DiyMobileAudio.com Car Stereo Forum banner

1 - 19 of 19 Posts

·
Registered
Joined
·
116 Posts
Discussion Starter #1
So in my never ending thirst for change, I am going to try out a Mid tweet mid set-up in my front doors. I have a set of RS180-8 and currently use my fr89's. After a little research into MTM designs and their inherent dis/advantages I think its very doable and with some T/A and EQ might give me strong midbass without nearly as many components. My plan is two RS-180-8's in each door with a as small a center to center distance as possible. Normally, it is my understanding that 2K is about the highest a 6.5 MTM design can go for the lowpass. I had done two mach 5 mids befor in a similar fashion but unkowingly ran them to 3500 or so. . Didnt sound bad but not as clear as I like. I can only hope that with an active set-up, I can dial in the best crossover point. I will use the FR89 to not only lower the mids crossover point but also for loudness. A tweet playing means redustion in power handling/loss of sonic clarity at high spl. A 3 inch widebander playing that high allows more power handling/ sonic clarity. To be honest, I loved the FR89's high end. The cymballs sounded good. I had them crossed at 500 on my 4.150pdx with an sls 6.5. They were great until I cracked them. So a crossover of 1200 or so will help substantially with that clarity. (i hope) The rs180's will be one per channel on the PDX so about 75 watts a mid. I will also be trying my new ARc audio 1000.2T on my sub. I have an audiopipe 1800 mono here now @ 4 ohms for clarity. But after a long look at specs. The pipe puts out .2 distortion at 30% power at 4 ohms!!!!! Arc is .01 no impedance marked. I spoke to rep at arc audio(because that amp is special) and he said next time he talks with Robert(zeff) he would get me more exact distortion figures and efficiency numbers.

Speakers: Dual RS-180-8, FR89EX-4, RF T1 12 (4 ohms)
Amps: PDX 4.150(midbass), RF power T5002(mids/highs), 1000.2T(sub)
Crossover: Active 24 or 12 DB. RF can do both on input and pass through.

I figure, with my little bit of knowledge, that I should end up at around 1200 on the crossover. Thsi will work with my center to center spacing and still allow th fr89 clarity at high levels. Does anyone have a program to tell me my SPL limtis? IE wattage vs excursion on the RS 180's. My doors are well sealed with wood and the woofers will be mounted to a 3/4 inch baffle. Im hopin for a 60 hertz LR4 highpass but again, its all dependent on SPL.

As an aside, does anyone think that a 18DB slope may work better for the mids? I read somewhere that d'appolito himself suggests that it offers the best transition with the lower crossover and avoids off axis anamolies due to the MTM design when used horizontally instead of vertically.???? let me know. I will be posting a build thread with the project.
 

·
Registered
Joined
·
116 Posts
Discussion Starter #2
21 views no opinions? Darn.

Well if anyone wants to give me some thoughts on a vertical alignment vs horizontal in the door. I cant help but beg the question of comb filtering due to differences in distance between drivers and the listener?
 

·
Premium Member
Joined
·
6,034 Posts
I'm afraid this won't work.

Arrays are interesting because they do not 'sum' properly until you get to a certain distance. Until you reach that distance, the sound is completely chaotic; it can vary by a few dB when you move your head a single inch.

A good 'rule of thumb' is 10x the size of the array. So if you had an MTM with dual 16.5cm woofers, with 7cm for the tweeter, you'd want to sit about 400cm back. (13 feet!)

Obviously, this is completely impractical in the car. (I found this out the hard way. I built something like this in the 90s, and it sounded like ass)



David Smith from Snell has a really good grasp on arrays. He's posted about them on diyaudio quite a bit.

Smith has attacked the array problem by shrinking the driver size dramatically. In the pic above, you can see that he's reduced the dimensions to such a great degree, you can actually use this array in a 'real' room.

Also, have you considered using a horizontal MTM? In the car, there are a lot of advantages to using a horizontal MTM instead of a vertical. (Vertical MTMs have a wide 'sweet spot', which is nice for a home. But if you flip an MTM on it's side, you improve the *vertical directivity*, which is particularly nice if the loudspeaker isn't located at eye level. You're trading horizontal directivity for vertical.

You can also maximize both horizontal *and* vertical directivity by making the MTM like a cross:



But once you go that route, coaxials start to look pretty good.

 

·
Premium Member
Joined
·
6,034 Posts
Also, as far as the amplifier distortion, that's a complete non-issue at low frequencies. It's not unusual for a sealed sub to produce distortion on the order of twenty or thirty percent! Even worse, the audibility of distortion goes up as frequency goes up. Which means that the distortion components are often more audible than the fundamental!

So, yeah, don't sweat subwoofer amplifier distortion.
 

·
Registered
Joined
·
1,090 Posts
Also, not sure how you have your widebands mounted now, but in door the top end is gonna be way more rolled off compared to pillars/dash.
 

·
Registered
Joined
·
116 Posts
Discussion Starter #6
Patrick: So you are saying I am better off with the horizontal MTM. I have tried that before with two mach 5 mids, and it sounded great. The only difference between now an then is I plan to run a much lower crossover point on the mids. Also, I dont think that I can make a cross shape with three speakers. I loved the information in your post though as I dabble in my home audio as well. Thats were all my car project woofers and tweets go :). What if I set my T/A to to like ten feet to make the sound act as if it were coming from far away? MYBE????


ecbmxer: Honestly, I loved the FR89s in the door a little bit above knee level. My stage still sat up pretty high. Maybe its cause I sit gangsta or a preference of a more rolled of upper spectrum. The only dislike I had with them was that they sounded stressed on passages that had a hard guitar, like metal.
 

·
Registered
Joined
·
39 Posts
Do you have DSP power to TA each speaker in MTM independently? Varying TA for each speaker might give you wanted directivity... newer tried myself, can be PITA to set it correctly...
 

·
Premium Member
Joined
·
6,034 Posts
Patrick: So you are saying I am better off with the horizontal MTM. I have tried that before with two mach 5 mids, and it sounded great. The only difference between now an then is I plan to run a much lower crossover point on the mids. Also, I dont think that I can make a cross shape with three speakers. I loved the information in your post though as I dabble in my home audio as well. Thats were all my car project woofers and tweets go :). What if I set my T/A to to like ten feet to make the sound act as if it were coming from far away? MYBE????
Not necessarily. Horizontal has it's advantages, YES, but the main problem with this array is that you'll have to listen to it from fifteen feet away. The only way to minimize that is to use smaller drivers, or figure out a way to get them closer together. (The array in my car uses 5" woofers, but I'm reducing the size of the cone so that it 'seems' to be about 2" in diameter.)

Arrays are fine and dandy as long as they're carefully designed, and you don't sit too close to them. But if you sit too close, they sound awful.

DSP can't do anything to change this; it's all a factor of the size.

Picture if you were trying to make a single beam of light out of an array of LEDs. If you're a few feet away, it works fine, but if you're sitting an inch away, it's going to be blinding, and the intensity will vary if you move your head one inch.

It's the same thing with arrays.

It's the reason the midrange array on the Snell is so small. There's an interview with David Smith in Stereophile, where he talks about how the MTM in that speaker wasn't even possible until they could make the drivers small enough.

Snell Acoustics XA Reference Tower loudspeaker | Stereophile.com

"I tried a number of alternative schemes," said Smith, "but none of these alternative systems looked promising, so a more conventional symmetric array was settled on...The simulations showed that constant directivity came from having a constant element spacing relative to the crossover frequency. That is, if the highest crossover frequency is six times the next crossover frequency, the mid-to-tweeter spacing should be one sixth the woofer-to-tweeter spacing. Both of these spacings should be one third (at most, one half) the crossover wavelength. In practice, this means the mids had to be placed very close physically to the tweeter, and the top crossover point should be as low as practical.

"The one breakthrough was finding a high-output, low-resonance tweeter from Audax that has a compact-diameter magnet structure. By rear-mounting this driver on a computer-cut aluminum MTM plate, the Vifa midrange units' driver baskets can overlap the tweeter's metal frame to keep the center-to-center spacings minimized."


One thing you *do* have going for you is that you can use a very low crossover with your 'tweeter'. For instance, if you have a center-to-center spacing of 13cm, you'd need to use a xover frequency of 871hz. That's not practical with a conventional tweeter, but with the Fountek it's doable. So in that respect, it's a good choice for the array.

 

·
Registered
Joined
·
116 Posts
Discussion Starter #9 (Edited)
First: Patrick, Thank you for providing your experiences and information for this project. My major concern was where center to center spacing. Without your Input, I would never have known Distance effects summation.


I wonder how the RS150 would work. these daytons have a smaller cone than most equivilant drivers. have you worked with the RS150 or something similar? what kind of crossover/ midbass response should I expect?

Also, Where did you get woofer sizes? I did R=SQRT(126/pie) for the RS180 as SD=126CM. That gives 6.34=r, and D=12.7. Fountek is D=5.9. Thats 18 or 180 cm away(6 Feet). With seating and speaker placement I will be 3 feet or more away to my ears. Now I wonder if this is where my vertical aligment of the midbasses will help. With a vertical alignment I can push the woofers up toward the front of the door(near the jam) and thusly gain about 5 more inches of distance. . Also If I do a MMT design Vertically, I wil have a closer Center to center spacing. which should help with my FR response and allow more flexibility in crossover selection. I really wish I knew of a local person with a "real" good set of ears. Then I could build both designs and demo them. As it sits, my wife is the best audiotioner I have.
 

·
Registered
Joined
·
1,301 Posts
So it's down to the wiring! I got everything installed tonight and will be wiring it all up on Thursday. I also managed to go ahead and replace the dustcaps on my JBL 1000GTi's that I'll be using. I'll try and get a better picture in the morning before heading to work. Hopefully I'll have some daylight to work with.








The closest i've seen to this is PIONKEJ's build its also where i got the inspiration for my current build.
 

·
Premium Member
Joined
·
6,034 Posts
First: Patrick, Thank you for providing your experiences and information for this project. My major concern was where center to center spacing. Without your Input, I would never have known Distance effects summation.


I wonder how the RS150 would work. these daytons have a smaller cone than most equivilant drivers. have you worked with the RS150 or something similar? what kind of crossover/ midbass response should I expect?

Also, Where did you get woofer sizes? I did R=SQRT(126/pie) for the RS180 as SD=126CM. That gives 6.34=r, and D=12.7. Fountek is D=5.9. Thats 18 or 180 cm away(6 Feet). With seating and speaker placement I will be 3 feet or more away to my ears. Now I wonder if this is where my vertical aligment of the midbasses will help. With a vertical alignment I can push the woofers up toward the front of the door(near the jam) and thusly gain about 5 more inches of distance. . Also If I do a MMT design Vertically, I wil have a closer Center to center spacing. which should help with my FR response and allow more flexibility in crossover selection. I really wish I knew of a local person with a "real" good set of ears. Then I could build both designs and demo them. As it sits, my wife is the best audiotioner I have.
I'm building an array for my car at the moment, so I answered your questions in my forum:

Audio Psychosis • View topic - Under Dash Horn Alternative

It's a long read - took me nearly two hours to type. But you're asking a lot of good questions, so I thought it would be interesting to explore how arrays work!
 

·
Registered
Joined
·
1,301 Posts
PB's post is a REALLY good read. it touches on a lot of critical stuff.
Patrick Bateman said:

If you screw around with arrays for a while, you'll start to see that using the same size for all the elements simplifies the design a great deal. For instance, it's very hard to get the array to behave if you use a 17cm woofer and a 2.5cm tweeter, but it's easy to get the array to behave if you use a 8cm woofer and a tweeter in an 8cm waveguide. (The waveguide changes the directivity of the tweeter so that it matches the woofer, in addition to all the other benefits.) If you look at the Snell array, you'll notice that Smith used a waveguide that's the same size as the midranges. That's by design - it simplifies everything.
This is what guided my final decision to go with multiple 3 in full range speakers for my front and center stage. PB is a smart cookie.
 

·
Registered
Joined
·
116 Posts
Discussion Starter #13
Quietfly: The only issue I see with the multiple fulrange driver setup is the lack of lowend output. Sure you can tune an enclosure but that is beyond waht I have planned for my doors.

Pat: Couple of points and maybe some more picking of your brain.


"Because the crossover is so critica"l, the obvious solution is to simply use extremely steep crossover slopes. Taken to the extreme, with an infinite crossover slope, the elements in the array would not interact with each other."
I will be using LR4 active slopes. I like these for power handling and there is no phase change to my understanding.​

"All loudspeakers are arrays. This may seem non-intuitive, but all loudspeakers are arrays. For instance, a simple two-way loudspeaker is an array./B]

This is why my project isnt a bad idea. a typical three way setup is still an array and realistically I can achieve much better CTC distance with my dual mids. Can I achieve the same mid-bass though​

"if you flip an array on it's side, the distance changes, because the vertical height of the array changed."

This is the part I now am trying to figure out. I understand I may gain distance from the listener if I make a vertical array vs the horizontal. It would be worth about a foot in distance. But how will my height affect my response(theory of course) and how might my image get affected. I figure my image will sit hi on the door as my array will be 16 inches vertical. Now if I make this a MMT setup, i will avoid the off axis FR response issues that MTM designs commmly have. I also opine that the height of the vertical MMT will make up for whatever heightening effect the horizontal MTM might have provided. let me know​
 

·
Registered
Joined
·
116 Posts
Discussion Starter #14
I have found an interesting read on D' appolito arrays and my math is good but my understanding of theta in the followin is lacking, any thoughts?

fx = c/2 * |{[(L cos [theta])2 + (d + L sin [theta])2]1/2
- [(L cos [theta])2 + (d - L sin [theta])2]1/2}-1|

where fx is the lowest frequency of total cancellation in Hz, [theta] is the vertical displacement (ranging from -90 deg. to +90 deg.), d is the vertical separation between the high-frequency driver and one low-frequency driver, L is the distance from the plane of the drivers to the on-axis design point, and c is the velocity of sound. For a given d, as [theta] increases in magnitude, fx decreases. Clearly, if d and [theta] are sufficiently large, it is possible that fx will occur at a frequency where the output from the high-frequency driver is negligible, and the system response at that frequency at that listening angle will effectively drop to zero.

IF anyone can fill me in that would be appreciated , I am assuming that Theat may be the OF AXIS ANGLE? so, in a car and with the array in the doors, that may give a resultant theta of pie/4 or possibly pie/3. This also asssumes the above formula is in radians.I understand that the instructions say degrees but I have never calculated a theta in degrees outside of geometry. Also, L would be measured distance from the door panel to the my ears??? plz let me know
 

·
Registered
Joined
·
1,301 Posts
Quietfly: The only issue I see with the multiple fulrange driver setup is the lack of lowend output. Sure you can tune an enclosure but that is beyond waht I have planned for my doors.
I have mine running from 190 up and everything lower being fed to my L6se's.i have no issue with low end. My original plan was for MTM setup for the L4se's but it just was not do-able in my car. so i switched to using 3 pairs of Dual 3 inch drivers and having L6se's in my doors.
 

·
Premium Member
Joined
·
6,034 Posts
Quietfly: The only issue I see with the multiple fulrange driver setup is the lack of lowend output. Sure you can tune an enclosure but that is beyond waht I have planned for my doors.

Pat: Couple of points and maybe some more picking of your brain.


"Because the crossover is so critica"l, the obvious solution is to simply use extremely steep crossover slopes. Taken to the extreme, with an infinite crossover slope, the elements in the array would not interact with each other."
I will be using LR4 active slopes. I like these for power handling and there is no phase change to my understanding.​
For the past year or so, I've been learning all I can about phase response. I have *personally* noticed that getting a loudspeaker to perform better in the time domain makes an audible difference. Having said that, the situation will likely vary for every listener, and some will be more sensitive to phase than others as.

As for the phase change of high order crossovers, this is a known problem. High order crossovers are notorious for screwing up the phase response. Companies like Thiel and Dynaudio have been saying this for years. Siegfried Linkwitz, the co-inventor of the LR4 crossover that's so popular, has said I indicated my preference for a low order crossover between woofer and midrange, because of its reduced phase distortion. In the extreme this would argue for a 6 dB/oct 1st order acoustic crossover, which is exceedingly difficult and costly to realize, because it places stringent demands on linearity and frequency response of the drivers used."

That quote is from Linkwitz' site : Phase distortion

In my situation, I used to use 4th order crossovers in the 90s, and slowly switched to lower order crossovers because I could never get 4th order crossovers to sound 'right.'

A big part of this was likely because the 4th order crossovers are less forgiving. (These were passive, and due to the high number of components, they're more prone to strange interactions with the driver's inductance.)

But as I learn more about phase, I'm noticing that low order crossovers are quite usable as long as you get the drivers very VERY close together. IIRC, Smith used '3rd order acoustic' on the Snell.

When I say '3rd order acoustic', I mean that the combined response of the drivers and the crossover forms a 3rd order slope. This is one of the less obvious things about crossovers; if you use a 2nd order crossover but the driver is already rolling off, you will end up with a final crossover slope which is much higher.

If you read the white papers on Linkwitz Riley crossovers, a lot of the benefits that are mentioned depend on the drivers having perfectly flat response. In the real world, we're usually crossing drivers over at a point where they're rolling off already, so the combined slope is never the same as what we'd intended.

This is one of those things that DSP solves very nicely. You can basically 'dial in' a crossover slope and delay. It's pretty incredible, but *does* require a microphone and some patience.


"All loudspeakers are arrays. This may seem non-intuitive, but all loudspeakers are arrays. For instance, a simple two-way loudspeaker is an array./B]

This is why my project isnt a bad idea. a typical three way setup is still an array and realistically I can achieve much better CTC distance with my dual mids. Can I achieve the same mid-bass though​

"if you flip an array on it's side, the distance changes, because the vertical height of the array changed."

This is the part I now am trying to figure out. I understand I may gain distance from the listener if I make a vertical array vs the horizontal. It would be worth about a foot in distance. But how will my height affect my response(theory of course) and how might my image get affected. I figure my image will sit hi on the door as my array will be 16 inches vertical. Now if I make this a MMT setup, i will avoid the off axis FR response issues that MTM designs commmly have. I also opine that the height of the vertical MMT will make up for whatever heightening effect the horizontal MTM might have provided. let me know​


If you use very tight spacing on an array, you can get it to behave like one big driver. For instance, with the Snell array, the two midranges and tweeter are basically functioning as one wideband driver.

There's a big caveat though - the spacing must be very tight, and the crossover must be chosen carefully. Spend some time with the array simulator from FRD Consortium and you'll see what works and what doesn't.

But the same rules of physics still apply; the transition from nearfield to farfield gets higher and higher as the array gets larger and larger. This applies to ALL arrays, not just MTMs. It applies to TMs too.

And when you play with the array simulator from FRD, you'll see that arrays have wide directivity HORIZONTALLY, but narrow directivity VERTICALLY.

Now flip the array on it's side, and you just flipped the lobes.

Now the VERTICALY directivity is wide, and the HORIZONTAL directivity is narrow.



This is what I am listening to right now. It's an MTM array from JBL which works on either axis.

If you go on the home audio forums you'll hear a lot of people complain about 'MTM lobing.' Just keep in mind that there are situations where we WANT the loudspeaker to lobe. The WHOLE REASON the JBL has narrow horizontal directivity when layed flat is because of lobing.

For instance, what if the loudspeaker is in a room where the sidewalls are too close? Narrow horizontal directivity can improve your imaging by reducing reflections off the side walls.

Or what if the loudspeaker is in a room where you can't easily place it at ear-level? Then wide vertical directivity can compensate for a loudspeaker that's placed too low.

Are you seeing where I'm going with this? :D

Our cars are basically a room where the loudspeakers are too close to the sidewalls, and too low. It's for this reason that horizontal MTM arrays can work nicely. In car audio, we spend way too much time listening to the home audio types. The problems in car audio are a lot different than home audio.

The JBLs have such wide vertical directivity, I find that they're pretty darn listenable on the floor or on the ceiling. When placed on the floor, the soundstage does NOT seems to be that low. The key is to use a boundary to good effect (corner loading.)

 

·
Premium Member
Joined
·
6,034 Posts
I have found an interesting read on D' appolito arrays and my math is good but my understanding of theta in the followin is lacking, any thoughts?

fx = c/2 * |{[(L cos [theta])2 + (d + L sin [theta])2]1/2
- [(L cos [theta])2 + (d - L sin [theta])2]1/2}-1|

where fx is the lowest frequency of total cancellation in Hz, [theta] is the vertical displacement (ranging from -90 deg. to +90 deg.), d is the vertical separation between the high-frequency driver and one low-frequency driver, L is the distance from the plane of the drivers to the on-axis design point, and c is the velocity of sound. For a given d, as [theta] increases in magnitude, fx decreases. Clearly, if d and [theta] are sufficiently large, it is possible that fx will occur at a frequency where the output from the high-frequency driver is negligible, and the system response at that frequency at that listening angle will effectively drop to zero.

IF anyone can fill me in that would be appreciated , I am assuming that Theat may be the OF AXIS ANGLE? so, in a car and with the array in the doors, that may give a resultant theta of pie/4 or possibly pie/3. This also asssumes the above formula is in radians.I understand that the instructions say degrees but I have never calculated a theta in degrees outside of geometry. Also, L would be measured distance from the door panel to the my ears??? plz let me know
I can't seem to find the JBL paper right now :(

I use the 'rule of thumb' because the math on the nearfield transition is too much to do in my head. "ten times the array height" is easy for me to calculate :D

The array simulator from FRD Consortium is awesome.

For instance, a lot of people in car audio obsess about aiming their drivers. But look at Gary Biggs - his drivers are pointed waaaaaaaaay off axis. Over 45 degrees in fact.

So what's the deal with Gary Biggs?

I'm guessing that Gary designed his system so that you're listening off-axis.

In other words, there's absolutely no reason you must listen to you loudspeakers on-axis. It's perfectly valid to optimize the crossover for off-axis listening, in fact, there's a lot of advantages to listening off axis.

The array simulator from FRD can help a lot with this - you can easily tweak driver spacing and crossover slope to get a good response off-axis.

But this is one of those things that's damn near impossible to do without a computer. There's waaaaaaaaaay too many variables.

When you play with the simulator from FRD, you'll notice that EVERYTHING matters:

  • The driver spacing matters a LOT. Move the drivers an inch and the response changes.
  • The directivity of the drivers matters. You can simulate a waveguide-loaded tweeter by simply changing the diameter of the tweeter. (a 3/4" tweeter in a 3" waveguide basically has the same directivity as a 3" driver.)
  • The crossover slope matters a LOT
  • flipping the polarity of the drivers has some unexpected effects, sometimes positive
  • Most importantly of all, the location of the listener is paramount. In an array, the orientation of the array changes the sound dramatically. This effect grows more extreme as the center-to-center spacing grows. But there are benefits to this, as noted in my last two posts. You can basically 'narrow' the sound on one axis.
 

·
Registered
Joined
·
116 Posts
Discussion Starter #19



And when you play with the array simulator from FRD, you'll see that arrays have wide directivity HORIZONTALLY, but narrow directivity VERTICALLY.

Now flip the array on it's side, and you just flipped the lobes.

Now the VERTICALY directivity is wide, and the HORIZONTAL directivity is narrow.



This is what I am listening to right now. It's an MTM array from JBL which works on either axis.

If you go on the home audio forums you'll hear a lot of people complain about 'MTM lobing.' Just keep in mind that there are situations where we WANT the loudspeaker to lobe. The WHOLE REASON the JBL has narrow horizontal directivity when layed flat is because of lobing.

For instance, what if the loudspeaker is in a room where the sidewalls are too close? Narrow horizontal directivity can improve your imaging by reducing reflections off the side walls.

Or what if the loudspeaker is in a room where you can't easily place it at ear-level? Then wide vertical directivity can compensate for a loudspeaker that's placed too low.

Are you seeing where I'm going with this? :D

Our cars are basically a room where the loudspeakers are too close to the sidewalls, and too low. It's for this reason that horizontal MTM arrays can work nicely. In car audio, we spend way too much time listening to the home audio types. The problems in car audio are a lot different than home audio.

The JBLs have such wide vertical directivity, I find that they're pretty darn listenable on the floor or on the ceiling. When placed on the floor, the soundstage does NOT seems to be that low. The key is to use a boundary to good effect (corner loading.)

due to my car being a sports coupe, My setas place the Seakers very close to level with the driver and passeger. For this reason, I am doing the Vertical array. My off axis response will be superb all the while my vertical reflections will be reduced. Now if I had an SUV, the horizontal array would be better. Check out the build log I am doing on here. I plan to wrap the drivers in clothe and test the best placement.

I cant use the simulator ont hat site as I run linux and Wine does not wanna run it. So...you see my equation.
 
1 - 19 of 19 Posts
Top