Mark Audio Alpair 10 MLTL Design

[jij3]

In the course of this project I have learned more than I ever wanted about PVP pipe specs. It all started with a local source of "Schedule 40" 6" PVC pipe. The first Tees I bought off of ebay turned out to be for drain pipe, and did not fit. I was however able to jury-rig one up with a scrap piece of Sch 40 pipe to mount one of my Alpair 10.2 EN drivers (from Dave) for break-in. It is on the left of the photo. Shorter than optimal, but actually sounding very good!

I then bought some real Sch 40 Tees (much more expensive...) and came up with the speaker on the right. It is close to Jim Griffin's dimensions (L=46.1", Zd=16", Zp=46", So=Sl=182 cm2). I am planning to cut and inch off the top to give L=45.1, Zd=15, Zp=45. (I have found that Martin King's sheets don't like L=Zp.) There are 25 lbs of lead shot (inside old socks) wrapped completely around a 5" vertical port (1" radius) firing out the bottom. It stands on 1 3/8" spikes. It is still breaking-in, so I can't comment on sound yet.

This is truly an easy project for those of us wood-working challenged. The only real trick is getting the driver mounted securely and as close as possible to the vertical column. The drain pipe Tee was much easier to work with, and may be practically identical in performance. The Sch 40 Tee presented an oblong opening after I cut off the horizontal arm. I used 1" long rings cut square (with a band saw) off the ends of the 6" PVC pipe, fitted and glued to the inside of the Tee opening, and screwed the drivers to their cut ends. I cut open each ring to the proper length (circumference) and placed them in a 300F oven for about 10 min. They come out as soft 1" wide strips of plastic, maleable to any shape. I molded each to the opening let it cool and glued (welded) it inside the opening. When dry it was ready for pilot holes and mounting screws. PVC requires more torque when turning the screws in the pilot holes than does wood, and I stripped one screw head. Fortunately Dave sent me a couple of relacements, thanks Dave.

OK they are cosmetically challenged. I am soliciting advice on beautification. I am considering fitting a whole body sock for each speaker made of grill cloth fabric.

Jack

[Octavia]

Another possibility might be wrapping the top and bottom sections with carpet and allowing the t and end caps to remain as is. Really looking forward to hearing your comments on sound quality once the drivers break in. I'm guessing the rounded shape will help them disappear and image like gangbusters!

[GM]

Originally Posted by jij3

(I have found that Martin King's sheets don't like L=Zp.)

I am soliciting advice on beautification.

???

Slick looking!

Whatever works best overall with the room decor/WAF unless you want them to stand out like a 'sore thumb', in which case I'm partial to brite white, accent shaded with Dayglo orange to create a Creamsicle 'look' .

GM

[Octavia]

One added benefit of using automotive carpet is the backing might be helpful in dampening any structure borne resonances. Might be subtle but then again sometimes all those little things add up. Another thought too is you might want to try some kind of a "footing" under the speaker, such as a wood, marble, granite, that will allow the port opening to fire into the room more without getting muffled by the carpet and its underlying padding. This would be easy enough to try both ways and see what you prefer. Again "kudos" for a nice design!

[jij3]

Thanks for the advice. Once the speaker is broken in, I will try setting it on wood as opposed to carpet.

Interestingly, the short jury-rigged prototype using the sewer Tee is sounding very, very good, now that its driver is fairly broken in.

I think I am going to be very happy with these speakers. Their only limitation is likely to be that sometimes I like to listen at high volume, and these drivers do have limited SPL. So, my next thought is to add a second driver to each speaker. It would be almost trivial to stick one on the top firing up. Given the Tinker-Toy ease in taking the speaker apart, changing tube lengths is easy, but changing the cross-sectional area is not.

I have Martin King's MathCad sheets, and there is a sheet for a MLTL bipole with 2 drivers mounted on the front and back, but nothing for one driver at the end of the TL and the other mounted at 90 degrees further down the line. Does anyone know how to use one of these sheets to model such a speaker? Any thoughts on whether it would be worth the trouble?

Thanks, Jack

[Octavia]

I enjoy the high volumes at times too. I mean lets be honest a concert grand piano at real life volumes is a joy that not many speakers can pull off without some nasty distortion creeping in. I'd love to know the results you get with adding a second driver facing the ceiling. My guess is if you do it right you will add immensely to the scale and realism as long as most of the up firing signal can bounce around for 10 mil sec or more before arriving at your listening position along with the main forward firing drivers. Please post pics of whatever you come up with.

[planet10]

Originally Posted by jij3

I have Martin King's MathCad sheets, and there is a sheet for a MLTL bipole with 2 drivers mounted on the front and back, but nothing for one driver at the end of the TL and the other mounted at 90 degrees further down the line. Does anyone know how to use one of these sheets to model such a speaker? Any thoughts on whether it would be worth the trouble?

You have some options:
1/ model an ML-TL with 2 end loaded drivers (or with 1 and then double the cross-section). Similar to Castle microTower.
2/ move the driver on the back to the side
3/ model an 2 driver ML-TL, then top mount one driver, and then place the 2nd such that the average position is at the specified Zd. The further the drivers apart the more the actual build will deviate from the sim. As well front driver may end up kinda low.


dave

[jij3]

many thanks, Dave

For option 1, using the model with one speaker and doubled Sd, if I wire the speakers in parallel, should not the Re be halved? Should the Vad be doubled as well as Sd? If I do this, the sim looks pretty good, with the double speaker mounted at the top or at 15" down from the top. Do you think I could actually mount another driver to the top of the existing column as is and expect good results?

For options 2 and 3,are you referring to the TL_ML_Bipole_Corner_7_03_09 file?

Jack

[planet10]

I don't know any of the MJK files (i'm on a Mac), i farm out all my TL modeling to Scott.

Just plopping a driver on the top moves the apparent Zd, so ripple cancellation will change.

dave

[jij3]

I went over to MJK's Yahoo Group, and with Martin's guidance, I have modeled the speaker using a double-sized driver positioned halfway between the forward-firing driver and the top of the TL. Yes, there is more ripple because of the higher position of the driver on the column. The bass also begins to fall off 10 hz higher because of the relatively small cross-section. The only way to prevent this is by doubling the cross-section of the TL and the port.

I may try it anyway with the two drivers I have, since anything I do with the PVC is easily reversible, and I do use a subwoofer for the lowest bass.

Jack