Hi guys,
I am planning an open baffle speaker with a 15" BMS woofer, a 18s nmba520 10" mid and the 18s nds1095 CD with XT1086 horn.
The Baffle for the 15" will be a U-frame to be determined. The one for the mid is supposed to be as small as possible.
I wondered now if the speaker could potentially profit from a second mid driver in a MTM configuration. With a crossover at 1400hz a tweeter woofer distance of 23cm is advised which seems doable with the 10" and size of the horn. I am not too familiar with MTM designs and well, speaker design in general so is that a stupid idea? What are your thoughts?
I ordered the parts a little prematurely a few weeks ago so that is nothing to be discussed. I could only order a second nmba520.
Thanks!
I am planning an open baffle speaker with a 15" BMS woofer, a 18s nmba520 10" mid and the 18s nds1095 CD with XT1086 horn.
The Baffle for the 15" will be a U-frame to be determined. The one for the mid is supposed to be as small as possible.
I wondered now if the speaker could potentially profit from a second mid driver in a MTM configuration. With a crossover at 1400hz a tweeter woofer distance of 23cm is advised which seems doable with the 10" and size of the horn. I am not too familiar with MTM designs and well, speaker design in general so is that a stupid idea? What are your thoughts?
I ordered the parts a little prematurely a few weeks ago so that is nothing to be discussed. I could only order a second nmba520.
Thanks!
The layout you plan is not unconventional, J. D'Appolito used it and others too.
But the M was smaller which mean lower center to center distance ( less destructive interference).
What kind of filtering do you plan ( passive, active, dsp)? Steep slope will help to minimize artefacts.
Now the choice of size of drivers: it is usually wise to double membrane area each time you go down an octave frequency for low end ( to lower cone movement which means less distortion).
With your driver choice it would need 2x15" to double the mid surface of 2x10" ( the choice made by B&W when they switched from 1x15" to 2x10" in the 800 serie).
If you want absolutely 2x10" it wont be a big deal but the weakest link will be your low way ( max spl).
Have you already heard an MTM? Some people dislike the rendering ( i do like it as this is a good way to approximate a point source with multiple 'conventional' drivers ( read not coaxial).
But the M was smaller which mean lower center to center distance ( less destructive interference).
What kind of filtering do you plan ( passive, active, dsp)? Steep slope will help to minimize artefacts.
Now the choice of size of drivers: it is usually wise to double membrane area each time you go down an octave frequency for low end ( to lower cone movement which means less distortion).
With your driver choice it would need 2x15" to double the mid surface of 2x10" ( the choice made by B&W when they switched from 1x15" to 2x10" in the 800 serie).
If you want absolutely 2x10" it wont be a big deal but the weakest link will be your low way ( max spl).
Have you already heard an MTM? Some people dislike the rendering ( i do like it as this is a good way to approximate a point source with multiple 'conventional' drivers ( read not coaxial).
Last edited:
yes indeed. My main concern is if the mid is too large for a MTM configuration.
I will use the Hypex plate amps with DSP.
you are right, it probably isnt really worth it to use a second mid driver. I was just appealed by the point source character of MTM if done right. Though i have never heard it...
Do you have any other design features in mind i could think of to make the most out of this?
I thought about angling the 15" to reduce floor reflections but i am not too sure yet if thats a good idea and how to do it right.
I will use the Hypex plate amps with DSP.
you are right, it probably isnt really worth it to use a second mid driver. I was just appealed by the point source character of MTM if done right. Though i have never heard it...
Do you have any other design features in mind i could think of to make the most out of this?
I thought about angling the 15" to reduce floor reflections but i am not too sure yet if thats a good idea and how to do it right.
That 10" driver is very expensive ��
Considering the total system sensitivity, I would go for a second 15" woofer: WWTM
Considering the total system sensitivity, I would go for a second 15" woofer: WWTM
True, but two 15" per side is really getting big and simulation with EDGE shows it should be alright using one per side for -6db at 30hz at 100db and reasonable xmax
In general, whether OB/dipole or in a box, a 10" driver used in an MTM format needs to be crossed over much lower than 1.4kHz to avoid off-axis holes in the upper part of the passband of the two 10" drivers.
The center to center distance will be minimum 10" plus half the width of the horn. Likely an inch or two more because of physical constraints involved in placing the drivers very close together. So this will be about 16 inches. I simulated this in TheEdge for a baffle shape like you mentioned. The bottom is wider and can fit a 15" driver. The two 10" drivers are in a slimmer upper part, spaced 16" apart.
The horizontal directivity was done on the tweeter axis out to 45 degrees. Looks like it will just be OK for a 1.4kHz crossover point, but this is not where the MTM configuration has its "pattern holes". Instead we need to look in the vertical axis.
In the vertical axis, a sensible (to me) way to look at it is that you will either listen sitting down (ears at about 1m above the floor) or standing up (ears around 2m high), roughly speaking. As before, on the tweeter axis is OK (at ~1m above the floor) but as you move up to 2m you can see that the "hole" moves lower in frequency and is centered at about your crossover frequency. This will make the speaker sound dull. Also, sound is radiated at larger angles even though your ear will not be positioned there, and this will reduce the sound power at around 1kHz, making the "in room" sound a bit dull. But this might match the directivity of your horn tweeter. Unfortunately, I have found many DIY speakers using horn tweeters (but not shallow waveguides) to sound very dull in the room because of the large difference in radiated power between the low and high frequencies. It's kind of the opposite approach compared to dipole systems, so I tend to knock it. But you might like that kind of thing.
If you design the system as an WMT instead, you can "aim" the in-phase spot of the M-T crossover at the listening location using delay, etc. Without the second M there will no "holes" generated by interference. It's the better approach IMHO. The 10" will be able to be used down to a couple hundred Hertz where it can easily be crossed over to the woofer, so there really is not a need to add a second 10" mid. The woofer sensitivity will be the limiting factor in the overall system sensitivity.
The center to center distance will be minimum 10" plus half the width of the horn. Likely an inch or two more because of physical constraints involved in placing the drivers very close together. So this will be about 16 inches. I simulated this in TheEdge for a baffle shape like you mentioned. The bottom is wider and can fit a 15" driver. The two 10" drivers are in a slimmer upper part, spaced 16" apart.
The horizontal directivity was done on the tweeter axis out to 45 degrees. Looks like it will just be OK for a 1.4kHz crossover point, but this is not where the MTM configuration has its "pattern holes". Instead we need to look in the vertical axis.
In the vertical axis, a sensible (to me) way to look at it is that you will either listen sitting down (ears at about 1m above the floor) or standing up (ears around 2m high), roughly speaking. As before, on the tweeter axis is OK (at ~1m above the floor) but as you move up to 2m you can see that the "hole" moves lower in frequency and is centered at about your crossover frequency. This will make the speaker sound dull. Also, sound is radiated at larger angles even though your ear will not be positioned there, and this will reduce the sound power at around 1kHz, making the "in room" sound a bit dull. But this might match the directivity of your horn tweeter. Unfortunately, I have found many DIY speakers using horn tweeters (but not shallow waveguides) to sound very dull in the room because of the large difference in radiated power between the low and high frequencies. It's kind of the opposite approach compared to dipole systems, so I tend to knock it. But you might like that kind of thing.
If you design the system as an WMT instead, you can "aim" the in-phase spot of the M-T crossover at the listening location using delay, etc. Without the second M there will no "holes" generated by interference. It's the better approach IMHO. The 10" will be able to be used down to a couple hundred Hertz where it can easily be crossed over to the woofer, so there really is not a need to add a second 10" mid. The woofer sensitivity will be the limiting factor in the overall system sensitivity.
Attachments
Thanks, as always very helpful!
So i think i will better ditch the MTM idea since the benefit is questionable and the driver is really expensive.
Do you have more concerns with this plan (WMT) or ideas how to improve it? I am really reading a lot about open baffle speakers but 90% are just a plain plank with holes in it...
Can the overall lover system sensitivity be a problem besides the power requirements?
So i think i will better ditch the MTM idea since the benefit is questionable and the driver is really expensive.
Do you have more concerns with this plan (WMT) or ideas how to improve it? I am really reading a lot about open baffle speakers but 90% are just a plain plank with holes in it...
Can the overall lover system sensitivity be a problem besides the power requirements?
What model is your woofer? Unless it has high xmax and low Fs it will max out very easily on OB use.
Here is how you can check this for yourself, for the driver you intend to use.
As a first approximation, a woofer in a planar open baffle will have a near field response around Fs that is very similar to the "free air" response. This is simple to compute using any box modeling software. Simply plug in the TS parameters for your driver and then set the "box volume" to at least 1000 times Vb. I just put in 99999999 or something like that. Ignoring the effects of the baffle itself, what the driver sees is something like a box with "infinite" internal volume. This usually means a "drooping response", meaning the frequency response falls off slowly starting far above Fs, because Qts is less than 0.5. When Qts is low, eg below 0.3, this effect is really impactful.
On top of that, the system will incur "dipole cancellation" effects. It's a "baffle step" effect for open baffle or dipole systems. This is what you would hear in the "far field". You have to add the baffle losses (assuming the responses are given in decibels, e.g. the SPL response) to the "infinite box" response you calculated above. You get the dipole cancellation using a diffraction modeler that can simulate open baffle systems. I like to use "TheEdge" to simulate these effects, but there are other programs that can do this as well. The smaller the overall baffle dimensions, the more the low frequency loss.
As you go lower and lower in frequency, losses will eventually become insurmountable no matter how much power you have available. Some designers like to try to make the baffle larger, fold it back, etc. and make do with what that brings. I like to "throw in the towel" somewhere between 80-100Hz and in doing so can use a relatively small baffle (or even no baffle). Then I cross over to either a closed box subwoofer, or a very large H-frame dipole sub, depending on how large of a space the loudspeakers will be used in (larger = H-frame, smaller = closed box). This is because, unless you make the front to back pathlength quite large, it is difficult to get sufficient low bass SPL.
Totally agree with CL...
A typical 15" PA woofer can max out at even normal listening levels with bass heavy music on OB. Maxing out is when the voice coil starts hitting the pole plate. A little more volume and you risk blowing your driver. This is because there is no force to dampen/stop the cone (no high pass effect provided by the enclosure). To stop maxing out you will have to do one or more of the following:
Listen at very low volume
Listen to music with no bass
Apply a high pass filter
In my experience, a single 15" usually requires a high pass at 50hz. Double 15" can probably handle 40hz HP.
Pick your compromise.
A typical 15" PA woofer can max out at even normal listening levels with bass heavy music on OB. Maxing out is when the voice coil starts hitting the pole plate. A little more volume and you risk blowing your driver. This is because there is no force to dampen/stop the cone (no high pass effect provided by the enclosure). To stop maxing out you will have to do one or more of the following:
Listen at very low volume
Listen to music with no bass
Apply a high pass filter
In my experience, a single 15" usually requires a high pass at 50hz. Double 15" can probably handle 40hz HP.
Pick your compromise.
Stick with one 10".
You will most likely get the best results between the 15" and 10" if you overlap the crossover a bit.
300hz LP on the 15" and 200hz HP on the 10" for example.
I would make u-frames for both the woofer and mid. Instead of having the wings level at the back, make them variable length to break-up resonances.
Don't worry too much about the exact baffle shape and size. There are compromises to any design. Make the baffle just a bit wider than the woofer if you want it compact. You can make the 10"/horn baffle the same width, narrower, pyramid shape...up to you.
It will make for a very nice system once the tuning is sorted out.
You will most likely get the best results between the 15" and 10" if you overlap the crossover a bit.
300hz LP on the 15" and 200hz HP on the 10" for example.
I would make u-frames for both the woofer and mid. Instead of having the wings level at the back, make them variable length to break-up resonances.
Don't worry too much about the exact baffle shape and size. There are compromises to any design. Make the baffle just a bit wider than the woofer if you want it compact. You can make the 10"/horn baffle the same width, narrower, pyramid shape...up to you.
It will make for a very nice system once the tuning is sorted out.
I had the best results in EDGE with a crossover at 190hz for the 10" and 145 or so with the 15" using a 2nd order slope. But that will be probably easy to sort out when a prototype baffle is done.
Variable length sounds good. I thought about that and as long as the program can be trusted it has a really nice effect on spreading that pipe resonance. Though i read that if a U-frame is folded "back" again to simulate it in EDGE the distance to be used has to be different from the actual baffle. Do you know something about that?
Regarding the mid baffle i will probably first produce a baffle i like aesthetically and then try different kind of wings to put on the back. Good advice!
Some guys and i think Linkwitz himself proposed magnet mounting the drivers. Do you think thats worth it?
With the baffle i intended i can reach 110db or something at 40hz with pretty minor corrections in eq according to edge. So i don't really worry about that since i don't need that kind of SPL or sub bass. At 100db and 40hz EDGE calculates an excursion of 1mm. Fine for me!
Last edited:
Magnet mounting of driver could be interesting for the 10" yes but you may need to lower xover a bit: the point is to lower ( make it disapear) a resonance ( within the driver structure) which is usually located in the range 150/400hz ( most often around 200hz). But i'll depend from the driver construction and behavior.
The other interesting point is not to have your frontplate/face to radiate sound ( by architectural/structural coupling) and give a secondary source of emmision ( as you need to decouple gasket from it's the front baffle thanks to neoprene or other resilient material).
You may have to prototype and check for it: it appears as a wiggle in the impedance plot in the frequency range of interest.
The other interesting point is not to have your frontplate/face to radiate sound ( by architectural/structural coupling) and give a secondary source of emmision ( as you need to decouple gasket from it's the front baffle thanks to neoprene or other resilient material).
You may have to prototype and check for it: it appears as a wiggle in the impedance plot in the frequency range of interest.
Last edited:
Thanks, so i will just magnet mount every driver. It doesn't seem to have any drawbacks.
What do you think of slotloading the 15" bass in a very shallow slot, so basically a rearmounted woofer on a 2" thick baffle with a slot. According to Nelson Pass that might increase sensitivity by quite a bit. With such a shallow slot the resulting resonance should be so high that it wont bother me too much and maybe an exponential slot opening could further improve on that.
edit: okay Qts increases and fs is lowered with decreasing slot size. What are the drawbacks? The efficiency gain is probably nonsense
The Slot Loaded Open Baffle Project Article By Nelson Pass
edit2: okay forget it.
What do you think of slotloading the 15" bass in a very shallow slot, so basically a rearmounted woofer on a 2" thick baffle with a slot. According to Nelson Pass that might increase sensitivity by quite a bit. With such a shallow slot the resulting resonance should be so high that it wont bother me too much and maybe an exponential slot opening could further improve on that.
edit: okay Qts increases and fs is lowered with decreasing slot size. What are the drawbacks? The efficiency gain is probably nonsense
The Slot Loaded Open Baffle Project Article By Nelson Pass
edit2: okay forget it.
But its a interesting read: https://www.diyaudio.com/forums/multi-way/259627-slot-loaded-infinite-baffle-care-throw-sanity.html
Attachments
Last edited:
I have GPA 416-8b and faital 18fh500.
Sans high pass filter, Kollektiv Turmstrasse max out both at the slightest krank up of volume (with modest 6db dipole correction). My 15" require HP at 50hz and my 18" at 40hz to handle any self respecting bass note. Now trust me or your Edge, you will hopefully find out yourself soon.
Sans high pass filter, Kollektiv Turmstrasse max out both at the slightest krank up of volume (with modest 6db dipole correction). My 15" require HP at 50hz and my 18" at 40hz to handle any self respecting bass note. Now trust me or your Edge, you will hopefully find out yourself soon.
Well what SPL are you talking about? As i said i don't want to reproduce 25hz at 110db or something. I also don't know how deep kollektiv Turmstraße goes with their songs. They're awesome btw!
A Thread for those interested in PPSL enclosures
This is a variation around the theme of slot loaded. The interest is it lower 2nd order harmonic distortion ( by cancelation of drivers non linearity and a mechanical lowpass ( the cavity). There is a link between the width of slot ( needed for driver insertion ) and highest freq range of the sub. And if it doesn't do for you, you can still go 'clamshell' style ( by reverting the inversed driver).
Wait for Djk (rip) appearing in the thread as he give most infos needed ( including his own way of doing it -6th order assisted ( with eq) for highest possible efficiency, but you can do as you want, ob, closed, even horn load the cavity... principle is always the same). And no need to magnet mount, the inverse they must be in tight contact to the structure. May be easier to achieve ( built).
This is a variation around the theme of slot loaded. The interest is it lower 2nd order harmonic distortion ( by cancelation of drivers non linearity and a mechanical lowpass ( the cavity). There is a link between the width of slot ( needed for driver insertion ) and highest freq range of the sub. And if it doesn't do for you, you can still go 'clamshell' style ( by reverting the inversed driver).
Wait for Djk (rip) appearing in the thread as he give most infos needed ( including his own way of doing it -6th order assisted ( with eq) for highest possible efficiency, but you can do as you want, ob, closed, even horn load the cavity... principle is always the same). And no need to magnet mount, the inverse they must be in tight contact to the structure. May be easier to achieve ( built).
Last edited:
The GPA has 4mm xmax i just noticed. So i can imagine that it maxes out. but an 18" with 9mm linear xmax? That got to be a really minimalist baffle or huge SPL. How large is your baffle or do you may have a link to your project?
Last edited: