With more then one exciter on the same panel there is no way one can completely stop the cancellation/reinforcement effect, but there are ways to REDUCE it to minimize the effect.
The only true way to completely stop the effect is with the use of separate panels with one exciter on each panel used with a crossover like most conventional cone speakers utilize. This is one of many reasons why I keep on saying that DML's/BMR's physics are very similar to that of conventional cone speakers.
Insufficient excursion would be a limiting factor I think, along with the difficulty of getting sufficient air movement out of DML bending in low frequencies unless it's a very large panel. The panel would have to flex a lot and this leads I believe to the inevitability of pistonic motion to get decent output.
As above for panel bending limitations
Regarding multiple exciters, I didn't find it harder to get a good response when using 4 exciters compared to 1. It is like when setting up sound in a room. One might think it is easier with fewer sources to avoid phase issues. However the opposite is true since the reflections in the room will cause more severe phase issues with longer delays, and several sources will actually help to mask those issues. Of course one can find really bad configurations where multiple sources are worse than a single, and it takes a bit more tweaking to get it right, but I think on average you would end up with a more even response when using multiple exciters even if you placed exciters randomly. There are so many parameters, so it is very hard to draw conclusions what works best from a few test and some youtube videos though since, but in my limited experience, the kind of skewed rhombus pattern used by Tectonic, and also recommended as an alternative by Dayton, seems to be quite easy to get right.
As I understand a DML panel will operate mostly using the piston principle in the sub region. So a close group of exciters will pretty much work as one driver of that piston motion but with more motor force. At least I found that when it comes to bass as expected more exciters meant higher levels also in that region, however playing around with exciter placement it didn't register anywhere close to the difference it made modifying the suspension of the plate. When it comes to bass, I think placing a single or cluster in the middle of the plate will be fine. You just make sure to have enough force to push the plate and the air in front of it, and that the suspension is compliant enough.
The more you work on a DML dedicated to bass, the closer you will come to a woofer design probably
As I understand a DML panel will operate mostly using the piston principle in the sub region. So a close group of exciters will pretty much work as one driver of that piston motion but with more motor force. At least I found that when it comes to bass as expected more exciters meant higher levels also in that region, however playing around with exciter placement it didn't register anywhere close to the difference it made modifying the suspension of the plate. When it comes to bass, I think placing a single or cluster in the middle of the plate will be fine. You just make sure to have enough force to push the plate and the air in front of it, and that the suspension is compliant enough.
The more you work on a DML dedicated to bass, the closer you will come to a woofer design probably
The other contributing factors are a high aspect ratio (approximately 4:1), and attaching the panel to a rigid frame around most of the perimeter with a good damping material.
Spedge said:
even more validation. I pretty quickly ruled against experimenting using multiples on a panel, but my fundamental drive isn't engineering. My gut says that multiple panels with one exciter each is a better use of them, to create a bigger wall. I have four of the Dayton thrusters, and will have 4 ~2x2' panels. I have 10 foot ceilings in my shop and I'd do 8 panels, but my desk is in the way for the lower portion and without that the upper is too much of a pain for what it would add.
switching topics, has anyone compared their best DML build with magneplanars?
I have removed the second exciter and have placed it on the front of the panel so that the two exciters are in push pull mode and are acting as one exiter.
The comb filtering has now disappeared !!!
with cone speakers on a baffle, the comb filtering only travels through the air.
with two exciters attached to a dml panel ,the comb filtering travels through the panel itself and also in the air.
My exciters are pistonic in motion ( push pull) this drives the area of the exciter in pistonic motion ,tectonic call this the primary drive area.
For the cleanest sound and spatiality I find a single exciter on a panel sounds best for me.
steve.
This is a plot of the two exciters in push pull mode .
As you can see the frequency response goes a bit awry above about 2.5k , this is because of the exciter body in front of the drive area.
But below the 2.5k point the response seems fine.
I was playing some loud music last night and it look a lot to get the exciter warm.
This worked far better than the two exciters side by side as they are now working together as one.
They also don't seem to get so hot, but I need more listening to be sure.
The down side is there is now a harshness to the sound above 2.5k which I would not be able to put up with.
To do this properly I would have to join the two exciters together for low frequency use somewhere Below 2k, sub use probably ?
Hopefully a more powerful driver of some sort could drive the low end below the 40hz point ?
Picture to follow.
Steve.
As you can see the frequency response goes a bit awry above about 2.5k , this is because of the exciter body in front of the drive area.
But below the 2.5k point the response seems fine.
I was playing some loud music last night and it look a lot to get the exciter warm.
This worked far better than the two exciters side by side as they are now working together as one.
They also don't seem to get so hot, but I need more listening to be sure.
The down side is there is now a harshness to the sound above 2.5k which I would not be able to put up with.
To do this properly I would have to join the two exciters together for low frequency use somewhere Below 2k, sub use probably ?
Hopefully a more powerful driver of some sort could drive the low end below the 40hz point ?
Picture to follow.
Steve.
DML analogy is like dropping the stone into the pond. Drop a single stone in the middle of the pond and it will make a uniform circular ripple that will flow to the edges. Drop more then one stone into the pond at any location and its ripples will clash and overlap each other in random chaos. The more stones you drop into the pound the more the ripples will clash and overlap each other. In doing so some frequencies will be cancelled out while some will be amplified.
There is no way one can completely stop multiple ripples in the same pond from clashing/overlapping with each other. There are ways to reduce it though.
This is the reason a single exciter on a panel is the most coherent sounding.
There is no way one can completely stop multiple ripples in the same pond from clashing/overlapping with each other. There are ways to reduce it though.
This is the reason a single exciter on a panel is the most coherent sounding.
Audiofrenzy.
The waves , the peaks and troughs can be clearly seen using pink noise , as you move the microphone around the panel you can see the wave moving up and down the frequency band.
On the proplex panel , which had the exciters as close as possible, I could see the wave moving from about 2k to 20k.
Steve.
The waves , the peaks and troughs can be clearly seen using pink noise , as you move the microphone around the panel you can see the wave moving up and down the frequency band.
On the proplex panel , which had the exciters as close as possible, I could see the wave moving from about 2k to 20k.
Steve.
I was a little disappointed with the bracing on my canvas panels.
It didn't cure the problems I was having with the small 6x4 inch panel.
The large area of canvas is not tensioned enough, so makes a noise on very loud low frequencies down to 40hz.
But is ok if xo about 80hz .
The canvas panel with the larger ply panel is ok but my small exciter is having trouble trying to drive the lower frequencies, it is there but at a lower level
Steve.
It didn't cure the problems I was having with the small 6x4 inch panel.
The large area of canvas is not tensioned enough, so makes a noise on very loud low frequencies down to 40hz.
But is ok if xo about 80hz .
The canvas panel with the larger ply panel is ok but my small exciter is having trouble trying to drive the lower frequencies, it is there but at a lower level
Steve.
The plate is not infinite, and will have modes that depend on the dimensions, so it is like a small pool rather with waves reflecting back almost immediately.
So you will not have perfect ripples anyway, but a chaos of intersecting waves very quickly with a lot of amplification and cancellation. And of course if power of a single exciter per channel is not enough, you either have multiple exciters per plate, or more plates. Now figure out how those pools interact together if you have several of them
Luckily since DML is not phase coherent to start with, neither multiple exciters or multiple speakers seems to be that much of an issue.
I have seen people putting multiple exciters in a line or spread out on the panel, and the little testing I did with that didn't show much potential. When I used four exciters are in a close cluster they act more like one large exciter...just like you will get a very similar pattern in the pool when throwing multiple rocks next to each other instead of a single rock.
Also, the typical strong modes of a plate doesn't seem to be in the region that fits between the exciters, and instead it appears to helps enforce the high end a bit. But the overall response has not been less smooth in my experience.
Subjectively I think my 4x exciter plates sounds great, but they do still require quite a bit of EQ. And I was in a rush when I made them and not very scientific about my measurements that led me to the configuration, so I reserve the right to be wrong
A single exciter will have the least amount of cancellation and amplification when compared to multiple exciters. Even Stevie wonder can see it. LOL
You can only place a exciter so close to each other and the larger the exciter the further away they will be from each other with around 1 inch or more gaps between voice coil of each exciter. If I were to drop 4 pebbles in a pond in your exciter configuration it would not be more uniform then a single pebble being dropped in a pond. There is no overlapping or clashing of waves when a single pebble is first thrown into the pound especially in the center as that is were the coherency starts.
Did you do a proper A/B test where you compared a single exciter to multiple exciters? use test tracks with good clean female vocals and then tell me which one you think sounds more coherent/clear.
This is one reason a lot of people like full range conventional cone drivers as well as coaxial drivers with the tweeter in the middle like kef as they have good coherency as the sound radiates from a single point source.
You can only place a exciter so close to each other and the larger the exciter the further away they will be from each other with around 1 inch or more gaps between voice coil of each exciter. If I were to drop 4 pebbles in a pond in your exciter configuration it would not be more uniform then a single pebble being dropped in a pond. There is no overlapping or clashing of waves when a single pebble is first thrown into the pound especially in the center as that is were the coherency starts.
Did you do a proper A/B test where you compared a single exciter to multiple exciters? use test tracks with good clean female vocals and then tell me which one you think sounds more coherent/clear.
This is one reason a lot of people like full range conventional cone drivers as well as coaxial drivers with the tweeter in the middle like kef as they have good coherency as the sound radiates from a single point source.
@Audiofrenzy
It is not that simple. It might seem obvious, and many make the same assumption when talking about point source speakers in a room, but it is not true there either, since multiple sources also help even out the room modes and reflections.
Also you have to consider the speed of waves propagating in the plate material compared to its size. A 1x1m plate would be more like throwing a stone in a tea cup looking at how fast the reflections start to ineract. Within a few ms you will have the plate full of waves interacting with each other, and those non-overlapping initial waves will hardly be part of the audible sound since they happen to fast for us to notice.
I have not done proper AB tests, just testing different configurations on different materials listening and recording response graphs.
My goal has not been only good fidelity though, and power density has been a critical factor. So when I determined that multiple exciters didn't seem obviously detrimental to the response I didn't dig in to the differences more. In my tests the results was better, but like I said there are so many parameters so that doesn't say that much. But it is the configuration Tectonic use, so I guess it has some merit. If you are satisfied with the power of one exciter it is probably not worth pursuing though, and if you are not you don't have much choice.
What configurations have you tested, and what measurements do you have showing the issues you mention?
It is not that simple. It might seem obvious, and many make the same assumption when talking about point source speakers in a room, but it is not true there either, since multiple sources also help even out the room modes and reflections.
Also you have to consider the speed of waves propagating in the plate material compared to its size. A 1x1m plate would be more like throwing a stone in a tea cup looking at how fast the reflections start to ineract. Within a few ms you will have the plate full of waves interacting with each other, and those non-overlapping initial waves will hardly be part of the audible sound since they happen to fast for us to notice.
I have not done proper AB tests, just testing different configurations on different materials listening and recording response graphs.
My goal has not been only good fidelity though, and power density has been a critical factor. So when I determined that multiple exciters didn't seem obviously detrimental to the response I didn't dig in to the differences more. In my tests the results was better, but like I said there are so many parameters so that doesn't say that much. But it is the configuration Tectonic use, so I guess it has some merit. If you are satisfied with the power of one exciter it is probably not worth pursuing though, and if you are not you don't have much choice.
What configurations have you tested, and what measurements do you have showing the issues you mention?
I recall seeing a Tectonic video where they talk about how the center area between the exciters helps the response by making it act like a larger version of the central area or an exciter, which is what provides the initial transient information. So as they explained it, multiple exciters are beneficial in their design. Of course they have to use multiple exciters to get the power needed, and they would highlight any benefit and maybe not discuss the drawbacks as much.
But one thing I found really spectacular when setting them up in all four corners of the dancefloor is how I could hear the stereo image even when standing very close to one stack, which doesn't really happen with normal speakers. I think that it is possible that effect is more pronounced when multiple exciters, and should then be apparent as better imaging also when using home listening levels.
But one thing I found really spectacular when setting them up in all four corners of the dancefloor is how I could hear the stereo image even when standing very close to one stack, which doesn't really happen with normal speakers. I think that it is possible that effect is more pronounced when multiple exciters, and should then be apparent as better imaging also when using home listening levels.