (near the end of the day, or maybe tomorrow) I will set up some full range drivers and show the results of playing them together.

Will FR125 full range drivers be okay? If not I have various 3" full range drivers as well.

[Paul Hynes]

Danny,

Thank you for the physics tutorial also backed up by actual measurement information. It is very useful for those enthusiasts working on their own line arrays.

I was aware that you had been involved with the design of the line arrays you have linked to and that they had been well received by many.

Considering the latest information you have presented I think the rising response of the B200 through the midrange and lower treble, before the inevitable drop off in the upper octave, and the LF reinforcement of the array, is enabling the overall response to around 10 Khz to be reasonably well balanced. You did suggest this might be the case. Drive unit burn-in has certainly smoothed out the top end and in practice the line array does not sound ragged as the manufacturers curves would suggest. It also does not appear to lack treble output compared to the live acoustic performances I have attended. That said, at my age, my hearing is no longer what it was above 15 KHz.

When I get some spare time I think I may try and graft some HF transducers onto the arrays but they will only stay if this can be achieved seamlessly. One thing is certain, I would find it difficult to go back to other types of loudspeaker, having experienced the capabilities of the line arrays.

I was aware of the electrical issues regarding series/parallel wiring and as you suggest parallel wiring could also be working in the favour of the B200 drivers.
 
Regards
Paul

[Paul Hynes]

FULLRANGEMAN,

Over the years I learn the treble extension is not important, but the musical integrity of the fullrange is paramount.

I have to agree with this statement.

Regards
Paul

[AK]

I thought lower frequencies don't cancel each other that easily. plus horn provides directivity.
so, Danny, why did you design arrays with multiple tweeters then?

[FullRangeMan]

If all LAs have this CTC problem, what is the Filter Corretive suited ?? 

[HT cOz]

I thought lower frequencies don't cancel each other that easily. plus horn provides directivity.
so, Danny, why did you design arrays with multiple tweeters then?

I'll take a stab at this question.  The tweeters used in Danny's Line Arrays are ribbons and thus the spacing of them is very close together.  Basically the CTC of the ribbon tweeter becomes a non issue because they start to act like one giant tweeter.  The woofers are crossed low enough that the size of the wave means you are basically sitting in near field to them and combfiltering is not an issue.

[HT cOz]

If all LAs have this CTC problem, what is the Filter Corretive suited ??

Use ribbon tweeters and cross low enough to the woofers so that they will not comb in the listening area????

[Danny Richie]

I thought lower frequencies don't cancel each other that easily. plus horn provides directivity. so, Danny, why did you design arrays with multiple tweeters then?

The lower waves will couple, but the higher ones will cancel out.

Try this: Take your center to center spacing and get the distance. Then go here and see what wavelength that distance is: http://www.soundoctor.com/freq.htm

Now if the delay in time is equal to this distance then your cancellation will be at this given frequency.

And on the long line of planar magnetics, the voice coil area of those drivers are space less than an inch apart. So cancellation is minimized.

[AK]

So, how do I know what frequencies will couple and what frequencies will cancel each other out?
I may just build horns for appropriate coupling frequency...
so, do those stacked planars actually increase high frequency sensitivity?
sorry to bug you Danny. you are just providing some very useful information, which is not easy to find.

[FullRangeMan]

Use ribbon tweeters and cross low enough to the woofers so that they will not comb in the listening area????

Sorry HTCOZ, I was talking of corretive filters for fullranges as the B200. I hate tweeters, the guy that invented it must be shoted at the paredon. The tweters sound do not spread in throughout the room.
Instead of various ribbons tweeters I would prefer the Carver Amazing 60 inches ribbons and a half dozen of subs like the Gilmore Audio Model1, now with a SPL of 91,5dB. Anyone know if this Carver 60'' is sell as raw driver??
Regards.

[HT cOz]

So, how do I know what frequencies will couple and what frequencies will cancel each other out?
I may just build horns for appropriate coupling frequency...
so, do those stacked planars actually increase high frequency sensitivity?
sorry to bug you Danny. you are just providing some very useful information, which is not easy to find.

Not to jump in here but earlier in this same thread Danny said that "21 Neo 3 tweeters and each one with a sensitivity of 96db. Note the total sensitivity of the whole speaker is only 96db" Sounds like multiples do not increase sensitivity.  Probably just lower distortion and are easier to drive,,, but that is me guessing.