Need T-Line help for Full Range T-Line

[wanton]

Hi,


I want to build a TL based on MarkAudio Alpair 10.2 (Markaudio) drivers.

The TS parameters are the following

Alpair 10.2

Revc= 7.200 Ohm
Fo= 35.386 Hz
Sd= 88.250 cm2
Vas= 28.119 Ltr
Cms= 2.543m M/N
Mmd= 7.480 g
Mms= 7.956 g
BL= 5.606 T·M
Qms= 2.530
Qes= 0.41
Qts= 0.35
Levc= 93.626u H
No= 0.297 %
SPLo= 87 dB
X max= 8.5-mm (1 way)
Power = 35 watts nom. 70 Max
Coil = 25-mm alu body with rectangle copper winding.

As You can see the Vas is quite high and the stated frequency range is very nice.

I used the excel spreadsheet made by Mr Keith Webb on Quarter Wavelength Loudspeaker Design, that simulate Mr Kings tables.
I have calculated a TL line length of 2,25m, which would have a ratio of the driver end to the port opening (SL/S0) of 0,8.
According to Vance Dickason this would be a 1,25 Sd to 1Sd tapered line.
The S0 (area right behind the driver )of the line according to Mr Kings tables turns out to be quite bigger than the Sd shown in the TS parameters.
That would be 300 cm2 compared to the 88cm2 of the Sd

The same goes for the SL (area of the port opening).
That would be 244cm2 compared to the 88cm2 of the Sd

That would mean that the line would be a tube ~ 3,6: 3Sd, this has me confounded

This is totally different than what is written in Vance Dickasons Loudspeaker cookbook and what You have on Your site.
Where it's stated that the speaker should be on the end of a tapered line but 1,25 - 2,5 Sd:1Sd.

This put me in a bit of a fix because I don't know whether to do it by Mr Kings Tables or per the suggestions in the Vance Dickason book.

Can anyone explain where this difference comes from ?

Also is it worthwhile to adjust the resistance of the speaker so it becomes exactly 8ohms instead of the 7,2 stated in the parameters ?


Thx for Your Help

[chrisb]

Originally Posted by wanton

Hi,


I want to build a TL based on MarkAudio Alpair 10.2 (Markaudio) drivers.

The TS parameters are the following

Alpair 10.2

Revc= 7.200 Ohm
Fo= 35.386 Hz
Sd= 88.250 cm2
Vas= 28.119 Ltr
Cms= 2.543m M/N
Mmd= 7.480 g
Mms= 7.956 g
BL= 5.606 T·M
Qms= 2.530
Qes= 0.41
Qts= 0.35
Levc= 93.626u H
No= 0.297 %
SPLo= 87 dB
X max= 8.5-mm (1 way)
Power = 35 watts nom. 70 Max
Coil = 25-mm alu body with rectangle copper winding.

As You can see the Vas is quite high and the stated frequency range is very nice.

I used the excel spreadsheet made by Mr Keith Webb on Quarter Wavelength Loudspeaker Design, that simulate Mr Kings tables.
I have calculated a TL line length of 2,25m, which would have a ratio of the driver end to the port opening (SL/S0) of 0,8.
According to Vance Dickason this would be a 1,25 Sd to 1Sd tapered line.
The S0 (area right behind the driver )of the line according to Mr Kings tables turns out to be quite bigger than the Sd shown in the TS parameters.
That would be 300 cm2 compared to the 88cm2 of the Sd

The same goes for the SL (area of the port opening).
That would be 244cm2 compared to the 88cm2 of the Sd

That would mean that the line would be a tube ~ 3,6: 3Sd, this has me confounded

This is totally different than what is written in Vance Dickasons Loudspeaker cookbook and what You have on Your site.
Where it's stated that the speaker should be on the end of a tapered line but 1,25 - 2,5 Sd:1Sd.

This put me in a bit of a fix because I don't know whether to do it by Mr Kings Tables or per the suggestions in the Vance Dickason book.

Can anyone explain where this difference comes from ?

Also is it worthwhile to adjust the resistance of the speaker so it becomes exactly 8ohms instead of the 7,2 stated in the parameters ?


Thx for Your Help

I'd go with Martin's work, and don't worry about the nominal 10% difference in resistance - once you apply AC frequencies that number becomes a reactive impedance that can change dramatically in terms of amplitude and phase angle. A well designed and damped enclosure can tame that somewhat, but for any moving coil motor that measurement will never be ruler flat.

You're aware that there's an excellent MLTL (term coined I think by Martin?) design for the A10.2 by Bob Brines? I've built several enclosures for this driver, and these certainly deliver.

M10-A10

[Scottmoose]

The pensil and Bob's box are both variations on transmission lines / QW boxes so there's a couple of options there. Regarding the above:

-Forget Dickason's data unless it's the material in the latest Loudspeaker Design Cookbook, which is derived from Augspurger's alignment tables (it sounds like you've got an earlier version). If this is the case, go with Martin's tables, which are the more complete.

-The 1.25:1 taper ratio mentioned is an old rule of thumb; goodness knows where it came from, but it's essentially twaddle. The piston area of the driver has no direct relationship to the cross section of the line, although it's sometimes used as a convenient way of expressing it.

-A TL is no different from other vented boxes in the sense that it requires volume. Length & taper set the frequency the line is tuned to; if it lacks sufficient volume for the tuning frequency and driver in question, you'll simply choke the output & raise F3.

-A very shallow taper isn't worth the effort. 3:1 is really the practical minimum for a reverse taper line, assuming you want to get any benefit from it, and preferably greater.

-Use the driver's impedance, not the nominal 8ohm.

[planet10]

Originally Posted by wanton

According to Vance Dickason this would be a 1,25 Sd to 1Sd tapered line.

To further emphasis what Scott has already said, the old=school design scheme this rule-of-thumb came from was obseletted in Fall 1999 when both Augspurger & King contemporaneously came out with their independent but corroborating transmission line models, which both matched physical results, showed how poor the old design recipe was, and opened up huge new swaths of TL-space. Now one can, if they know what they are doing, sit down with the modeler and be pretty sure that the TL they design is a pretty good representation of what it will be when built.

You should just cut that section of Dickason out of your book and burn it. It is here for historical purposes only.

Note that the MJK tables only generate 1 of many possibilities for a TL -- a small easily tabled, subset of TL-space.

dave

[wanton]

Cheerio....
Thx for the good info ...
I will post pics of the ongoing construction.

The Alpairs are already on the way :O)

[wanton]

Posting a pic of my first idea ....

Comments are well appreciated

Construction material would be Birch plywood 20 mm on the sides.
3pcs of 8 mm Elastic ply would make the curvy parts.

According to Kings tables the parameters of the line would be the following

Line length 2,254m for a 0.8 taper ratio.
Start cross section of the line right behind the driver 304 cm2,
End cross section of the line at port opening 244cm2

[Scottmoose]

My initial query is what special advantages are you hoping to achieve with a 1:0.8 taper ratio? It's far too shallow to significantly lower Fp for a given length, so you're not going to end up with a much shorter pipe than if it were untapered, and since it will be damped (IIRC Martin's tables assume a uniform 0.5lbs ft^3 stuffing density of dacron hollow-fibre material), you're not really going to get any significant reduction in unwanted standing waves either.

[wanton]

Hi, Thx for Your reply

The idea to make Tls came during my studies, I wanted a speaker that played good ant looked different. I didn't wnat to make a box, so I started messing around with different geometric ideas.

My original assumptions were based on the 4th ed. of Dickasons cookbook, which turned out to be quite outdated.
That's were the 1,25:1 ratio came out of, I put my parameters into the spreadsheets @ quarter-wave and then tried to model a speaker to my liking.
Of course these are ideas, therefore can be changed
Do You advise to make the ratio more like 4:1 ?

For Fp I understand the resonant frequency of the enclosure ?

Based also on my old assumptions I was ready to experiment with different stuffing materials and densities, but now am all ears to suggestions as to optimize ...

Cheers

[planet10]

You should be looking at a minimum of a 4:1 taper, and you should also be taking advantage of a driver off-set. End-loading a TL is a compromise.

dave

[GM]

Originally Posted by wanton

Do You advise to make the ratio more like 4:1 ?

I advise more like 10-20:1.

GM