Compact TL’s for SS 18w8545k

[hihopes]

I would like to rehouse my 18w8545k’s in COMPACT transmission-line enclosures. They have to sit on top of my 60 litre bass boxes. I have spent weeks scratching around for TL info and now, using whatever information I have been able to find, I have designed a box as follows:-

SO = 362.5 cm2 (SD x 2.5)
SL = 72.5 cm2 (8:1 taper)
Length along centre of line = approx 1.2m
Driver centre @ 225mm from SO
Line has 3 folds
Width of line = 250mm
Driver specs:- Re = 5.5; Le = 0.4; Fs = 28Hz; Qts = 0.28; Bl = 8.2tm; Sd = 145cm2; VAS = 48L; X max = 10mm. (from datasheet) Drivers are EnAbl’d.

Using 22mm material, this gives me a box with outside height of 400mm and depth of 350mm - about as big as WAF will allow (with a bit of cajoling).

Drawings done on graph paper by hand, so cannot post at present.

Can some kind soul please sim this and see how it might perform? Any suggestions for improvements would be most welcome.

[pkitt]

First there is something amiss with your statements. You say the line has an 8:1 taper, but the SO/SL you listed is only 5:1; however, since SO is 2.5 x Sd as you stated, I changed SL to 45.3 sq.cm. to make the taper be 8:1. There is one piece of info you didn't state and that is the stuffing density and length. Anyway, with the change to SL I made, I modeled your tapered line using one of Martin King's MathCad-based TL worksheets, using a stuffing density of 1 lb/cu.ft. (a total of ~10 ounces of stuffing in the first half of the line's total length). This is not an absolutely accurate model in that it does not take into account the folds in the line, specifically how the cross-sectional areas around those folds should be modeled. If you want, based on how this looks at this point, I can model it more accurately with a different worksheet (you could send me a cabinet sketch to work from). I've attached the system response graph (red line) and overall it appears pretty good. There are the normal ripples above 500 Hz which could be reduced by increasing the stuffing density and/or length but at the expense of F3. OTOH, if you made the line shorter and increased the taper ratio, keeping the same volume within the line, that would also reduce the magnitude of the ripples without making F3 higher.
Paul

Edit: I modeled without any resistance added in series with the driver, but there will be some from crossover inductor(s) and should be included for more accuracy. Also, on the graph I attached please ignore the "20 Hz" label below the bottom left corner. The frequency does start at 10 Hz (it's a quirk in copying from Martin's King's software that I can prevent but often forget to do so).

Originally Posted by hihopes

I would like to rehouse my 18w8545k’s in COMPACT transmission-line enclosures. They have to sit on top of my 60 litre bass boxes. I have spent weeks scratching around for TL info and now, using whatever information I have been able to find, I have designed a box as follows:-

SO = 362.5 cm2 (SD x 2.5)
SL = 72.5 cm2 (8:1 taper)
Length along centre of line = approx 1.2m
Driver centre @ 225mm from SO
Line has 3 folds
Width of line = 250mm
Driver specs:- Re = 5.5; Le = 0.4; Fs = 28Hz; Qts = 0.28; Bl = 8.2tm; Sd = 145cm2; VAS = 48L; X max = 10mm. (from datasheet) Drivers are EnAbl’d.

Using 22mm material, this gives me a box with outside height of 400mm and depth of 350mm - about as big as WAF will allow (with a bit of cajoling).

Drawings done on graph paper by hand, so cannot post at present.

Can some kind soul please sim this and see how it might perform? Any suggestions for improvements would be most welcome.

[bjorno]

Originally Posted by hihopes

Can some kind soul please sim this and see how it might perform? Any suggestions for improvements would be most welcome.

Hi,hihopes,

See the picture.

b

[hihopes]

Gosh, you guys are quick - and very kind too! Thanks a ton. paul, sorry about the confusion re taper ratio. You guessed correctly to make Sl @ 43. I had a spreadsheet on which i was working out various dimensions and I read off from the 5:1 section while writing the post. The figure I got was 43.3. I did not mention stuffing because I intended to try various stuffing techniques. I see that Dave (Planet 10) has a method of heavily stuffing the port to create a type of variovent effect.
I have spent some time knocking together a drawing of the box and I see that I can't get the driver as far away from SO as I worked out in hand drawing. As the drawing stands, I get a distance of 195mm to driver centre. I could probably get a few mm more if needed, or if it makes a major difference, make the box slightly higher (ouch!).

Bjorno, your suggestions seem to be well thought out. However, I am more interested in getting a smoother FR than in extending the bass performance. The main reason I want to build TL's at all is to improve the clarity of the mids and the tightness of the bass. Your suggestion of a closed TL is very interesting. I never even thought of that -and I see the volume you mention is very much smaller @ 15L (My current boxes are 14litres vented enclosures). That would certainly make them very compact. Big question:- Would a closed tapered TL have the same kind of sound as an open tapered TL? I assume that I would have to make SO quite a bit smaller to get such a drop in box volume?

What happens if I reduce SO in my current design?

I have an image saved in jpg format. How do I attach it?

[hihopes]

I think I got the pic here. Pls excuse the rough and ready appearance. It is to scale, though.

[pkitt]

You really don't have to be concerned if you can't get the woofer's center exactly where you wanted it, and if I still had the modeling worksheet up I'd re-model it with the new woofer location. I'm not sure if you want me to do anything more for you at this time?
Paul

Originally Posted by hihopes

I think I got the pic here. Pls excuse the rough and ready appearance. It is to scale, though.

[hihopes]

Paul, thanks very much for your help. I have no idea what the most optimal woofer position might be. I just thought further down would probably be better.

I would think that stuffing about 2/3 of the length (1st 0.8 meteres) would kill frequency output above about 400hz and get rid of most of the ripple. Does it sim that way?

I don't mind pushing F3 higher - even up to 100hz if necessary, as I have bass speakers. I can adjust their x-over to suit. My main concern with these is getting the best possible midrange. If changing the design drastically gives much less ripple, then I will do that, provided they retain the best characteristics of the TL "flavour."

Thanks for your help so far. I don't want to put you and Bjorno to any unnecessary trouble. I guess there are a gazillion ways to do this and I can't expect you guys to just keep on simming whatever changes I think of. I just want to gather as much info as possible in order to come up with the closest to optimal design that I can within the parameters mentioned. (I hope I have clarified the parameters sufficiently clearly).

Any more suggestions for me to consider and research will be much appreciated.

[bjorno]

Originally Posted by hihopes

Big question:- Would a closed tapered TL have the same kind of sound as an open tapered TL? I assume that I would have to make SO quite a bit smaller to get such a drop in box volume?

What happens if I reduce SO in my current design?

No, a closed TL has often better sounding upper bass (80+ (Hz)) than an open TL but would sound like an full BW tapered open TL that is XO: ed (max 2nd order filter) at least an octave higher than the low end at mid-bass ~ upper-bass frequencies, about 70 + (Hz), provided: the TL has a smooth roll off like the low end of the improved TL I showed in my picture.

A peaked roll off like yours has phase issues at about 50 Hz and If not stuffed to be aperiodic, would interact with a LP filter at the upper midband frequencies, counteracting a smooth FR summing with the (sub) woofers.

IME for your application, closed TL’s (Your current design with a closed terminus will do the job, you don’t necessarily make your enclosure like a cone like I did) or closed boxes would behave far more smoother at the upper FR (denser stuffing’s near the closed terminus) end than if you stay with your current open TL and would be easier to XO complementing (sub) woofers as you probably don’t need to add any HP filters.

b

[pkitt]

You'll note in the modeling I already did for your design that I placed stuffing in just the first half of the line. Now if you really are willing to give up more of the bass response, stuffing more length of the line will reduce those higher frequency ripples more, while also increasing F3. Later today I'll model your TL some more and see what I can come up with.
Paul

Originally Posted by hihopes

Paul, thanks very much for your help. I have no idea what the most optimal woofer position might be. I just thought further down would probably be better.

I would think that stuffing about 2/3 of the length (1st 0.8 meteres) would kill frequency output above about 400hz and get rid of most of the ripple. Does it sim that way?

I don't mind pushing F3 higher - even up to 100hz if necessary, as I have bass speakers. I can adjust their x-over to suit. My main concern with these is getting the best possible midrange. If changing the design drastically gives much less ripple, then I will do that, provided they retain the best characteristics of the TL "flavour."

Thanks for your help so far. I don't want to put you and Bjorno to any unnecessary trouble. I guess there are a gazillion ways to do this and I can't expect you guys to just keep on simming whatever changes I think of. I just want to gather as much info as possible in order to come up with the closest to optimal design that I can within the parameters mentioned. (I hope I have clarified the parameters sufficiently clearly).

Any more suggestions for me to consider and research will be much appreciated.

[pkitt]

Just to establish a reference point, I modeled the same setup as I did previously but with the driver's center located 195 mm from the beginning of the line with a stuffing density of 1 lb/cu.ft. and the first half of the line stuffed (~9.5 ounces total). The first graph below shows the resulting system response which is very similar to what I posted earlier. Next I kept the same stuffing density but stuffed the first 2/3 of the line and that result is shown in the second graph (~11.5 ounces total of stuffing). The third graph represents the system response with the whole line stuffed at 1 lb/cu.ft. (~13 ounces total but no stuffing in the terminus itself). As you can see the magnitudes of the ripples above 400 Hz are decreased as the stuffing length is increased, and the shape of the lower end of the curve is changed along with an increase in F3.
Paul