But the trade off, small boxes heavily compensated....
Requring great dispacement capability, usually nowdays by high excursion drivers.
By simply looking at BL vs excursion curves of typical drivers there will shure be greater compression/distorsion compared to alignments that require less excursion for a given sound pressure.
Also, if we are going to compensate the low sensitivity of closed boxes at low frequencies a lot of power will be nessesary. Power compression by voicecoil heating is another factor to take into account.
The usually low sensitivity of high excursion drivers due to low ratio of voicecoil in gap will only add to power requirements.
Requring great dispacement capability, usually nowdays by high excursion drivers.
By simply looking at BL vs excursion curves of typical drivers there will shure be greater compression/distorsion compared to alignments that require less excursion for a given sound pressure.
Also, if we are going to compensate the low sensitivity of closed boxes at low frequencies a lot of power will be nessesary. Power compression by voicecoil heating is another factor to take into account.
The usually low sensitivity of high excursion drivers due to low ratio of voicecoil in gap will only add to power requirements.
The candidate has 100 points !!!!
Regarding power demands: Since the driver-box combination already went down to 60 Hz approx and the final Q was only 0.5 the rise in power wasn't very high for normal program material.
Listening to the infamous 1812 Telarc recording or some hip-hop tracks would change things a bit of course.
The important thing (as is true for most engineering tasks): Don't overdo it.
I think everyone can agree that expanding a woofer that is already capable of doing 60Hz down to 30 Hz is less proplematic than EQing a woofer - that does only reproduce 100 Hz by itself -down to 20 Hz.
And yes, LTF does influence group delay distortion. It depends on what one is doing. Transforming to a lower Q means better group delay properties.
Regards
Charles
LTF....
Well, sorry to pop in on this thread, but you're talking about the subject, so I thought....
I have some 'el-cheapo' 10" ers (mabe by some obscure company, and sold under the name of another obscure company) that I put into a WO-36 (Wicked One odd-thinggy) but I don't like it at all... Might be my not so perfect construction, or the fact that the WO is rahter car-oriented, what ever (More over, this huge anumal doesn't fit in my livingroom, so...)
So I thought, given these specs:
Vas =145 liters
Fs = 34 Hz
Qms = 3,6
Qes = 0,52
Qts = 0,45
Sens: 92 bD, 300W
(As claimed by the manufacturer!!!!)
I'd make a 190/200 l box (approx. 122x40x45 cm) But both in, use my cheap sub-plate amp, and a LTF like this
Fo = 54,70 Hz
Qo = 0,70
Fp = 18 Hz
Qp = 0,707 (K = 3,065)
This gives a max gain of 19,31 dB
My question is: Will it work? Will it sound? How can I determine if the drivers a willing to do what I'll ask them to do?
I don't feel like spendig 60 Euros on MDF, just to find out that I can shove it in the stove (Like my WO-36 is goning to end up like)
Cheers, Paul
OH, BTW: It's for music mainly, no high SPL's needed, rather large listening/living room, KEF CS7's as mains (Blame me, I STILL need to paint them after 2 1/2 years...)
Well, sorry to pop in on this thread, but you're talking about the subject, so I thought....
I have some 'el-cheapo' 10" ers (mabe by some obscure company, and sold under the name of another obscure company) that I put into a WO-36 (Wicked One odd-thinggy) but I don't like it at all... Might be my not so perfect construction, or the fact that the WO is rahter car-oriented, what ever (More over, this huge anumal doesn't fit in my livingroom, so...)
So I thought, given these specs:
Vas =145 liters
Fs = 34 Hz
Qms = 3,6
Qes = 0,52
Qts = 0,45
Sens: 92 bD, 300W
(As claimed by the manufacturer!!!!)
I'd make a 190/200 l box (approx. 122x40x45 cm) But both in, use my cheap sub-plate amp, and a LTF like this
Fo = 54,70 Hz
Qo = 0,70
Fp = 18 Hz
Qp = 0,707 (K = 3,065)
This gives a max gain of 19,31 dB
My question is: Will it work? Will it sound? How can I determine if the drivers a willing to do what I'll ask them to do?
I don't feel like spendig 60 Euros on MDF, just to find out that I can shove it in the stove (Like my WO-36 is goning to end up like)
Cheers, Paul
OH, BTW: It's for music mainly, no high SPL's needed, rather large listening/living room, KEF CS7's as mains (Blame me, I STILL need to paint them after 2 1/2 years...)
Hi,
that's a lot of gain. Do you realise the message it's telling you.
+19.31db is almost 100times.
For a 300W driver limit that means the rest of the frequency range is working upto a maximum of 3W when the bass driver hits the +19.3db induced thermal limit.
If the frequency is quite low your driver may run out of Xmax before you even get to 3W. For a midband or upper bass sensitivity of 92db/W then your EQed bass driver runs out of puff before you have reached a peak SPL of 97db. I think you would be very disappointed with that limitation after all your building effort.
I would try for less bass extension and much less gain. Maybe just +10db and see how the numbers stack up. Even 10db less gain will only increase your peak SPL to about 107db. Headbangers will think this is too quiet.
All you speaker heads.
Have I got this right?
that's a lot of gain. Do you realise the message it's telling you.
+19.31db is almost 100times.
For a 300W driver limit that means the rest of the frequency range is working upto a maximum of 3W when the bass driver hits the +19.3db induced thermal limit.
If the frequency is quite low your driver may run out of Xmax before you even get to 3W. For a midband or upper bass sensitivity of 92db/W then your EQed bass driver runs out of puff before you have reached a peak SPL of 97db. I think you would be very disappointed with that limitation after all your building effort.
I would try for less bass extension and much less gain. Maybe just +10db and see how the numbers stack up. Even 10db less gain will only increase your peak SPL to about 107db. Headbangers will think this is too quiet.
All you speaker heads.
Have I got this right?
Yes Andrew spot on. A pair of cheap 10 inch speakers are not going to hit 18Hz in a sealed box at any worthwhile SPL, certainly not to keep up with some decent main speakers.
Whilst the high gain is not that good, it's not a show-stopper as long as you have the speakers with capabilities to handle the power and SPL requirements.
Whilst the high gain is not that good, it's not a show-stopper as long as you have the speakers with capabilities to handle the power and SPL requirements.
Oups....phase_accurate said:
I forgot to mention I'd but BOTH drivers in the box... // so I end up having 4 ohms.
(Andrew's values would thus become: 19 dB, 600W, 6W)
I'll fiddle with the LTF S-sheet to obtain some 'reasonable' values.
I don't bang my head quite often :-D Me thinks maybe around 97 or 100 dB'd do the trick for me...
Bye, Paul
(Ow, and, yes, those KEF's are like 87 dBw-1, so....)
bibster said:
IMO the place to start when designing an LT system is to look at the max SPL that is available from your driver and box combination. If this does not meet 97dB at your chosen -3dB or (even higher up the range), you will just be wasting time with the LT spreadsheet and it's time to find new drivers or change your box size, usually the former.
If the Kefs are floorstanders and vented, I think you will end up having to buy new drivers for your sub unless you go for an alternative box design.
From recent posts on here of what WinISD says against what people seem to be realising in the actual houses, it would seem that a lot of the SPL predictions are wildly optimistic from numerous software packages. Linkwitz has some equations on his site which seem to be more accurate.
edit: on the Pluto+ page he has two spreadsheets he used to compare the XLS12 and XLS12, either of these can be used. They are extremely comprehensive.
Once upon a time I used this tiny sub:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=25015&highlight=
But richie convinced me that LT can do the same job as ACE electronics. ACE electronics measure the current in through the speaker and provide by complex feedback any complex output impedance of the amp. Means you can adjust the system to any desired effective Thiele/Small parameters.
LT is less difficult and can deliver the same results. LT is a 4th order filter. Two orders are exactly compensating poles and zeros of your sealed box system without correction. This means also phase shift and group delay ! Cool !! And the 3rd+4th order define any new desired transfer function of the system.
No matter which electronic compensation you use for an undersized box, you will need high power in the low frequencies. Also distorsions usually increase.
y current approach is IRPS with LT. IRPS=in resonance principle speaker. If we look at the efficacy of a sealed box, you are getting quite values around the resonance frequency. The impedance is high, current and power lower, but sonic output strong. Originally I was thinking to add mass on the cone and my measurements indicated even in this situation a better efficacy than living with high natural fres and electronic correction. But in the end I decided to make the enclosure large enough to have a reasonable natural resonance frequency of the sealed box around 40Hz without adding mass. 40Hz is a frequency where still a good portion of bass signal can be found in normal music. In addition I am adding a LT to this natural IRPS system which is changing the resulting roll off frequency down to 12Hz.
All together a trade off between size, required power, possible SPL.
Up to now I enjoy this trade off quite much.
I must agree to statement that micronizing the box to exzessive small values and correcting with brute force electronics does sound less impressive than larger sized designs. I definitely hear this by comparing my new design (Two cubes 40cmx40cmx40cm, with one Adire Sadhara driver each) to my older supersmall 12liter/11driver system. The new system offers much more relaxed and impressive bass reproduction also at low signal levels, where the distorsions should low in both systems. Up to now I did not fully analyze the reasons for this difference, but at least my ears tell me very clear...
For my current situation the two IRPS Sadhara cubes with LT are doing a good job and in my living two 12" drivers with+/-27mm lin Xmax are mostly sufficient. I don't want it giant.
Hm,... 8Hz you say... for pipes... I could change the LT to go deeper. But my reason NOT to go so deep was the fact that our human body is sensitive to infrasound. From what I know the most dangerous frequency range of infrasonic weapons is somewhere between 5Hz....10Hz. How can our pope support such things in the church???!!!
http://www.diyaudio.com/forums/showthread.php?s=&threadid=25015&highlight=
But richie convinced me that LT can do the same job as ACE electronics. ACE electronics measure the current in through the speaker and provide by complex feedback any complex output impedance of the amp. Means you can adjust the system to any desired effective Thiele/Small parameters.
LT is less difficult and can deliver the same results. LT is a 4th order filter. Two orders are exactly compensating poles and zeros of your sealed box system without correction. This means also phase shift and group delay ! Cool !! And the 3rd+4th order define any new desired transfer function of the system.
No matter which electronic compensation you use for an undersized box, you will need high power in the low frequencies. Also distorsions usually increase.
y current approach is IRPS with LT. IRPS=in resonance principle speaker. If we look at the efficacy of a sealed box, you are getting quite values around the resonance frequency. The impedance is high, current and power lower, but sonic output strong. Originally I was thinking to add mass on the cone and my measurements indicated even in this situation a better efficacy than living with high natural fres and electronic correction. But in the end I decided to make the enclosure large enough to have a reasonable natural resonance frequency of the sealed box around 40Hz without adding mass. 40Hz is a frequency where still a good portion of bass signal can be found in normal music. In addition I am adding a LT to this natural IRPS system which is changing the resulting roll off frequency down to 12Hz.
All together a trade off between size, required power, possible SPL.
Up to now I enjoy this trade off quite much.
I must agree to statement that micronizing the box to exzessive small values and correcting with brute force electronics does sound less impressive than larger sized designs. I definitely hear this by comparing my new design (Two cubes 40cmx40cmx40cm, with one Adire Sadhara driver each) to my older supersmall 12liter/11driver system. The new system offers much more relaxed and impressive bass reproduction also at low signal levels, where the distorsions should low in both systems. Up to now I did not fully analyze the reasons for this difference, but at least my ears tell me very clear...
For my current situation the two IRPS Sadhara cubes with LT are doing a good job and in my living two 12" drivers with+/-27mm lin Xmax are mostly sufficient. I don't want it giant.
Hm,... 8Hz you say... for pipes... I could change the LT to go deeper. But my reason NOT to go so deep was the fact that our human body is sensitive to infrasound. From what I know the most dangerous frequency range of infrasonic weapons is somewhere between 5Hz....10Hz. How can our pope support such things in the church???!!!
In reply to Bogie, you do not need many speakers, and you do not need a lot of equalization to get this sytem to work. I have heard IB systems with only one high Xmax 15" sound very loud very clean, and get down to 16 Hz. I only use one notch filter at 31.5 Hz with 5 dB of cut to get rid of a room resonance. See Audio Magazine December 1999 issue article by Tom Nousaine for review of four IB subwoofers with complete measurements of output, frequency response and distortion. At normal listening levels, my installation has a reasonably flat response down to 6 Hz, with a system impedance peak at 12 Hz, demonstrating a considerable response one octave below fs.
Hi Rek,
great that you are pointing out the room resonances.
Most normal living rooms have their three main modes all in the range of normal music bass.
Even large rooms do not bring down the modes low enough to ignore.
I.e. floor-ceiling 3m: 57Hz
length 7.3m: 23Hz, urghs also double is within normal bass range
width 7.1m: 24Hz, + double
Altogether causing fat booming between 40Hz..60Hz, if the subs are in the wrong position. Up to now I am fighting the resonances just by finding proper positions for the subs.
But I am thinking about three notch filters (some simple OP amp solutions), which can be adjusted to the room. I do not have pratical experience with them. Normal room resonances have a Q between 5 up to 15. Do you think Q should be adjustable, or is it sufficient to settle it to Q=10 and just have the frequency + attentuation adjustable?
Bye Markus
great that you are pointing out the room resonances.
Most normal living rooms have their three main modes all in the range of normal music bass.
Even large rooms do not bring down the modes low enough to ignore.
I.e. floor-ceiling 3m: 57Hz
length 7.3m: 23Hz, urghs also double is within normal bass range
width 7.1m: 24Hz, + double
Altogether causing fat booming between 40Hz..60Hz, if the subs are in the wrong position. Up to now I am fighting the resonances just by finding proper positions for the subs.
But I am thinking about three notch filters (some simple OP amp solutions), which can be adjusted to the room. I do not have pratical experience with them. Normal room resonances have a Q between 5 up to 15. Do you think Q should be adjustable, or is it sufficient to settle it to Q=10 and just have the frequency + attentuation adjustable?
Bye Markus
ChocoHolic said:
The problem in "curing" room response that way is that you'll end up with one hotspot where it sounds good. Move away from that spot and you'll wonder where all the bass went. Better to treat the acoustics of the room with bass traps (which might just mean strategically placed bookshelves, sofas etc.)
...right.
Of course, I also play around with arranging the furniture. But possibilties are limited.... also with combination of acceptable furniture set up and sub position I am not always happy. Furniture usually is of limited use, because wave length are much larger than furniture size. You mostly reduce Q, which is already helpful, but does not avoid the room modes. I would need to install two non parallel pseudo walls and hang down the ceiling in a strange way....
That's why I am thinking about filtering the signal.
From my understanding if the room resonances are filtered out, then most of these hot spot effects should be reduced to a low level. Simply because there is almost no signal anymore, which which could power into this resonances. Or does your experience show opposite results?
Of course, I also play around with arranging the furniture. But possibilties are limited.... also with combination of acceptable furniture set up and sub position I am not always happy. Furniture usually is of limited use, because wave length are much larger than furniture size. You mostly reduce Q, which is already helpful, but does not avoid the room modes. I would need to install two non parallel pseudo walls and hang down the ceiling in a strange way....
That's why I am thinking about filtering the signal.
From my understanding if the room resonances are filtered out, then most of these hot spot effects should be reduced to a low level. Simply because there is almost no signal anymore, which which could power into this resonances. Or does your experience show opposite results?
Markus, I have been considering an approach much the same as yours, after coming to conclusions much the same as yours. My lounge has to be lived in and I suspect that the floor-ceiling resonance is my main problem as I get a boom about 50Hz. I don't really care too much if using an EQ to tame the peak makes the sound worse elsewhere, as usually the sofa is where people sit.
As the resonance is a vertical one it seems equally bad throughout the room, so I think a bit of EQ should actually sound better for everyone anyway.
As the resonance is a vertical one it seems equally bad throughout the room, so I think a bit of EQ should actually sound better for everyone anyway.
In reply to Chocoholic, I was lucky that the room resonance that I encountered at my listening position was right at a slider frequency of my 1/3 octave equalizer, 31,5Hz. Careful measurements should be made with a sine wave generator and a microphone that has reliable frequency response at frequencies below 100 Hertz. Many high quality microphones used for recording have intentional low frequency roll offs that can taint the measurements. The Panasonic capsule that Linkwitz uses is really pretty good in that range, and does go down to the 5 to 6 Hz region. I would recommend a parametric equalizer for subwoofer channel so that the room effects can be worked out. I had to modify my 1/3 octave equalizer since it has a 10Hz to 250Hz variable two pole high pass filter that cannot be bypassed. I added capacitors in parallel with the high pass section filters to make it 1Hz to 25Hz instead. I currently have it set at 4Hz. The other thing that you need to know is the actual low frequency roll off of the power amplifier you are using for the subwoofer channel. I had to select among my power amplifiers to get the one with the lowest –3dB frequency. The one I use measures 3dB down at about 1 Hz. This would be an ideal application of the venerable Crown DC300 amplifier that was flat down to DC! There are a few parametric equalizers out there like http://www.rane.com/pe17.html that have center frequency capability at 10Hz.
...back from holidays...
Yes REK, such parametric EQ is looking pretty much what I want to try. I think for my purpose three bandpass/notch filters should be OK.
Up to now I plan to set up that stuff on my own.
BTW: Why is the low cut of the power amp relevant?
As long as it is much lower than the room modes it should not affect the booming. (I think..)
In my case the power section will be a DC coupled class D amp.
Depending on the application, I plan to have a 6th order 1db ripple tschebbychef around 8 or 9 Hz - cutting off the subsonics. And additional may be a 'party power filter', = 2nd order butterworth below 40Hz, in order to allow high SPL without exceeding Xmax of the speakers.
Yes REK, such parametric EQ is looking pretty much what I want to try. I think for my purpose three bandpass/notch filters should be OK.
Up to now I plan to set up that stuff on my own.
BTW: Why is the low cut of the power amp relevant?
As long as it is much lower than the room modes it should not affect the booming. (I think..)
In my case the power section will be a DC coupled class D amp.
Depending on the application, I plan to have a 6th order 1db ripple tschebbychef around 8 or 9 Hz - cutting off the subsonics. And additional may be a 'party power filter', = 2nd order butterworth below 40Hz, in order to allow high SPL without exceeding Xmax of the speakers.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.