Yet Another Kappalite Tapped Horn

KTOKTO

Member
2014-09-06 11:36 pm
I just picked up an Eminence 3015LF and would like to use it as a learning experience building a SS15 style TH using Brian's spreadsheet for the layout. My woodworking skills are up to par but I am going to need some assistance in making sure the design is sound before I proceed.

I am planning on using braced 1/2" ply like the original SS15. I don't have any minimum volume or single sheet requirements. If the build is successful, my plan is to build 4 to be used in a small live act (guitar, bass, drums) crossed around 100-120Hz.

Here is what I have thrown together thus far using the published T/S for the 3015LF.

[IMGDEAD]http://i66.tinypic.com/108bdb7.png[/IMGDEAD]

[IMGDEAD]http://i68.tinypic.com/os4a4p.png[/IMGDEAD]

[IMGDEAD]http://i67.tinypic.com/2l8wydd.png[/IMGDEAD]

I measured the driver in DATS after break in and threw my measured values in to HR. Everything was close to the published specs but the bump at 55 was less pronounced and the sim was up around 1 db at 40. Sorry, I forgot to save the measurements so I can't post the sim with my measured values at this time.

One concern I have right away is the compression ratio I should be shooting for due to the light cone. I didn't search too long but I couldn't find the value used in the original SS15 with my initial search.

Like I said earlier, I am a rookie so any suggestions or criticism on what I need to change here are greatly appreciated. Thanks.
 
Last edited:
Like I said earlier, I am a rookie so any suggestions or criticism on what I need to change here are greatly appreciated. Thanks.

The physical layout looks Ok. See if you can reduce the "error" associated with the driver's baffle by changing its length a little.

The sim'd results however with that 2dB bump at the lower end suggest that this may not be the best alignment for that driver however. I suggest aiming for a smoother passband and cutoff.
 

KTOKTO

Member
2014-09-06 11:36 pm
The sim'd results however with that 2dB bump at the lower end suggest that this may not be the best alignment for that driver however. I suggest aiming for a smoother passband and cutoff.

I did some messing around and it looks like the 3015LF wants a much larger S2 than I have been able to get with the spreadsheet. I think if I could move the point where internal panel 2 attaches to panel 1 I would be in business. (Basically, shift internal panels 2-4 away from the baffle a bit.)

I'm not sure if that would be a huge programming undertaking to add a field to make panel 2 move left and right like panel 5 moves up and down, or if that makes sense. Just a thought.
 
Last edited:
I did some messing around and it looks like the 3015LF wants a much larger S2 than I have been able to get with the spreadsheet. I think if I could move the point where internal panel 2 attaches to panel 1 I would be in business. (Basically, shift internal panels 2-4 away from the baffle a bit.)

Looks like you're referring to "Panel G" in the spreadsheet. You can shift it away from the baffle ("Panel B") by increasing the value for "Panel G Offset". This will also increase S1 of course.

You can also increase S2 by reducing "Expansion 1" a bit. S1 will remain unchanged in that case.
 

KTOKTO

Member
2014-09-06 11:36 pm
Looks like you're referring to "Panel G" in the spreadsheet. You can shift it away from the baffle ("Panel B") by increasing the value for "Panel G Offset". This will also increase S1 of course.

You can also increase S2 by reducing "Expansion 1" a bit. S1 will remain unchanged in that case.

Geeez, how did I miss that, It was staring right at me :) Thanks again.
 

KTOKTO

Member
2014-09-06 11:36 pm
OK, things are starting to click. Here is my second attempt.

QzJ87zw.png


435bTqU.png


I would like it to have a little more power at 40 and sacrifice the sensitivity up top. I am messing around in hornresp and I like the look of things once I get the passband around 100db by lengthening the horn path.

I am guessing I can get this done by making the front panel E longer, making the box taller and less deep, or a combination of these. Is this a correct assumption on where to start?
 
Last edited:

KTOKTO

Member
2014-09-06 11:36 pm
Set "Expansion 1" to be equal to "Expansion 2", unless you really want to pinch the horn at S2 like that. Then enter a figure for the "cone volume adjustment". This will pinch the horn at that point to compensate for the volume of air in front of the cone.

Thanks for the advice, how are things looking here? Just wondering if I am on the right path? Is the large spike where panel F meets panel G1 a concern?

n2lgu0.png
 
Thanks for the advice, how are things looking here? Just wondering if I am on the right path? Is the large spike where panel F meets panel G1 a concern?

[IMGDEAD]http://i63.tinypic.com/n2lgu0.png[/IMGDEAD]

The large spike is fine. If anything, the bit of extra volume there will lower the horn's primary resonance frequency a bit. Some choose to smooth off that area with an additional cross panel.
 

KTOKTO

Member
2014-09-06 11:36 pm
Thanks again for the help Brian. I think I am getting close to having something with a 100db passband -3db @ 40. I made the box an inch taller and am working on lengthening the path. It is really interesting learning what parameters affect the response curve in the sim.

[IMGDEAD]http://i67.tinypic.com/ace4j8.png[/IMGDEAD]

[IMGDEAD]http://i65.tinypic.com/9tg65w.png[/IMGDEAD]

I have to work on fixing the error on panel I but I think I am on the right track. I just need to see if I can squeeze another 1/2 db @ 40 out of it at this size. Any suggestions are appreciated.

Thanks again.
 
From a few quick HornResp sims, it looks like you might need to aim for a cab that's about 290L litres net to get a somewhat decent response with this driver in a 40 Hz TH. (see attached for example). The issue here is this driver's high Qts, a result of its relatively low BL. You can probably squeeze it down to 250L, but you'll be making some compromises along the way .

(BTW - thanks once again to David for the ability to now illustrate the prime resonance frequency of the horn right in the Loudspeaker Wizard's Acoustical Power window - that's made a significantly positive difference in using the wizard!).
 

Attachments

  • 2016-02-05 (1).png
    2016-02-05 (1).png
    23.9 KB · Views: 168

KTOKTO

Member
2014-09-06 11:36 pm
Here is 278L. This ends up being 1" wider and 3" taller than the original ss15 I believe. The crossover point is pretty low though. I normally cross my tops at 120 but they can go to 100.

[IMGDEAD]http://i64.tinypic.com/2hicgeq.png[/IMGDEAD]

[IMGDEAD]http://i65.tinypic.com/4kwhef.png[/IMGDEAD]
 
Last edited:

KTOKTO

Member
2014-09-06 11:36 pm
OK, I ran through a bunch of variations this weekend and I think I am going to settle on a net volume of 250L. This is the last variation I tried but I keep running in to the same dip in the upper band.

In hornresp, if I take the parameters below and decrease L23 and increase L34 and L45 a bit things start to look really good. The lower dip up and the upper dip down smooth out quite nicely when I do this. Unfortunately, I am having a hard time making this happen in the spreadsheet.

Do you have any suggestions on what I need to change on my layout? I must be missing something because I have tried making the box deeper, adjusting the size of the front panel but I keep getting in a loop and am ending up with similar results no matter what I try.

[IMGDEAD]http://i68.tinypic.com/2qi93mf.png[/IMGDEAD]

[IMGDEAD]http://i67.tinypic.com/2ltn66s.png[/IMGDEAD]
 

KTOKTO

Member
2014-09-06 11:36 pm
KTO, looks great! Mind posting the HR info so I can fool? From memory and very basic HR skills I remember getting similar results when I deliberately tried to get the smoother or shallower knee (which I like a lot > 90db @ 30hz).

The hornresp inputs for the box are listed on the spreadsheet screen shot above. The ts params I used are from eminences website.
 

KTOKTO

Member
2014-09-06 11:36 pm
Mind posting the HR info so I can fool?

OK, I got back to my computer. Here are the inputs for post #17 above. If I use the loudspeaker wizard and take a little length out of L23 and add it back in at L34 and L45 things look very good. The low end looks very similar to above but the dip in the upper pass band is pushed up quite a bit. I just cant seem to figure out what I need to tweak though to move that path length from L23 to L34 and L45 like Brian recommended.

wwkwvd.png
 
Last edited: