Ultraslim tapped horn

This is not a high spl subwoofer, it's purpose is to give my alpair7 tl's support from 30-60hz.

Using two 4 inch drivers (W4-792sj) seems a bit weak in subwoofer terms and when simulating in hornresp they x-max-out at 7-8W in this setup.
The reason I picked 4-inch woofers is to keep the total hight of the box below 14 cm which happens to be the space available under my sofa :)

Well plugged in and running, crossed at 60 hz 24 oct/db, I'm more than surprised over the efficiency of a tapped horn. I would guess that the cone displacement is just over 1mm at max level of the alpairs (subwoofer level tuned at my flavor of course) and in my way of listening to music thats l o u d.

Just wanted to share this since I'm so happy with it!

inpu_para.jpg


01.jpg


08.jpg


03.jpg


05.jpg


07.jpg


09.jpg


11.jpg




Regards [email protected]
 

David_Web

Member
2008-09-27 1:17 pm
Looks neat!
I am interested in that Hypex amp due to it being relatively cheap.
How do you think it fair in a TH?
Did you defeat that boost at 25Hz or did you do any other modifications to make it suitable for a tapped horn?
IIRC the built in LP did get a lot shallower when crossed below it's max setting.
Can't remember exactly as my main PC died where all links and docs are.
 
The amp is straight out of the box. No mods.
The spec says there is a bass boost at 25 hz but only +2db so I decided to not take that in consideration. At some resellers it's claimed +6db at 25 hz but the manual states 2 db and I think thats the correct value. I see now that LP-filter is 18 db/oct but It's probably flatter.

It would have been nice to disable the bass boost and perhaps kick in the subsonic-filter at 30 hz instead of 20. I have not really checked out the amp-module but there is alot of smd components on it and thats where I'm usually give up when it comes to electronics :)

Regards [email protected]
 
Funny, I was interested in trying to use two Silverflute 5.5" drivers for this same purpose but for the CHR-70's in my office. I was even contemplating the Volvotreter 30hz tapped horn using the Tang-Band 6.5". Hows the sound? Output, Linearity, definition ect.

When crossed low at 55-60hz I really like whats coming out of it. Sounds distinct and solid. If I try to crossover it any higher it sounds hollow and even with blindfold there would be no problem to point out it's exact position.

This is really an one octave sub.

I dont know about the output power but as mentioned earlier it has no problem to cope up to the levels the alpair does.

Regards [email protected]
 
..This is really an one octave sub..

My first impression too..

..If I try to crossover it any higher it sounds hollow and even with blindfold there would be no problem to point out it's exact position..

You might find the reason here:See the picture.

Did you use measured TS in your simulations?

b :)
 

Attachments

  • 2xTB-W4-792sj_TH.JPG
    2xTB-W4-792sj_TH.JPG
    732.2 KB · Views: 217
When crossed low at 55-60hz I really like whats coming out of it. Sounds distinct and solid. If I try to crossover it any higher it sounds hollow and even with blindfold there would be no problem to point out it's exact position.

If possible, try LIGHTLY stuffing the first 1/3 to 1/2 of the horn, starting from the throat, to see if that will make a difference to that "hollow" sound. See The Subwoofer DIY Page v1.1 - Projects : A "Proof of Concept" tapped pipe** - introduction - I was able to substantially change the output at higher frequencies (and even eliminate a deep notch) by stuffing.

20090829-tp-fr4.gif
 
My first impression too..
Did you use measured TS in your simulations?
b :)

The limited bandwidth is not a problem since it will just fill up the interval between 30-50 (or maybe 60) hz. The output is more than enough as well.

Whats worse is if I have used wrong ts parameters. Where did you find yours? I have used this data sheet:

Europe Audio

From what I can see bl and mmd differs from your values.

If possible, try LIGHTLY stuffing the first 1/3 to 1/2 of the horn, starting from the throat, to see if that will make a difference to that "hollow" sound. See The Subwoofer DIY Page v1.1 - Projects : A "Proof of Concept" tapped pipe** - introduction - I was able to substantially change the output at higher frequencies (and even eliminate a deep notch) by stuffing.

That could be useful info if I, at some point, need it to go higher i frequency.
Inserting the stuffing would be a problem thou since the box is glued.

Regards [email protected]
 
So did I ,
I also consistency checked the data: Double click Sd in HR and you will find out why I used different parameters.
b

Okay I see now that you have entered Fs and Qes in addition of a assumed correct Cms and Re to calculate BL. Is that preferable over entering manufacturing sheet values direct into the BL field?

Regards [email protected]
 
You might find the reason here:See the picture.

Did you use measured TS in your simulations?

I noticed this yesterday, but didn't have time till now to post about it. As soon as I saw the 0.17 Qes/0.16 Qts I knew there was almost positively something wrong with either the published specs or his, if measured.

Using the EZH spreadsheet and the more likely specs (though best as always to use measured), I get a near perfect match to my calculations once S1 is increased, so maybe a good enough alignment if others want to try it or one can see if there's any benefit to a four section:

GM
 

Attachments

  • TB W4-792SJ 17.34 Hz conic TH [increased S1 EZH - published].gif
    TB W4-792SJ 17.34 Hz conic TH [increased S1 EZH - published].gif
    7.4 KB · Views: 383
Last night I did a quick measurement. Mic was flush the terminus (or mouth the correct term?). It seems like Bjorno's first sim, with my faulty ts parametrs, is pretty much what comes out of the horn. One have to take the measurement with a pinch of salt, of course, but the similarities to the sim is quite significant. Filter on amp is set on 150hz.

Thanks for pointing this out! My next build had likely suffered from faulty ts parameters if this hadn't come to my knowledge.

I noticed this yesterday, but didn't have time till now to post about it. As soon as I saw the 0.17 Qes/0.16 Qts I knew there was almost positively something wrong with either the published specs or his, if measured.

Using the EZH spreadsheet and the more likely specs (though best as always to use measured), I get a near perfect match to my calculations once S1 is increased, so maybe a good enough alignment if others want to try it or one can see if there's any benefit to a four section:

GM

I will definitely look into measure the ts parameters.


Regards [email protected]
 

Attachments

  • simvsreality.jpg
    simvsreality.jpg
    146.2 KB · Views: 387