2 way vs. 2.5 way vs. MTM (1 Viewer)

[Phil M]

Just a general question I guess about monitor designs.

What are some inherent advantages/disadvantages of a good 2 way monitor (like the AV1), vs. a 2.5 way monitor (like a Paradigm studio 40) vs. an MTM design like the Seas Odins or AV1+)?

I'm not real concerned with low end extension, but more for accuracy, openness, and detail.

I just didn't know if one design has specific advantages over another.

Phil

 

[AndersP]

Hi.

I just checked your Jaguar project out. It seems to me that you´ve become quite a handy - man. I really don´t know what I´m most impressed with, your carpentry or the webpage.
Keep it up!!

2 - w monitors: good imaging, comp. uncomplicated filter and box design, good imaging and holography and low price but also suboptimal low bass extension if good offaxis performance is on your wishlist ( size of midbass driver )

2,5: Bigger box that is more difficult to brace and damp, more drivers, more complicated filterdesign ( esp. the big inductor for the low bass ) but they also give more low end extention and take more power before failure

MTM: like 2w with more drivers and more power tolerance. Plays louder ( more efficiency )Differs in horisontal ( wider ) and vertical ( less wide ) radiation. Used in THX - applications because of this more controlled dispersion pattern

 

[Phil M]

AndersP

Thanks for the reply, and the compliments!

I was beginning to think that it was really a dumb question, or nobody new an answer?

I'm just getting into DIY and was wondering just how much effect (if any) the design of the speaker had on its sonics and output.

Phil

 

[Patrick Sun]

Okay, more thoughts on the different types of speaker configuration:
2-way: usually a tweeter and a midwoofer, depending on the baffle step compensation (BSC), you end up with a speaker whose sensitivity can be a bit on the low side if done properly. A lot of 2-ways are not done properly, and sound thin on the bass, and forward on the midrange and highs, but will claim to have a higher sensitivity to impress novice speaker buyers. Most likely, depending on crossover points and driver roll-offs, the target acoustic slopes can vary from 1st order to 6th order (though that's pretty drastic), commonly, you'll see either 2nd order or 4th order acoustic slopes.
2.5-way: like a 2-way, but the baffle step compensation is achieve by a lower "0.5" midwoofer that only contribute up to around 500Hz of output from the speaker (you will usually end up rolling off the lower 0.5 midwoofer sooner and will work hard in integrating and blending its output with the 1.0 midwoofer. This allows for a bit more sensitive speaker because you are not shelving down the main midwoofer's output like you would in the 2-way. The crossover target slops is pretty much the same was a 2-way (i.e. dictated by drivers and crossover point).
MTM: usually in the vertical orientation, the tweeter is sandwiched by 2 identical midwoofers. The crossover is effectively a 2-way, and baffle step compensation is required, but you gain almost 3dB sensitivity overall with this setup (wiring up 2 midwoofer in parallel will net you 6dB in output, but BSC will claim around 3dB of that gain). The crossover in a MTM will usually target 3rd order acoustic slopes between the tweeter and midwoofers to achieve the correct lobing characteristics of the output of the speaker.
3-ways: usually feature a large woofer that is used to fill in the baffle step loss, and the midrange and tweeter are used like a 2-way without needing to worry about the BSC. It's a bit tougher to integrate all 3 drivers and yield a good sounding speaker due to the interaction of the 3 drivers and the chance of comb effects in the overlap of frequency regions for each driver.
Center-to-center (ctc) driver spacing plays a part in selecting the highest crossover point (rule of thumb for some people is to keep the crossover frequency under the wavelength dictated by the ctc distance (in terms of inches, frequency = 13585/(ctc distance)). This is just a guideline, the world won't come crumbling down if you miss the ctc distance/XO frequency target.
There are diffraction effects from the placement of the drivers on the front baffle, you can minimize those diffraction effects by not placing the drivers symmetrically on the baffle (i.e. not on the center line, but offset but a little). There's a pretty cool Baffle Diffraction Simulator that you can play with, though I think it requires the latest version of Excel to run it. There's a bunch of other tools from DIY speaker enthusiasts looking to create a unified approach to DIY speaker design (from simulating boxes and its bass response, to crossover filter designs, etc).
More often than not, a tweeter's diffraction can be minimized by recessing the tweeter so that it's flush-mounted to the baffle. The woofer isn't as critical, and it's up to you if you want to flush-mount the midrange/midwoofers (doing so may decrease the relative acoustic centers of the tweeter with the other drivers, though).
Lining up the acoustic centers leads some people to slant the front baffle so they all line up relative to each driver and the listener. The acoustic center is the center of the driver where the sound emanates from, and for tweeters, it's relatively close, but for midrange/midwoofers and woofers, their acoustic center is further back from their surround material/ring near the voice coil of the driver. Thus the design consideration for the slant in some baffles (like the Jaguars).
Also, to combat diffraction effects, rounding over the front baffles edges is another technique that people use to smooth out the response of the tweeters. 3/4" roundover being the minimum for any benefit from rounding over the edges (1" roundover being more of a starting point, but that's a bit roundover bit!)
To smooth out the bass response (to get rid of boominess), people will stuff the enclosures with polyfill or fiberglass to slow down the soundwaves and make the enclosure "appear" larger than it is. How much to stuff is entirely an experimental exercise for most cases, but the rule of thumb is that for a 100% fillrate, you are using 1 pound of filling per cubic foot of enclosure volume. You can experiment with 25% fillrate, 50%, 100%, etc. to tune in what best sounds good to you.
I haven't even touched on all the knickknack filters and circuits you can use in a crossover to achieve the target slopes at the crossover frequencies, while taming peaks in drivers that otherwise would be too peaky.
There's a bunch to learn about DIY speakers, but it can be a bit expensive in terms of buying drivers, and test gear, and finally XO design software. But it can able be a cool hobby if it gets you up out of bed on a Saturday morning, and you find that you enjoy making a lot of sawdust to build that really neato speaker enclosure that maximizes the performance of the drivers in order to integrate them into the entire speaker system.
Phil, BTW, nice to see another photo-documentation nut on the forum.

 

[Phil M]

Patrick:
Thanks for the info, it is basically what I was looking for.
And like I said earlier, your site is what inspired me to do this in the first place. I'm hoping that maybe mine can encourage someone else to make some sawdust too!

 

[Brian Bunge]

Pat,

Great information! I've also seen 2.5-way MTM's. Two examples are the PSB Stratus Bronze and Silver speakers. Both have MTM aligned drivers but are also 2.5-way speakers. The 1.0 midwoofer is the top one and the .5 midwoofer is the bottom one in each speaker.

Brian

 

[Phil M]

Brian:

I think that the new Polk Lsi9 speakers are the same way in that they are 2.5 way in an MTM configuration.

I'm hearing great things about the Seas Excel W18E001 woofers and XT25 & Hiquphon OW1 tweeters.

I was toying with the idea of building something with them, but I think that crossover design may have way too steep of a learning curve for me.

Dave Ellis's 1801's use the W18 & OW1 in a (2) way, and the Seas Odin Mk3 kit, uses a pair of W18's in a MTM design with a Seas tweeter.

Anyone ever actually listened to any of the above drivers or kits?

Phil

 

[Ron D Core]

So would an MTM design also benefit form the slanting (like the JMLabs setups)?

 

[AndersP]

There are different opinions about it. JM-lab claims that they get the drivers centers of radiation more aligned by this slanting. Others claim that this solution can get you mere diffraction problems. It´s probably a question of taste.
Phil: Have you ever considered going active with your project? The SEAS magnesium driver are a bit demanding designing passive filters for, because of the nasty breakup medes, though it sure can be done successfully with alot of measuring.

 

[Greg Monfort]

>There are different opinions about it. JM-lab claims that they get the drivers centers of radiation more aligned by this slanting. Others claim that this solution can get you mere diffraction problems.
====
You get both. If the radius of the arc matches the distance to the listening position and it also matches the requirement of time alignment at the XO points, then you get an on axis ~ acoustic phase coherent mimic of a point source driver. The downside is increased off axis comb filtering, which all combines to make for a fairly limited 'sweet spot'. The trade off is plenty acceptable for me in most applications.
If you just pick a curve that looks good to you and don't address the phase issues at the XO point(s) you get at best a ho-hum performing speaker and are better off sticking with a conventional MTM IMO.
GM

 

[AndersP]

Greg: Very well put and I of course totally agree, but I have 2 snobbish remarks:

1. Time alignment at XO points tend to be less important with higher order filters. Dickason states that it is irrelevant with a 4:th order. You just need to change the phase of the drivers to match.

2. I read somewhere that the canadians had a state funded agency that had done some serious psycho acoustic research of the importance of on- and off axis sound. Conlusion was that the off axis sound was more important for " good sound " than the on axis sound, according to a test audience.

Wish I could write english like you guys.

 

[Brian Bunge]

Anders,

I think you are referring to the NRC (National Research Center) in Canada. They allow speaker manufacturers to use the facilities for free, I believe. I know Paul Barton of PSB has used it for extensive testing.

Brian

 

[AndersP]

Brian,

Exactly, I think I saw it in an Energy Veritas brochure a year or to ago. Maybe an explenation to why they stand so strong in speaker technology as they do.

\Anders

 

[Greg Monfort]

>Greg: Very well put and I of course totally agree, but I have 2 snobbish remarks:
====
Hmm, not 'snobbish' IMO, just counterpoints.
====
>1. Time alignment at XO points tend to be less important with higher order filters. Dickason states that it is irrelevant with a 4:th order. You just need to change the phase of the drivers to match.
====
Tell that to a prosound speaker designer. If you're lucky, he'll be polite. Anyway, let's look at some facts, then you decide:
A 4th order slope represents a 360deg or 1Hz phase shift between the two source's acoustic centers for a given frequency, so the first thing you have to do is find this to get the initial setback, then calculate the additional setback due to the XO's electrical phase rotation:
(speed of sound/crossover frequency)/(360/required phase).
So for a typical two way XO point of 2.2kHz: (13560/2200)/(360/360) = ~6.16" + whatever the difference is between the HF's acoustic center and the midbass driver's is at this frequency. Time wise this probably represents

 

[AndersP]

OK Greg I´m more or less with you all the way, but I feel that I need some more time to provide a reply up to your standards. It´s to be about parts of the 6 point something inches. You must be a very good typist.

I´ll be back!

 

[AndersP]

Hi again Greg.

For the first time on this forum, I´m now writing off line, as I sensed the seriousness in your last post. Unfortunately I´m of course gambling with the risk of the thread becoming cold and that nobody will read this, but I still think that your post demands a response from my part.

Thank you for sharing some of the maths behind even order filters phaze caracteristics, but most of it I already know. It´s possible I was a little unclear in my statement about higher order filters to achieve an interesting debate, since I feel the control freak in you Greg.

Time alignment doesn´t TEND to be less important, It IS less important the higher the order of the filter because of the reduced overlap. All this is of course mathematics, which on its own is pretty uninteresting as you are implying. I believe it´s called the electric transfer function, which soner or later, combined with the acoustic transfer function of the drivers in their baffle, comes in contact with reality. ( if you´re lucky )

Here´s where the 6,16” comes in. I made my first ”Tom Danley experiments” years ago and this to is very familiar to me, but thanks again for sharing.

I was a little puzzeled that you ” proved” my point for me with this (low order) example:

>At the Atlanta DIY2001, someone brought in some beautifully made speakers with 2nd order XOs. He had the tweeter connections reversed just as he's supposed to since the HF leads the LF by 180deg, but the HF's response was shelved down! When they were wired in-phase they gave a flatter FR. IOW, textbook filter design not only assumes a resistive load, but an *acoustic phase (time) correct* one.
==============

Excellent!!

Baffle design is only one of an aboundance of different means to an end in solving time problems in speaker design. They are described in detail in every good text book on speaker design on the market, not just Dickasons, so I´ll not tape it in here. If this was not the case the market would be full to the brim with stepped and slanted baffles and it isn´t.

>Tell that to a prosound speaker designer. If you're lucky, he'll be polite. Anyway, let's look at some facts, then you decide:
==============

Well Greg, prosound ( and vacuum tube fans ) have another problem to deal with ( in a perfect world ), namely phaze problems with integrating horn amplified drivers in their rigs/systems. The prosound people that I know and worked with (in my teens) solved their problems with multiple boxes, stacked on each other. Guess who had to adjust the phaze by moving the boxes around? ( I was in highscool and didn´t know better )
This is another magnitude of phaze problems that I don´t want to dig deeper in to, because of lack of theoretical experience, though I still feel it in my back sometimes.

But all the same, after having seen page after page with almost catastrophic results of measurements on different Klipsch speakers in scandinavian magazines, but all over, top results in the listening tests and of course amazement and respect for the sensitivity they provide, you start to wonder.

It seems to me that the leaders in design today, more and more start to design their projects of axis, Revel is one example, and for a good reason. This is my point.

Thank you very much for the kind words on my english. Sometimes I loose you Greg, but then it´s when you use short forms that I should have known if I´d hung out here a little more.

\Anders