TH with very short L12 and L34 ?

I had a few hours in an airport and decided to play a little with Hornresp.
I have very limited experience so please correct my mistakes.

Target is "micro-PA-sub". 40Hz low end would be great, 50-60 will probably do fine for pop-disco-fun.

The target-driver is one I have already and not the "typical" TH-candidate.
10" from Monacor/StageLine: SP252E, 4Ohm, 75W.

Impedance (Z) : 4 Ohm
Resonant frequency (fs) : 42 Hz
Music power (PMAX) : 150 WMAX
Power rating (PRMS) : 75 WRMS
Sensitivity : 94 dB
Suspension compl. (Cms) : 0.33 mm/N
Moving mass (Mms) : 43 g
Mech. Q factor (Qms) : 3.13
Electr. Q factor (Qes) : 0.60
Total Q factor (Qts) : 0.50
Equivalent volume (Vas) : 57 l
DC resistance (Re) : 3.75 Ohm
Force factor (BxL) : 8.5 Tm
Voice coil induct. (Le) : 1.1 mH
Voice coil diameter : 35.5 mm
Voice coil former : aluminium
Linear excursion (XMAX) : ±4 mm
Eff. cone area (Sd) : 353 cm2
Magnet weight : 960 g
Mounting cutout : Ø 232mm
Mounting depth : 110mm
Dimensions : Ø 255 mm x 110 mm

First the driver in a 50L BR to get a baseline for comparison: HelmHolz= 40Hz.
HR record attached.
[IMGDEAD]http://i64.tinypic.com/33li0xx.png[/IMGDEAD]
Lets call that a 92db/1W sensitivity. Not "flat" by any standard, but most likely OK for the purpose.
 

Attachments

  • 252E-BR-50L-92dB.txt
    923 bytes · Views: 10
So, the TH part. I started with the "THAM10 redesigned big version from post 12": http://www.diyaudio.com/forums/subwoofers/182209-tham10-10-compact-th-proposal.html

That gave a pretty rugged response and its a little big for the purpose.

After a lot of playing with the HR parameters I ended up with this:
241v43n.png

v78snl.png


lets call that 95dB from 60Hz.

Odd thing is that i end up with very short L12 and L34 so the horn-bend is not very typical. HR record attached.

Trying to fold the horm I get something like below. Not final, - but total volume is close to simulation.
2ih9w6p.png


Outer dimensions are approx 50*50cm. Width is 30 and subtracting wall-thickness that end up close to 58Liter as the sim.

It take 50W to reach 4mm excursion and do around 112dB in 2Pi. I have no idea if this "good" or not.

Anyone tried a fold like that? Does it still qualify as a tapped horn?

And is it complete madness to have a low cutoff around 60Hz?

Kind regards TroelsM
 

Attachments

  • 252E-TH-60L_1.txt
    958 bytes · Views: 14
Last edited:
Odd thing is that i end up with very short L12 and L34 so the horn-bend is not very typical. HR record attached.

Trying to fold the horm I get something like below. Not final, - but total volume is close to simulation.

The Hornresp sim and the fold drawing are not even close to the same. Not even vaguely similar.

In the sim L12 is very short - too short in fact to even fit the driver. In the drawing L12 is very long, maybe 30 cm or so?

In the sim there is an abrupt and severe flare rate change at S2. in the fold pic there is no flare rate change at S2 at all.

The mouth area in the picture isn't really accurate to the sim either.

You did a very very bad job of folding this, the folded plan isn't even close to the sim in a variety of ways.

About the sim in particular - you have a 58 liter tapped horn tuned around 50 hz but with a low knee closer to 70 hz.
And you have a 50 liter ported box tuned around 40 hz.

The tapped horn is moderately louder at some frequencies but the ported box can play almost an octave lower. I'd much prefer the ported box, 70 hz tuning is way too high for my taste.

If you want to make a fair comparison make the tuning (and ideally also the low knee) the same on both the tapped horn and the ported box. And if you do that you will find that the tapped horn won't have much (if any) advantage unless you make the tapped horn much larger than the ported box, at least not in the sim itself. That's just the way physics works. The tapped horn will always have some amount of advantage due to less port compression (and usually also less power compression) but that doesn't show in the sim.
 
Last edited:
@justaguy: thanks for the informative reply

I see that the fold and sim are not similar. Mostly the folding-sketch was a test to see if a really short L34 could be implemented. L12-lenght do not make much difference in the sim, so the original question mostly relates to L34.

About the TH-tuning: what determines that frequency?
I'll be happy to trade some of the low end for higher overall efficiency. You are probably correct that 60-70Hz is too high.

Surely the BR goes deeper and have a broader range. I'm not expecting magic here. I (sort of) understand that the TH can provide a higher eff in a narrow band, but it dos not come free.

About your last paragraph: as I understand it, BR will always have a limited efficiency no matter how large the box is. A horn or TH can in some cases become more efficient with a larger box. The question is how large a box will it take to make the TH more efficient/better compared to the BR.
 

just a guy

Member
2006-05-12 6:59 pm
@justaguy: thanks for the informative reply

I see that the fold and sim are not similar. Mostly the folding-sketch was a test to see if a really short L34 could be implemented. L12-lenght do not make much difference in the sim, so the original question mostly relates to L34.

About the TH-tuning: what determines that frequency?
I'll be happy to trade some of the low end for higher overall efficiency. You are probably correct that 60-70Hz is too high.

Surely the BR goes deeper and have a broader range. I'm not expecting magic here. I (sort of) understand that the TH can provide a higher eff in a narrow band, but it dos not come free.

About your last paragraph: as I understand it, BR will always have a limited efficiency no matter how large the box is. A horn or TH can in some cases become more efficient with a larger box. The question is how large a box will it take to make the TH more efficient/better compared to the BR.

The flare length and shape determine the tuning and the low knee frequency. Size vs tuning determines efficiency.

The reason the ported box has a limited efficiency is because there's only one impedance peak inside the passband so even if you make the box HUGE it won't have efficiency across the whole passband, it will only have the spike of efficiency down near tuning.

The tapped horn on the other hand has several impedance peaks inside the passband so if you make the box HUGE it's possible to achieve higher efficiency across the whole passband, up to 3 octaves of gain bandwidth.

Effective tapped horns are usually 2 - 4 times larger than ported boxes for the same driver and tuning. As mentioned, due to power and port compression the tapped horn will always have an advantage even if it doesn't appear so in the sim but if you want it to be really effective and have it show in the sim it needs to be significantly larger than a ported box tuned to the same frequency.
 
Accepting a larger box and trying to get the sim closer to a fold that is doable I'm experimenting with something like this:
[IMGDEAD]http://i64.tinypic.com/2rxejc4.png[/IMGDEAD]

[IMGDEAD]http://i64.tinypic.com/1z6qz5c.png[/IMGDEAD]

Size would be 55*55cm and 45cm wide. That would give approx 100Liter internal volume minus approx 8-10Liter for "horn-path".

Fold is NOT correct in the sketch, but I think L12, L23, L34 and total volume are close.

[IMGDEAD]http://i68.tinypic.com/2h4du8p.png[/IMGDEAD]

So basicly twice the volume of a decent BR and approx 3-4dB more volume, -dos that make kind of sense?

Another thing: apparently this driver do not agree with high compression, in fact the sim is with S2 higher than SD. If S2 becomes to low the "passband" will get a 10dB suck-out in the "middle". To little motor-force (BL)?

Kind regards
 
You still have a MASSIVE flare rate change at S2 that's not reflected in the drawing. And then you have a large flare rate change after the second bend in the drawing that's not there in the sim. In fact there are several flare rate changes in the drawing that are not in the sim. And the whole mouth area still isn't really accurate to the sim.

The drawing still isn't even remotely close to the sim. At this point I'm not sure why you are drawing the fold when you are still experimenting with box size and comparing to ported box, but I guess it's all good practice. Just not there yet with the folding skills.

Twice as large as the ported box should give a bit of sensitivity gain, yes.
 
@Justaguy: wow. Your fast and honest. Very Honest. Thanks.

The sim do not suggest much difference with regards to the first section from S1 to S2.

The mouth area is also of little consequence according to the sim. I would guess that there is some error in the way the mouth-area is calculated as the flare is massive and short.

The folding was merely a sketch to verify it was possible. If total volume and section-lengths can fit in there, I guess that the folding will be possible in some way.

Still just experimenting and learning.

Kind regards TroelsM
 
The sim do not suggest much difference with regards to the first section from S1 to S2.

Really? Really?

[IMGDEAD]http://i66.tinypic.com/357krd2.png[/IMGDEAD]

The midpoint of the driver is S2. You see what's happening at S2? There's a MAJOR flare rate change. Look at your drawing. Is there a flare rate change at S2 12.9 cm from the closed end? No there is not. You have a huge flare rate change right after the 2nd bend which isn't supposed to be there according to the sim, and no flare rate change where the sim says there is supposed to be.

Now the schematic drawing is axisymmetric (round). When you draw that out as flat (rectangular) with 2 assumed parallel side walls the flare rate change at S2 is going to look a LOT more dramatic than shown in the schematic, because S2 is more than 5 times larger than S1, and S3 is not a whole lot bigger than S2.

Now as for the sim itself (assuming you entered the t/s parameters right and I copied them right as my eyes are a bit blurry right now) it took literally two seconds to change your sim to this.

[IMGDEAD]http://i68.tinypic.com/f1mbso.png[/IMGDEAD]

I put the sliders to "auto" so there is NO flare rate change in this at all. Folding it is as simple as adding S1 and S4 together (+ wood thicknesses) for the box cross sectional area, dividing the combined path length by 2 for the box length and drawing in the divider/baffle board. It literally could not be more simple to fold this.

But it's still not a good design IMO, as the tuning frequency is well below the low knee.

This is a bit better. Everything is still set to "auto" so it's a simple fold, and the low knee corresponds to the tuning (dip in excursion) frequency much better. But L12 is not equal to L34 so it's not a single fold design, it's slightly more complex to fold.

[IMGDEAD]http://i67.tinypic.com/219tgch.png[/IMGDEAD]

This second variation took an extra 5 seconds to complete so this is a pretty simple process when all the sliders are set to "auto". Maybe practice with the sims a bit more and put off the folding pics until later. The folding is going to need a LOT of work based on your first two attempts.
 
Hi.

I really (really! really!) appreciate you taking the time to do this.

I agree that your sims looks great, but when I enter the same numbers I get approx 240 Liter of volume. That double the volume I was simulating with.

While 240L may be the "optimal" volume for the driver I simply cannot build that big a box. That's the reason I was sketching along-side the simulations. to get a feel for the size. The huge increase in size gives no extra sensitivity, only lower knee. For this application I rally dont need much below 40-50Hz.

You mention that the tuning in my sim is below the lower knee. I sort-of understand that this may be a less than optimal arrangement. I see people talking about that TH-drivers should have high Fs. I wouls have guessed that 42Hz fall into "semi-high" Fs, but thats probably wrong.

Kind regards
 
Last edited:
Yeah, the sims I did are a lot bigger than yours.

It's not just about fs on it's own, the fs is supposed to be higher than the TH tuning. That's why your 42 hz driver sims so comfortably in a 30 hz TH but not so much in a 60 hz TH.

Anyway if I have a few minutes today maybe I'll see if I can make a sim work with 60 liters although it's not going to beat a ported box by much.
 
So here's a bunch of random stuff. Two of these are tapped, two are not. As you can see, tuning is quite similar in all these, between 40 and 44 hz (the big dip in excursion) and the low knee is fairly similar. The trade off is whether you want a very strong low knee or if you want sensitivity higher up in frequency, each of these could trade one for the other. All of these are around 92 - 95 db / watt. They are all between 76 and 102 liters. And excursion is quite similar in all of them.

That's the point. No matter what type of enclosure you want, if they are roughly the same size and tuning you will get roughly the same spl.

The one that looks the best to me is the last one, the smallest one (76 liters), it's the most similar to a ported box and it's got the smoothest, most extended response. This is why the ported box or reverse taper tl is a good option.

I didn't spend more than 10 or 12 minutes on all of these combined so none are optimized in any way, I just did a bunch of stuff to illustrate how spl is very similar regardless of enclosure type when limited by size and tuning.

For this driver when size is limited to 100 liters or less I probably wouldn't do a tapped design, I would probably do a reverse taper tl or some variant of ported box.

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