MLTL driver/port placement, floor reinforcement etc...

Okay, first post, here goes:

I'm planing a MLTL using the Jordan jx92s and have many questions. I would like something a little shorter than GM's 48" design but still want the driver at around ear level when seated (36"). Maybe around 40"H x 10"W x 6"D (internal) with a 2" dia port 2.5- 4" long depending.

After playing around with MJK's worksheets, I've figured out the general gist of driver and port offset values to best squelch unwanted overtones. Now, I noticed that the driver and the port are interchangeable with respect to distance from the ends of the box. For example, driver down 8" from top and port 3" up from bottom on 42" long, 60sq" box graphs the same as driver 3" down, port 8"up. (playing with the box length and port length will keep the driver at ear level and the graphs similar.) So, would not the lower port gain more floor boost, thus disrupting an optimally flat graph? I've noticed that most designs are using ports at or near the floor. Many fire straight at the floor. Is floor boost integral to their design calculations? How does carpet vs. hardwood affect the bass absorption/reflection? It seems to me that the higher port option would reduce this variable and realise in-room closer to mathcad models.

The width of the box is 1/4 the length. Am I going to have any surprise port output from standing waves along the width, not accounted for in the MJK ported box worksheet?

ps. Has anyone tested the T/S parameters for the jx92s? I'm working off Jordan's specs as I have yet to order the drivers (and wouldn't have the means of measuring them even if they were in front of me).

I found one website listing different figures here:

http://www.spectrumaudio.de/

But their Qm and Qe values don't even calculate properly to the Qt they provide. Looks like rounding error, or something.

More questions to come...

Max
 
Now, I noticed that the driver and the port are interchangeable with respect to distance from the ends of the box.

You are correct, the currently available versions of the worksheet do not take into account the driver in the room or the baffle step loss at lower frequencies. So you have to use some judgement in assessing the actual geometry used in your design and how the floor, the front baffle size and shape, and the distance between driver and port impact the idealized response.

The assumptions in my worksheets are consistent with most other freeware modeling tools. I do have newer versions of the worksheets, that include all of these effects, that I am currently using but have not decided how or even if I am going to make them available.
 

Nelson Pass

The one and only
Paid Member
2001-03-29 12:38 am
My own experience gives me a preference for the port at
the floor, firing toward the rear wall. It seems to load well,
you hear fewer of the higher frequency artifacts, and it
maximizes the apparent external distance between the port
and the driver.

As an alternative, I've also had good luck with a QWTP (with no
bottom) firing down into thick carpet. You can alter the port
dimension by raising and lowering the enclosure. :cool:
 
Nelson Pass said:
My own experience gives me a preference for the port at
the floor, firing toward the rear wall. It seems to load well,
you hear fewer of the higher frequency artifacts, and it
maximizes the apparent external distance between the port
and the driver.

With the proper port/driver offsets and stuffing, it seems that there won't be too much high frequency output from the port. Unless there are problems over 1kHz, which is where the MJK worksheets cutoff, so I can't predict this.

Why would you want to maximise the distance between the driver and port? Multiple drivers, according to common practice, should be as close as possible to each other (point source, line source). So, why not driver and port? What about phase issues of a rear firing port?

Max
 
baffle design idea

I m read some threads about baffle step in this and in other audio forum. This is new staff for me..

Anyway, for two weeks ago, I loaded down a "baffle step"-simulation program - "edge".

I get the best result with this baffle shape you can see at the picture. I made the picture in paint so its not perfect...

The intention is to build an ordinary box and later add "the rest" off the baffle. What do you think about this idea?
 

Attachments

  • baffel3.png
    baffel3.png
    2.1 KB · Views: 689
Why not just build GM's MLTL-31 design?
http://www.ejjordan.co.uk/diy/

As long as you keep internal dimensions the same, you can custom-design the cabinet for your desired driver height.

You could even put the thing on stilts, like in this concept drawing:
http://www.stat.ucla.edu/~erickson/jpgs/trapmltl.jpg

Building a wide-and-shallow enclosure, and adding wings, are decent ideas IMO. Since you can never entirely predict the response of the speakers in-room, however, the process is more trial-and-error than science.
 
"Building a wide-and-shallow enclosure, and adding wings, are decent ideas IMO. Since you can never entirely predict the response of the speakers in-room, however, the process is more trial-and-error than science."

Ok... otherwise I got a REALY good result when the shape of the baffle come close to samething that looks like a logarithmic spiral...well I have to find out an "easy modified" material for the "add-on-baffle"
 
A couple of projects in the following page have removable wings, might give you some inspiration:
http://melhuish.org/audio/oswaldsmill.html

I wouldn't get too hung up on "optimizing" the baffle cutoff. Unless you have a truly circular baffle, you're already spreading the cutoff over a range of frequencies. If you spread that cutoff freq over too wide a range, it might be difficult choosing how to implement electronic compensation.

For me, the best solution is wall mounting. Unfortunately, most of us have too much clutter against the walls to make this practical.

Another solution are coupling devices that explicitly control dispersion, i.e. horns, or dipole setups.

Next best thing (IMO) is either a totally wide baffle, or a totally rounded off narrow baffle. With narrow baffle, however, you are doomed to some sort of frequency compensation, and the cutoff freq is in the sensitive hearing band. If you push the cutoff freq down to bass freqs, the brain is less sensitive to whatever junk you put in the signal path. (Human hearing is optimized in the vocal range.)

OK I'm just rambling here.
 
Dumbass "I wouldn't get too hung up on "optimizing" the baffle cutoff. Unless you have a truly circular baffle, you're already spreading the cutoff over a range of frequencies. If you spread that cutoff freq over too wide a range, it might be difficult choosing how to implement electronic compensation."

To me it seems EASIER to implement electronic compensation in this way.....without ANY HUMP in the cut off frequence area...when adding a BSC...whatever...frequence response at the full range driver (Fostex FE167E) is far from perfect..ect..so its mayby no idea to make the baffle "to perfect"...
 
Frequence response with BSC (attached pic)...the bad thing there is the needed value for the inductor...somthing around 10H!!.....I know its possible to make an active circuit instead but than I have to buy a pre- and a power amp :bawling:

Well i will take a look on another baffle shapes..
 

Attachments

  • edge_baffle3_wcomp.png
    edge_baffle3_wcomp.png
    71.5 KB · Views: 562

McRat

Member
2004-07-13 9:26 am
Turku
Gamma:

Commenting nothing on how reasonable it is to design a "perfect" baffle for an imperfect driver, but looking at your baffle design, it really reminds me of a seashell (you know the one some snails carry on their backs :D ).

Knowing seashell shape follows the golden ratio (1:1.61803399), maybe this would be the optimal baffle shape too? Could you simulate this, as I haven't got that program nor do I know how to use it.

Just a thought from an ignorant newbie... ;)

-McRat-
 
McRat: "Knowing seashell shape follows the golden ratio (1:1.61803399), maybe this would be the optimal baffle shape too? Could you simulate this, as I haven't got that program nor do I know how to use it."

The shape "evolved" to a logarithmic spiral during the work in "edge" by mere accident ....width 57 cm hight 1,1 m...I'm not sure if i will use that design... mayby

soongsc: thanks for the link-tip
 
I read all the replies to this original questions, but I am not sure if I understood.

I recently got MLTL-48 cab built, but I found that there was a mistake in internal height. It should be 48", but now the outer height is 48". Since the thickness of the wood is 3/4", the internal height is 1 1/2" shorther, driver is closer to the top 3/4".

What would be the effect of this error? Cross sectional area is correct.

Thanks

Doug