[undertow]

I was noting there are quite a few good 2-way and 3-way designs on the market, but in general many suffer from lower efficiency or low frequency deficiency.

What does this have to do with anything? I guess I am curious as to why it would not work to buy 2 of a cheaper speaker that has an 8 ohm  nominal load and simply run them in parallel?
It would seem that takes them back to the more common these days 4 ohm load, adds a little efficiency, power handling etc…

I realize that the 2 tweeters in the system and acoustic alignment would suffer.
However, I happened to see this design from Daedulus  called the “Ulysses”.

http://www.daedalusaudio.com/Ulysses.html

I guess my point is prove me why not with just taking a design from the 8 ohm styles and double it?
Yes there is 2 crossovers and essentially you end up with 2 “Stereo Pairs” of speakers, but seems like in this Ulysses design specifically the way it’s is oriented in the cabinet it would not be of any significant issue.

It seems the giveaway that this is simply a separated or 2 mirrored with 2 crossover design inside is that it uses not only 2 primary tweeters, but 2 ambient tweeters as well saying to me this is some type of Dual 3-way or dual 4-way system in one cabinet on their own networks?

Would the extra entire crossover run in parallel like this have any detrimental effects to your amp? Seems it would just be a higher power handling / thermal handling, and efficiency per channel?
Also, out of the many 8 ohm monitors that you could essentially stack with the the tweeters flipped to the middle why not using the same theory?

This is more of a hypothetical and technical question more than putting it in physical action as of course as in many cases there will be challenges to this, but why not just design a good stable 8 ohm speaker and simply use two crossovers to parallel them together into one cabinet? Seems a bit more efficient and cost effective than many of the exotic designs out there.

Thanks

[Danny Richie]

You are not the first to consider this.

There are detrimental comb filtering effects though that drastically disrupt the response when you do that, especially in the upper frequency range.

Years ago there were some guys on the AV123 forum talking about doing that with pairs of the X-LS speakers. So I took a pair of our A/V-1's and set them on top of each other with the tweeters together. I then took measurements to show what happens. I will dig those up and post them for you with explanation.

[undertow]

I am sure I won't be the first or last

Comb filter huh?

But in actual production sitting there listening not measuring does it sound worse?

I wonder if simply the application of lower distortion, better efficiency etc... outweighs the small anomalies as you are now throwing out a lot more in room acoustic power and control?

When I was a kid I messed with some old gear that still had the "A-B" speaker switches on them and even put some old Advents or something together on a single channel sometimes loved it Of course I had no idea what I was looking for or listening too at that time, but in the end it did not seem to have bad effects accept on the hearing!

Thanks, I look forward to more about this.

[Danny Richie]

It's not a small anomaly. I just found the files. 

[Danny Richie]

Okay, whenever I test any speaker I measure on and off axis. And a very important measurement is the vertical off axis. This shows how well the drivers are in phase over a varying vertical range and tells a great deal about floor and ceiling reflections and what the overall in room response will look like.

Here is one of my A/V-1's measured on axis (red line) and then each additional measurement is made by moving the mic up 4". So orange, yellow and green are up 4", 8" and 12". These measurements are taken at one meter.



This will give you some point of reference. Hang in there. Here comes the good stuff....

[undertow]

I am sure there has to be issues so I believe you.

Just curious then how like as the example above do they get a double tweeter to work? Its on a single crossover? Why not use dual tweeters in every design if this is effective considering then you could compete with dual domes getting the efficiency vs. some of the higher efficiency Ribbons, AMT's and Horns?

Thanks

[Danny Richie]

So I stacked another A/V-1 (It's a matched pair) on top of it tweeter to tweeter.

I kept the mic on tweeter axis of the bottom speaker. The comb filtering effects were so great that I couldn't even go up 4" at a time. So I did it all over again only moving up 2" at a time.

This is the bottom speaker only.



It looks a lot like the one I posted above, but obviously not as bod as the change in height is less.

Now here is the same thing (same exact mic settings) with both speakers playing.



The purple line is still on axis of the bottom speakers tweeter. Then blue, green, and light blue are shows the response as I move the mic up 2" each time.

How's that?

Okay here they are overlaid.



Note that you do gain 6db of output in the lower ranges where the wavelengths are long and they couple with each other. But as frequency increases and the wavelengths get shorter then they are out of phase with each other due to a variance in distance. So they cancel each other out. That is comb filtering.

[Danny Richie]

Just curious then how like as the example above do they get a double tweeter to work? Its on a single crossover? Why not use dual tweeters in every design if this is effective considering then you could compete with dual domes getting the efficiency vs. some of the higher efficiency Ribbons, AMT's and Horns?

Thanks

It doesn't work. The company that you linked to that has a speaker with dual tweeters has this exact same cancellation effect in the off axis. The one that you linked to has the tweeters offset some and so the effect is in the vertical and the horizontal off axis.

[undertow]

Hmm, thats interesting as that speaker I found some great reviews!

[Danny Richie]

It gets even worse the further away the tweeters are.

Check out the measurements I took of the Axiom VP150. Here is a pic of the speaker.



The tricky part was trying to get absolutely dead center to minimize the effect.

Here is the on axis response.



Now here is what happens when I move the mic over 1" horizontally.



And here is what it looks like with a standard on axis, 10, 20, 30, and 40 degrees off axis.



Nice huh...

The mess in the lower ranges is a result of cancellation of the woofers as well.

[Danny Richie]

Hmm, thats interesting as that speaker I found some great reviews!

Try listening to it and moving up or down or back and forth. The change in the upper ranges becomes very apparent. And it is not hard to realize that you my be in an area where there is a hole in the response.

[undertow]

Oh I don't own the speaker, I was just commenting. So no way I can test this, but I believe you. So it really seems silly to use dual tweeters in general.

I know there are some other 4 and 5 way designs that use a tweeter, and then a super tweeter type array crossed at different levels of course. Even this seems a bit difficult to make work.

I have heard a few designs in the "Di-pole" design with simply an ambient tweeter on the back and a single front tweeter, honestly never found those anymore appealing either.

One tweeter rules I guess!

Thanks

[wgallupe]

Danny,

Since most real world listening is done much further away then the mic placement at one meter, would the graphs smooth out a bit of the mic was placed at, say, 3-4 meters?

[Danny Richie]

I know there are some other 4 and 5 way designs that use a tweeter, and then a super tweeter type array crossed at different levels of course. Even this seems a bit difficult to make work.

It is very difficult and there are trade offs.

I have heard a few designs in the "Di-pole" design with simply an ambient tweeter on the back and a single front tweeter, honestly never found those anymore appealing either.

You can add a rear firing or up facing tweeter without effecting the on axis or off axis response. It can add a bit of ambiance and in some cases can improve the imaging. But it can also be somewhat room dependent. 

[Danny Richie]

Danny,

Since most real world listening is done much further away then the mic placement at one meter, would the graphs smooth out a bit of the mic was placed at, say, 3-4 meters?

Yes, from further away effect is like being a meter away and going up just 2" at a time or something.

However, the measurements are really telling what the floor and ceiling reflections will be. If the vertical off axis has a huge hole in the response as shown in the 1 meter measurement then the reflection from the ceiling will reflect that hole. So the room reflections will not match (at all) the on axis response and the overall in room response will be messed up. And that does effect the imaging.

The same thing is true for a off axis peak. If you have a huge peak from a side wall reflection then it can really offset the tonal balance and cause listening fatigue.

[mojave]

How about the Tekton Pendragon? They have 3 tweeters. At what point do multiple tweeters become a line source and does a line source have comb filtering?

From the review by Stereo Mojo:

The tweeters are concentric radiators which allow for healthy linear excursion, outstanding 98 dB sensitivity while maintaining top end extension. Place them all in a D’Appolito design with the three tweeters sandwiched vertically between the two mid/bass drivers, controlling vertical response while offering smooth off-axis performance and acoustic power distribution in-room, and the sonic result shows this Jedi Master of Loudspeaker Design sees the forest through the trees.

[Danny Richie]

Yep, those speakers are going to also suffer from comb filtering in the vertical plane. Moving a foot up or down will result in 20db swings in the response. There is no way around it.

For a line source you really need to get the acoustic centers to within an inch or so to start with. So it is pretty tough to do with dome tweeters. With planar magnetic or ribbon tweeters the whole surface is the playing area so you need to get edge to edge within an inch or so if possible. Still there are cancellation patterns, but they are also filled in with the output from all the others. So in the vertical off axis or as vertical off axis changes you get about 3db or so of wiggle across the upper ranges and mostly top octave.

[undertow]

Out of whack!

Domes in a "Line Source" are something like the old Mcintosh speakers shown here. Does not look like they are 1" to each center either. This must have been a serious mess! Not to mention with all those drivers still only achieved 88 db efficiency.

[Danny Richie]

Because there are so many drivers in them there will be some averaged across the top end on those as well, but the wiggles in the response might be more like +/-6db wiggles.

Those wiggles as I call them will move around as height changes. So you might have a 3 or 4db peak here and a 3 to 4db dip there, and as height changes the peaks and dips may switch places.

[undertow]

So if you do 10 sideways MTM alignments its finally overkill and starts to reverse! JK