Hi,
Interesting paper with some interesting numbers in there. IMO it does not
sufficiently explore the rather more succinct method of an overdamped
low tuned 4th order system combined with a peaking high pass filter
to provide the EQ and the subsonic filter in one go.
IMO there are a lot of smoke and mirrors arguments in that paper, that
whilst true, are actually misrepresenting the practical reality by some
clever misdirection, the pudding is being severely over egged IMO.
It is a more flexible EQ system than an overdamped 6th order, and
more complicated, and doesn't include the subsonic filter, and it
would allow you to use driver not suited to what I'd call a sensible
6th order alignment in what is effectively a similar alignment.
However if you do go for an overdamped 6th order, you can make
the Q of the highpass filter adjustable, for room matching, but
you do need to choose the driver and enclosure carefully.
It does illustrate what you can do with low tuned overdamped vented
alignments compared to sealed boxes, the advantages are overwhelming,
as low bass actually needs SPL more than purely frequency extension.
rgds, sreten.
Vented boxes will end up rolling off at 24dB/octave. 6th order 36dB/octave.
The trick is to forget maximally flat alignments and overdamp the alignment
such that the roll-off starts off near first going towards second order for
most of the bass range of interest, accelerating to higher orders basically
below the real area of reproduction interest, combined with room gain works.
Last edited:
My ultimate preference would be for a sealed box Bessel alignment with LT and judicious EQ that takes room effects into account to produce a flat response. This does not have to involve very big boxes. For example, using two big woofers, one on each side wall, can make for a relatively slim and rigid box. Low Fs woofers should be used for the lowest possible group delay. This will minimize the boost required of the LT circuits and, if excursions are not very large, there will be relatively little air nonlinearity. A slick marketeer might even call it the "Stepford" 
alignment.
alignment.
Last edited:
Sending signal below the tuning frequency of a subwoofer is a bad idea. (unless someone likes distortion
) That's why I always have the High-pass set 4hz above the tuning frequency of a BR sub.
I've been building these for years. Bob is just describing a LLT sub + LT. I have no problems with his maths, although using a comparison sub tuned to 65hz is a little silly, Also using a very low Q sealed sub with LT would sound like
IMO. A properly designed LLT does not need a LT circuit to achieve flat in-room response to below 20Hz.
Last edited:
"Stepford"
I likeAs an added bonus if you buy in the next 20 min., they come with a wife who just loves to listen to you go on and on about the nuances of wizzer cones.
Sending signal below the tuning frequency of a subwoofer is a bad idea. (unless someone likes distortion
That makes very good sense. I think it is vital to use such filters for all systems I can think of but for BR, seems like an odd mixture of techniques.
But then why use BR if you are throwing away so much "baby with the bath water"?
Ben
Additional 2-cents:
while it is somewhat helpful not to direct any signal to the BR below some point (say, 4 Hz above resonance), it still does not control the cone below that point, the way a sealed box does. These are different techniques.
Perhaps the benefit of a BR-plus-high-pass arises from being able to create a pretty small BR box (without raising the system resonance, as with a sealed box), using the tuning to control the system resonance peak, an using the filter to limit uncontrolled cone motion in the low bass.
Ben
while it is somewhat helpful not to direct any signal to the BR below some point (say, 4 Hz above resonance), it still does not control the cone below that point, the way a sealed box does. These are different techniques.
Perhaps the benefit of a BR-plus-high-pass arises from being able to create a pretty small BR box (without raising the system resonance, as with a sealed box), using the tuning to control the system resonance peak, an using the filter to limit uncontrolled cone motion in the low bass.
Ben
I have to say, some of your analogies confuse me...
What is being thrown away?
I should point out that if I build a BR box they are semi-LLT tuned ~14hz and High-pass at 18Hz. I suspect you're thinking about tunings well above that. (which i don't like for many reasons)
You may be right, I was just thinking that with this extra low tuning you would still need a bit of help with the EQ as it rolls off, not quite able to keep up with the room gain without it. In my own situation the far bigger problem is the havoc created by the room modes...
With an undersized low tuned sub some EQ could be used as bob suggests.
If you can't treat the room properly (low WAF
) have you tried notch filters, I use them all the time and the brain dosen't notice the missing frequencies as much as it notices the drone of standing waves.When you wrote, "That's why I always have the High-pass set 4hz above the tuning frequency of a BR sub" I didn't for a second suppose you were talking about using a sub with a resonance of 14 Hz! It would be an understatement to say that is unusual.
Nor do I, or many people, have an opinion about the effectiveness of that kind of system since so very few people have ever been exposed to one.
If you were talking about boxes tuned for say, 30 Hz, then filtering out below 34 Hz would indeed be tossing out the content below 34 Hz along with the undesirable consequences of pumping signal into a BR below its resonance. All things considered, just maybe a good trade-off... I can't say. Perhaps for a small driver in a small box, a good design.
Bad me for not immediately supposing you were using a 14 Hz driver. Aside from my 12 Hz driver in an AR-1, can't say as I'd know where to find one.
Ben
Last edited:
Hi all,
A question that seems related to the topic being discussed here.
I was playing with the idea of building a (subwoofer) TL with LT. Tuning the TL so it's initial roll-off is ~12dB/oct (a small volume box with the line tuned to a low Freq with a taper of 10:1) and then boosting the drooping response with an LT. The benefit of using a TL is that it doesn't unload the driver below resonance (right?).
It seems that is almost the same as EQSS.
This gives a definite improvement in low freq SPL capability, GD is very well controlled, cone excursion is less then a sealed + LT for a given SPL.
Overall this should yield a system with more SPL capability and/or lower distortion, it almost seems to good to be true.. Am I missing something?
A question that seems related to the topic being discussed here.
I was playing with the idea of building a (subwoofer) TL with LT. Tuning the TL so it's initial roll-off is ~12dB/oct (a small volume box with the line tuned to a low Freq with a taper of 10:1) and then boosting the drooping response with an LT. The benefit of using a TL is that it doesn't unload the driver below resonance (right?).
It seems that is almost the same as EQSS.
This gives a definite improvement in low freq SPL capability, GD is very well controlled, cone excursion is less then a sealed + LT for a given SPL.
Overall this should yield a system with more SPL capability and/or lower distortion, it almost seems to good to be true.. Am I missing something?
Last edited:
TL do not control cone excursion below Fb, it rises like a BR.
Cone excursion would be less around FB in a TL or a BR than sealed, but more below Fb.
Hi Weltersys,
I did see in my sims that excursion rises below Fb, but I also noticed that the acoustic impedance of the line only drops slowly below Fb. This leads me to believe that indeed eventually the driver becomes unloaded at very low frequencies, but in the designs I simmed the impedance only get's real low when the frequency drops to single digit numbers. So to limit excursion a HPF filter is needed (in my case a second order around 14Hz).
Is my train of thought correct (the acoustic impedance of the line controls cone excursion), or am I way of?
Check out the tool I use to sim TL's, made by a fellow forum member, very nice piece of SW (it's the only free windows based tool I know of simulating TL's):
http://www.diyaudio.com/forums/software-tools/220421-transmission-line-modelling-software.html
I did see in my sims that excursion rises below Fb, but I also noticed that the acoustic impedance of the line only drops slowly below Fb. This leads me to believe that indeed eventually the driver becomes unloaded at very low frequencies, but in the designs I simmed the impedance only get's real low when the frequency drops to single digit numbers. So to limit excursion a HPF filter is needed (in my case a second order around 14Hz).
Is my train of thought correct (the acoustic impedance of the line controls cone excursion), or am I way of?
Check out the tool I use to sim TL's, made by a fellow forum member, very nice piece of SW (it's the only free windows based tool I know of simulating TL's):
http://www.diyaudio.com/forums/software-tools/220421-transmission-line-modelling-software.html
Your question would get a better answer from the people that design the simulation programs than me, but sounds like your train is on the right track.
I want to know if the use of a HP-Lowcut filter is relevant when implementing a LT. The scenario I am talking about is a 12" woofer with 300 Wrms and Xmax of ~7mm in a to small enclosure of 45 liters (1.6 cft). The woofer is crossed to a compression driver around 1-1.2 kHz and will play the mids. This is to be taken into considerartion here. Another option here could be a vented EQ-assisted enclosure following Don Keele.
I regularly listen to deep bass music (dub, D'n'B, all sorts of Electronica) which have very low frequencies in the source material. Now I do not mind if the speakers cannot reproduce (all of) them, but I do mind if 1) they get destroyed because the woofer is pushed to hard, 2) the higher cone excursion introduces considerable distortion and diminishes mid frequency reproduction.
So: How much does cone excursion matter? I found that if I installed a 2nd order BW high pass filter at 20 Hz I can limit cone excursion quite well. But group delay is shooting up and is getting even higher than in a vented enclosure from ~25 Hz downwards. I suppose this means the base will loose definition? If I do not use a lowcut, the woofer reaches 4mm cone excursion at 20 Hz and a peak of 5 mm subsonic. Allthough I wonder if this is fixed to any SPL or what SPL WinISD uses here. Thechnically the woofer permits such cone excursion, my question is: What does it mean soundwise?
I regularly listen to deep bass music (dub, D'n'B, all sorts of Electronica) which have very low frequencies in the source material. Now I do not mind if the speakers cannot reproduce (all of) them, but I do mind if 1) they get destroyed because the woofer is pushed to hard, 2) the higher cone excursion introduces considerable distortion and diminishes mid frequency reproduction.
So: How much does cone excursion matter? I found that if I installed a 2nd order BW high pass filter at 20 Hz I can limit cone excursion quite well. But group delay is shooting up and is getting even higher than in a vented enclosure from ~25 Hz downwards. I suppose this means the base will loose definition? If I do not use a lowcut, the woofer reaches 4mm cone excursion at 20 Hz and a peak of 5 mm subsonic. Allthough I wonder if this is fixed to any SPL or what SPL WinISD uses here. Thechnically the woofer permits such cone excursion, my question is: What does it mean soundwise?
Well, we're always trading efficiency for BW and since excursion increases 4x/octave for a given piston area [Sd] with sealed, one runs out of Xmax real quick down low, so seems prudent to use one.
Vented reduces it over ~1/2 octave IIRC, but really need to sim in Hornresp
or similar to see what your power handling limits are.
Vented reduces it over ~1/2 octave IIRC, but really need to sim in Hornresp
or similar to see what your power handling limits are.
LT transform is essentially a shelving filter, which increases signal level equally below the "transform" frequency. This will consequently increase excursion. For this reason it is wise to limit low frequency bandwidth by adding a high-pass filter. I started out by having both LT and HP filters, but realised that both functions could be achieved by using a single high Q HP filter. Yes, GD will increase, but it is important to compare before vs after.
GD is affected by the shape of the transfer function. Smooth transitions have lowest GD, and sharp deviations make it shoot up. If you consider that where you previously had a rolling-off curve, it is now flat, you will see that equalisation has in fact improved GD. Then you have to make a judgement in terms of the knee of your final HP curve. The sharper that knee, the more pronounced will be the GD. To every pro there is a con. It comes down to, as you have rightly alluded to: how much is too much? My advice: compare a few alignments and then take your pick.
Yes, a linkwitz transform, basically boost the schmoo to the bass.
Think bag end, take an 18-21", put it in a small sealed box.
It rolls off 12db/octave below 80hz.
Now put a low pass filter on it say starting at 20hz (so the driver will be 12db down at 40hz and 24db down at 80hz).
Put that filter with that sub/box and voila, flat response to 20hz, but.......
It took 24db of boost to get there, that eats a lot of power and xmax.
Bag end had a 21" sub reviewed in "way down deep ii" although it rolled at 64hz, 12db boost at 32hz, 12more db boost at 16hz. This added a lot of distortion past 100db compared to other subs.
Way Down Deep II Bag End S21E | Sound & Vision
Think bag end, take an 18-21", put it in a small sealed box.
It rolls off 12db/octave below 80hz.
Now put a low pass filter on it say starting at 20hz (so the driver will be 12db down at 40hz and 24db down at 80hz).
Put that filter with that sub/box and voila, flat response to 20hz, but.......
It took 24db of boost to get there, that eats a lot of power and xmax.
Bag end had a 21" sub reviewed in "way down deep ii" although it rolled at 64hz, 12db boost at 32hz, 12more db boost at 16hz. This added a lot of distortion past 100db compared to other subs.
Way Down Deep II Bag End S21E | Sound & Vision
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.