Mark Audio Alpair 10.3/A10p MLTL

I have just completed a MLTL design for the Mark Audio Alpair 10.3 (metal cone) and 10P (paper cone) driver. This design with the one change mentioned below will work with either cone variant. My model (internal dimensions of 40", width of 6", depth of 7.5"). The driver is centered an internal distance of 8" from the top. The port is centered and located an internal distance of 36" from the top. The port has a 1.25" radius (2.5" diameter) and is 6" long for the 10.3 and 5" for the 10P. A stuffing density of 0.75 lb-ft3 is placed in the top half of the box.

Thus the overall box size (with 0.75" thick material) is 41.5"H x 7.5"W x 9.0"D.

I have attached a pdf file for the 10.3 variant that shows the performance of the Alpair 10.3. These are simulations from the Martin J. King spread sheets. Several plots show the performance of an infinite baffle variant for comparison to demonstrate how the MLTL tuning performs.

The first plot shows the system response for a 1 watt input. The F3 (3 dB down point) on the low end of the band is in the 33-34 Hz area which is excellent. Above F3 the response is flat ( thanks to the 0.75 lb-cubic feet stuffing) and smooth.

The next plot separates the responses from the woofer and the port. Notice that the output from the port dips in roughly the 37 Hz area which where the box is tuned.

The impedance of the ported box shows the classic (and well centered) twin peaks which is typical of an optimally tuned box.

The final two plots were simulated for a 15 watts input to the MLTL. I'm looking for the performance with a higher level input and just where the displacement exceeds the spec value of 7.5 mm Xmax peak. The displacement plot in this case is the RMS level which is 0.707 times the peak value or 5.25 mm. Notice how the system response is about 100 dB SPL (not a sub woofer but relatively loud). At this point the RMS displacement at the very low end of the band (defined by the 33-34 Hz point) starts to exceed 5.25 mm around 30 Hz. Power levels greater than 15 watts below 30 Hz would cause the driver to be overdriven and eventually could damage the driver except for the arrestor that mechanically protects the Alpair 10.3 or 10P.

This design is not for commercial use.
 

Attachments

  • Alpair10.3Responses.pdf
    138.9 KB · Views: 3,420
Last edited:
Very nice. Did you already build it?
I did simulation in Leonard Audio and it looks also very nice. Here is picture of simulation.

What distance is ment as far field on your simulation?
 

Attachments

  • Alpari 10M - MLTL.png
    Alpari 10M - MLTL.png
    168.4 KB · Views: 6,881
I haven't built this MLTL but this is the same modeling that I have done many times before with exceptional results for a variety of speakers. You can search this forum for examples.

This MLTL is a straight box with no taper. It should be easy to construct. The MJK software includes the effect of stuffing so results are cleaner than the Leonard Audio results.

Far field starts at a distance which is roughly beyond the diameter of the aperture. We are talking a 4 inch aperture in this case. Typical far field measurements at a meter or so from the speaker will minimize reflections from nearby objects and boundaries.
 
Founder of XSA-Labs
Joined 2012
Paid Member
The Accidental MLTL technique may surprise you with how close it can come to a decent design using just a bass reflex model and approximate rules of thumb. More info here:
http://www.diyaudio.com/forums/full-range/231951-accidental-mltl-technique.html

Basically use a BR package like WinISD to calculate a bass reflex box volume and vent dimensions. Adjust the volume and vent dia and length to your liking to get the tuning you want. Now use the volume and set the length of the MLTL based on this approximate rule: 30in to 35in for 3 or 4in drivers, 35in to 45in for 5in to 6.5in drivers, and 45in to 60in for 8in drivers. Set the baffle width and use the volume to calculate the depth. Set the driver at 0.20x to 0.33x from closed end. Put the vent at about 0.8x to 0.9x from closed, add stuffing from closed end to just a few inches below the driver. That's it - the long aspect ratio helps to support quarter wave resonances which change how the driver is loaded and provides a nice smooth response with deeper bass extension. This method doesn't work well with low Qts drivers but moderate to high Qts drivers it seems to have a good success. If anything, you have a starting point for optimization with a model.

I have built several this way and am always surprised how good they sound. I can model them now in AkAbak easily enough - but still find it useful as a starting point.
 
Kazoo (or better John),

You asked: "Will this also work for the Alpair10 paper version?"

Excellent question. To address my answer I took my simulation model for the Alpair 10.3 and replaced its T/S specs with the Alpair 10P Gen. 1 ( the paper cone version) T/S specs. The box dimensions( given in my first post in this thread) were kept the same so I'm modelling the paper version in the same box as I derived for the 10.3 cone. The overall system response for both cone versions were compared and are essentially the same. Hence the system response plots virtually overlap. You can use either driver versions in this box and achieve excellent results.

If you really want to perceive a change between the metal and paper cone MLTL variants, I would only suggest a tiny change for the paper cone driver. Change the port tube length from 6" to 5" long. There is a small (about 1 dB downward) slope from 1000 Hz toward the 10 Hz division on the frequency axis (remember that F3 is at 33-34 Hz) of the paper cone system frequency response. You can flatten this slope by reducing the tube length. In the real world of an actual MLTL box you likely could not measure nor hear a change of this type over nearly two frequency response decades.

Bottom line is that this enclosure works for either the Alpair 10.3 or 10P drivers.

Jim
 
Last edited:
Yes, the outer frame diameter is large with these drivers. The size of the frame creates both external fit and internal fit issues to consider. Now I should have pointed out that you can trade width of the cabinet for depth of the enclosure if you are concerned about internal reflections. I modeled internal dimensions of 6" width x 7.5" depth or a cross-sectional area (CSA) of 45 sq. in. You could just as well select a 7.5" width and a 6" depth and have the same internal CSA. In either case the internal stuffing fills the cavity from the top of the box down to about 3" below the driver. So moving the walls (as this modification suggests) and the immersion in the stuffing material should alleviate internal reflections.

I would be also concerned about the frame thickness if you intend to flush mount the driver. The frame is 13/32" (10.5 mm) thick. With a 0.75" thick baffle you would have a 11/32" thick flange wall under the driver for the flush mount. While 6 screws will help to hold/stablize the driver in place, additional efforts may be necessitated to attain a firm fit. One solution is to radially glue 0.75" long pieces of 0.5" square hardwood dowel to the underside of the baffle under the mounting holes. Hence, the screws bit into these pieces of dowel when you tighten them.

Jim
 
Last edited:
On the 6" port length in an enclosure 6" deep: The port mounts to a cabinet wall which is 0.75" thick. Thus a 6" port extends 5.25" into the enclosure. That leaves 0.75" clearance. Perhaps not ideal but I've used this much clearance in a few cases. If you worry about such things, you can shorten the port tube to a length of 5" with minimal performance impact (you might raise the F3 point a couple of Hz or so).
 
I have a design for the A10m that also works for the A10P. Mine just happens to be 40" tall, but has a different alignment. Jim's seems to be close to QB3 while mine is somewhat overdamped. Different approaches.

In my designs, I find that that the optimum port placement is 35" from the top. An inch or two amongst friends is no big deal. With a port length that is nearly the depth of the cabinet, vent the port out the bottom. This puts the top of the port at nearly the correct position in the cabinet and two problems get solved at once.

Bob
 
Using the T/S parameters on the Madisound site, the TL alignment spreadsheet suggests a larger cross section area(67in^2) cabinet and larger volume(~1.5cuft) for the paper cone Alpair 10P Gen.2, than the metal cone Alpair 10 Gen.3 (45in^2 and ~1cuft).

Modeling room gain might change optimal port tuning goals, and some flared ports screw in from the front baffle to allow adjustment.

Might be worth some study for the paper cone Alpair10P before you cut wood.
 

Attachments

  • Alpair!0x  MLTL.jpg
    Alpair!0x MLTL.jpg
    126.4 KB · Views: 2,838
Plots for 10P MLTLVariant and Comments

I've done some bookkeeping in the first post in this thread (retitled to include the 10P design). As I could not change the original file download in that post, I have attached two performance files to this posting. One is the same file (but relabeled) as in the first post for the 10.3 MLTL performance while the second file is for the 10P MLTL.

I appreciate the inputs by Bob and several others. As we know a MLTL design is not unique as the modeler can vary different parameters--tube length, cross-sectional area, driver location, port location, port diameter, port length, stuffing amount, etc. You can change these factors and achieve the results that you prefer while others may prefer a slightly different design for their needs. You can also model a MLTL in your specific room with MJK's software so you can adapt it to your situation. Also as pointed out by Bob Brines you can alter the system response to suit your taste for damping.

Jim
 

Attachments

  • Alpair10.3Responses(1).pdf
    139.5 KB · Views: 468
  • Alpair10PResponses.pdf
    139.9 KB · Views: 451
brookhart- one of the very practical advantages of the ML-TL topology is that it is "scalable" for a wide range of driver models*, but of course, for optimal results, the math needs to be reworked for each

Jim and other folks have designs for the 7.3. including Scott Lindgren's Pensil which works very well

http://frugal-phile.com/boxlib/pensils/Pensil7s3-map.pdf

* edit: and as Jim points out, the designer can target a number of goals in optimizing his calculated compromises - of which there are always some
 

Attachments

  • 7.3 Penils MarKenT 003.jpg
    7.3 Penils MarKenT 003.jpg
    278.4 KB · Views: 2,855
Last edited:
leonard audio help

Very nice. Did you already build it?
I did simulation in Leonard Audio and it looks also very nice. Here is picture of simulation.


hi can you send me your file for this as i want to create a similar structure using a different size driver and amplifier settings. I am relatively new and from the image you put up it seems like something i would benefit from. If you have the file for it that would be great help thanks.

Ryan.