How to design a MTM speaker

[nikbrewer]

ok, i ordered the alesis 5.25" woofers and dome tweeters from clearence. A forum member is going to measure them for me. I figure this would be a good learning project on a MTM. how do i do this. do i have the drivers measured raw? or do i build a box and send it? dont i need a cross ove point to determine woofer spacing? also, why are tweeters off set, to allow the woofers closer together, or other reasons?

Also, i ried to do a sim in Win ISD, and it showed a 3 cubic foot box for a ported, and a wierd spkie in the low end. did i do somethign wrong, can someone confirm?

[philiparcario]

tweeter's dont have to be offset in mtm's. measurements should be freeair. then you will know the real vas qts fs

[nikbrewer]

oso, then from that point, i could design a crossover, break them in, then mail it off to be measured as a whole speaker?

[HareBrained]

Certainly, if you can measure later in the process, the less you have to simulate. But as it stands right now, differences between drivers (especially if from different lots) can be greater than the error in the simulations. This is demonstrated by Zaph's designs and his copious documentation. Measurements in the box means you'll know exactly the effects of the baffle step diffraction, edge effects, and vent/woofer response (assuming near- and far-field measurements are taken and then merged correctly).

Regarding offset drivers, you can see the effects in Response Modeler's BSC section. Offsetting a driver results in different edge effects that are mostly centered around 1000Hz (give or take, depending on the baffle width). For a tweeter, the impact decreases as the Fc increases. For something like the ND28, where there is already a boost between 1k and 3k, it is actually significant.

It's my understanding that offsetting the tweeter from the woofers also "tilts" the sweet spot from directly in front of the speaker. But I could be wrong.

[jclin4]

Nik,

lot's of useful info on offsetting tweeters in this thread, including discussion of whether to offset in an MTM or not:

http://techtalk.parts-express.com/showthread.php?t=205442

This specific post in the thread provides guidelines/rules for tweeter placement:

http://techtalk.parts-express.com/showpost.php?p=1508993&postcount=12

Also, play with the baffle edges in Frequency Response Modeler (i.e. no edge vs. chamfer vs various roundover radius') and note the diffraction effects on the tweeter response. It's pretty interesting and might influence your baffle/cabinet design.

- John

[Kapton Planet]

ok, i ordered the alesis 5.25" woofers and...

Also, i ried to do a sim in Win ISD, and it showed a 3 cubic foot box for a ported, and a wierd spkie in the low end. did i do somethign wrong, can someone confirm?

Those woofers have a high Qts not well suited to a vented box, thus WinISD can't find an optimum size. If you must port them, you can sim a more reasonable box size but you will always have some kind of peaking or boominess.

[nikbrewer]

well, ill get them tested free air, then play with the crossover spreadsheet and see what i get. Worst out come is im out 45 dollars for drivers, no biggie. So should i go sealed then? looks like 2 in a 1.25 cubic foot box is pretty flat with a f3 of about 76 hrz

[Chris Roemer]

ok, i ordered the alesis 5.25" woofers and dome tweeters from clearence. A forum member is going to measure them for me. I figure this would be a good learning project on a MTM. how do i do this. do i have the drivers measured raw? or do i build a box and send it? dont i need a cross ove point to determine woofer spacing? also, why are tweeters off set, to allow the woofers closer together, or other reasons?

Also, i ried to do a sim in Win ISD, and it showed a 3 cubic foot box for a ported, and a wierd spkie in the low end. did i do somethign wrong, can someone confirm?

You take the S.O.S. (13,500 ips), divide it by your C-C spacing between your tweeter and woofer (center of 5.5"d midbass to center of 3.75"d tweeter is 4.625", with frames touching), and you get your highest recommended Fc, which = 13,500 / 4.625 = 2.9kHz. You should also cross to a tweeter at least one octave above its Fs (1500 here) which puts it at 3.0kHz. So if you set your Fc at 2.95kHz, you'll be marginal on both accounts. You'll have to pummel your response curves to cross at 2.9 to 3.0 kHz. And you have to got at LEAST 2nd order on the tweeter when crossong it that low.

Also, you picked some midbasses with a Q (around .71?) that's meant for a closed box. Vented designs seem to like (IMO) a Q more in the 0.30 to 0.43 range. High Q, big box (with unsightly rollercoastering in a vented alignment). So I'd say you're running your WinISD alright.

Chris

[nikbrewer]

ok, so sealed box it is. So are these drivers not a good match for a mtm? would another tweeter be a better option? maybe a nother set of woofers to make some MT with the tweeters? I am all very nwe to this, and plan on just building these to put in the garage maybe. so i need to do a second order crossover at aroung 3k hetrz? does this sound doable? I dident understand on what my spacing needs to be.

[nikbrewer]

ok, so can i cross higher then that? or would they need to be close or farther apar to do that? Im taking notes on this as well.

[Æ]

ok, i ordered the alesis 5.25" woofers and dome tweeters from clearence. A forum member is going to measure them for me. I figure this would be a good learning project on a MTM. how do i do this. do i have the drivers measured raw? or do i build a box and send it? don't i need a crossover point to determine woofer spacing? also, why are tweeters off set, to allow the woofers closer together, or other reasons?

Also, i ried to do a sim in Win ISD, and it showed a 3 cubic foot box for a ported, and a wierd spkie in the low end. did i do something wrong, can someone confirm?

Why does everybody think they have to "space" the woofers?
Mount them together as close as possible. Period. Even overlapping the tweeter flange if possible.

Tweeter offset is to ameliorate diffraction. Remember that the front baffle is relatively wide compared to the shorter wavelengths of treble frequencies.

[nikbrewer]

what if i want some room between the drivers, as i like the look sof this, like the tritrix. basically just get them as close as possible, but the tweeter over a little bit and call it a day?

[Æ]

what if i want some room between the drivers, as i like the look sof this, like the tritrix. basically just get them as close as possible, but the tweeter over a little bit and call it a day?

Good idea.
Just remember though, "good looks" seldom has anything to do with "good sound." Aesthetics is pleasing to the eyes. But we hear with our ears!

[nikbrewer]

thats why i want to know why the spacing matters, so i can try to do both. Just saying make them touch menas nothing if i dont know why they need to be close

[philiparcario]

also nik if these need to be 2 in a sealed 1.25 cu ft box with an f3 of 76hz that does not mean they will sound bad. with a sub crossed at 80hz they will probably sound good. Since you can make boxes with ease so to speak make one pair at 1.25 cu ft the so called optimum size and make one pair at .9 cu ft. you can give yourself lessons in aperiodic venting. stuffing with poly-fil, that is a way to make a box seem bigger. you can hear for your self if you like a bit of boom in your speaker. etc.

[dthomas]

thats why i want to know why the spacing matters, so i can try to do both. Just saying make them touch menas nothing if i dont know why they need to be close


The reason the spacing matters is that the woofer above and below the tweeter will interfere with each other in the vertical plane. The good advice you have gotten here is to keep the spacing no more than 1 wavelength in distance hence a crossover point of around 2900Hz. While it is agreed that this is the maximum acceptable distance it will still exhibit anomalies in the response as you move above and below the tweeter axis in the vertical plane. So it is imperative that you place these so the tweeter is at ear level. The ideal scenario is to use a tweeter than can cross at about 1/2 a wavelength or in the 1500Hz range this will greatly reduce the interference issues in the vertical plane.

Another option if the tweeter you are using requires a 3000Hz crossover point is to do a WWT 2.5 way design. This way the lower woofer only assists the top woofer in the frequency range below baffle step where it is most needed and since only one woofer is playing up to the tweeter you won't have the issues you will in an MTM. You also net about the same final sensitivity with this design approach when compared to an MTM.

2900Hz is a very high crossover point for an MTM and I think most here would agree that shooting for a lower crossover point of below 2000Hz (the lower the better) is a good practice. You really need very small woofers and a small format tweeter to get the spacing close enough for a 2900Hz crossover point. If you were asking me I would do it would be a WWT 2.5 way with that driver combo for the optimum final result. I have used that Alesis tweeter very nice tweeter for the price. Just needs a crossover point of around 3000 to 3500Hz.

[nikbrewer]

Thanks for the suggestion. I like the idea of a WWT 2.5. i will mail offf the speakers to be measured on friday, then i will post up that info and we'll see what we can do

[Æ]

thats why i want to know why the spacing matters, so i can try to do both. Just saying make them touch means nothing if i don't know why they need to be close

Spacing does matter, the closer the better.
That is all you really need to know!
Keep It Simple. . . .

[nikbrewer]

ok, you are talking to the guy who builds cnc's in his garage, do you think i keep things simple?

[Æ]

ok, you are talking to the guy who builds cnc's in his garage, do you think i keep things simple?

I don't really know. ;)
Technology often simplifies. Labor saving devices!
Simple enough if you are all setup and know how.
I was thinking of having you do some baffles for me too.
For whatever reason many people think that MTM's should be spaced out to a certain distance. In fact they should be de-spaced!

[nikbrewer]

i just liek knowing therory. now, if i was using 2" mids and a 1" tweeter, would i jam them all together to? How about in a mmt, do i space the upper m and t like a 2 way? where do i put the other driver in relation? i know thay are not jammed togther in the pics i see

[Æ]

i just like knowing theory. now, if i was using 2" mids and a 1" tweeter, would i jam them all together too? How about in a mmt, do i space the upper m and t like a 2 way? where do i put the other driver in relation? i know they are not jammed together in the pics i see

It's not how physically close the woofers and tweeters are, but how acoustically close they are. Two large woofers separated by several inches are acoustically close because of the rather long wavelengths involved, assuming that you are crossing over in the bass region.

Tweeters can be physically close together but still be some distance apart acoustically, because of the much much shorter wavelengths.
Drivers that are separated end up with poorer lobing patterns especially in the crossover region and with multiple drivers, you will get comb filtering.

Really never any problem having all your drivers as close together as possible, except for the manufacturing, cutting of the holes in the baffle. It can be quite tricky cutting two different holes/rabbets so that the woofer flange can overlap the tweeter flange. That is one reason why there are now tweeters manufactured with truncated face plates.

Of course aesthetics is important, especially for a commercial product. Strategically spacing the drivers looks good, but the farther the drivers are apart the more difficult it is optimally placing the loudspeaker in the listening environment. You either have to raise, lower or tilt the loudspeaker to exactly the right position in order to achieve the preferred "sweet spot."

One way to get drivers physically close together is to cut their flanges!

[billfitzmaurice]

Why does everybody think they have to "space" the woofers?
Mount them together as close as possible. Period.+1. Even Joe D'Appolito does it that way. If it sounds good it is good. OTOH determining the correct crossover from the midbasses requires measuring them in the cab to determine the -6dB at 30 degrees off-axis frequency, which is as high as you want to go.

You should also cross to a tweeter at least one octave above its Fs
+1 with a 2nd order HP, but as you increase slope you can decrease the Fc. If you really must get down to fs x 1.2 you can do it with a 5th order high pass. That's Neville Theile's preference. Yes, that Neville Theile, not to be confused with a speaker manufacturer of the same last name.

[arlis_1957@yahoo.com]

just a thought. build the bigger box and dato in a shelf to achieve the smaller cf. if it has a removable back the shelf can be removed for testing. if the lager size is used in the final product you could just cut a big hole in it ,add a little glue and name it............. brace.