I've been looking at an article on qb5 alignments, http://sound.westhost.com/qb5align.htm (excellent site, IMHO)
I'm wondering about applying that to a subwoofer. The basic idea is that you tune the box so there's a hump down low in frequency response, them use eq to smooth it down to the
response you actually want. All the eq is cut so you are reducing some risks which are increased by schemes that use boost (Linkwitz transform & most other bass eq); exceeding either the amp's ability to deliver power or the speaker's excursion limits. Sounds useful.
Looking at tables in the article, it appears there are a couple of "magic" values of Qt where F3<=Fs, around Qt=.32 in Group 1 and Qt=.445 in Group 2.
Parts Express have woofer a selection guide:
http://www.partsexpress.com/resources/indexes/sb_guide.html
Click on any parameter heading to sort by it -- most people would likely look at "sealed F3" and "sealed volume" or the corresponding "vented" numbers.
I sorted on Qts and looked for .32 or so with a low Fs. Lots of possibilities, including some smaller drivers that might work for folk who need a compact sub. I'm not much worried about size, so the one I like is a Pioneer 15 inch
Qts=.32 so F3<=Fs=24 Hz, efficiency is pretty good at 96 Db 1W/1m, and it is only $50.
Box size comes out at around 8 cubic feet, not ideal but not horrible for a 15" woofer. A 2-foot cube or a more-or-less reasonable length of sonotube.
So this isn't a compact sub and any vented vented system gives somewhat poorer transient response than a correctly designed sealed box. I can live with those, unless the transient response really sucks.
Other than that, it looks amazingly good. 24 Hz F3 and 96 Db efficiency with a $50 woofer and dead simple eq. Wow!
What am I missing here? Are there problems I have not noticed?
Has anyone tried something like this? How did it work?
I'm wondering about applying that to a subwoofer. The basic idea is that you tune the box so there's a hump down low in frequency response, them use eq to smooth it down to the
response you actually want. All the eq is cut so you are reducing some risks which are increased by schemes that use boost (Linkwitz transform & most other bass eq); exceeding either the amp's ability to deliver power or the speaker's excursion limits. Sounds useful.
Looking at tables in the article, it appears there are a couple of "magic" values of Qt where F3<=Fs, around Qt=.32 in Group 1 and Qt=.445 in Group 2.
Parts Express have woofer a selection guide:
http://www.partsexpress.com/resources/indexes/sb_guide.html
Click on any parameter heading to sort by it -- most people would likely look at "sealed F3" and "sealed volume" or the corresponding "vented" numbers.
I sorted on Qts and looked for .32 or so with a low Fs. Lots of possibilities, including some smaller drivers that might work for folk who need a compact sub. I'm not much worried about size, so the one I like is a Pioneer 15 inch
Qts=.32 so F3<=Fs=24 Hz, efficiency is pretty good at 96 Db 1W/1m, and it is only $50.
Box size comes out at around 8 cubic feet, not ideal but not horrible for a 15" woofer. A 2-foot cube or a more-or-less reasonable length of sonotube.
So this isn't a compact sub and any vented vented system gives somewhat poorer transient response than a correctly designed sealed box. I can live with those, unless the transient response really sucks.
Other than that, it looks amazingly good. 24 Hz F3 and 96 Db efficiency with a $50 woofer and dead simple eq. Wow!
What am I missing here? Are there problems I have not noticed?
Has anyone tried something like this? How did it work?
Kick,
I am looking into QB5 alignments for a threeway big experimental Onken style cabinet. I am looking for the upper and lower limits of the test enclosure volume. There is some information on QB5 here. Satellites and Subwoofers
I can simulate QB5 in Aj-horn with a series capacitor. But Aj-horn does not simulate the group delay of the QB5 alignments. Are there simulators that will show the group delay of QB5? So i can get a visual on what i am doing? Would be nice to tune the delay peak out of the problem freq. range / tune for lowest group delay.
Winisd on my computer is stuck on the driver editor, i cannot fill in a name for the driver...
If WinIsd would work properly, would Winisd do the trick? Or is there another way to simulate this? Hornresp? Maybe someone with a bit more math and Akabak knowledge can help?
I am looking into QB5 alignments for a threeway big experimental Onken style cabinet. I am looking for the upper and lower limits of the test enclosure volume. There is some information on QB5 here. Satellites and Subwoofers
I can simulate QB5 in Aj-horn with a series capacitor. But Aj-horn does not simulate the group delay of the QB5 alignments. Are there simulators that will show the group delay of QB5? So i can get a visual on what i am doing? Would be nice to tune the delay peak out of the problem freq. range / tune for lowest group delay.
Winisd on my computer is stuck on the driver editor, i cannot fill in a name for the driver...
If WinIsd would work properly, would Winisd do the trick? Or is there another way to simulate this? Hornresp? Maybe someone with a bit more math and Akabak knowledge can help?
Last edited:
I usually use the Excel spreadsheet from here: The Subwoofer DIY Page - Ported Systems
It has several filter options which you can use to implement the EQ.
It has several filter options which you can use to implement the EQ.
I just tried Winisd beta on an older computer, nope dead in the water... Even tried the Alpha version still left on the old computer no result 
. Thanks Don for the excel diysub.org suggestion, much appreciated!
Will try the alignment by trail and error, hope to make some sawdust today.
Bernd Timmermans from Hobby Hifi published formulas for QB5 vented alignment. His approach is the other way around compared to the soundwesthost article. Driver provides the Q peak, not the enclosure or an opamp / circuit. These formulas work for drivers with Qts between 0.5 and 0.7 and stiff suspension.
Cabinet volume Vas * 3
Fb = 0.6 *fs
Fb = -3dB
The series capacitor is heavily influenced by the inductance of the voice coil. The value must be found emperical. This is a ballpark starting point for the value.
200.000
C = _______
Fb * Z
Z is driver Re, value of the capacitor (C) is in uF. The formula for the portlenght (L) in mm. d = diameter port in mm.
2.340 * d *d
L = ________- 0,8 d
Fb * Fb * Vb
Apologies for the strange Fb * Fb notation, read to the power of two. Layout limitations.
Sadly he got an enormous amount of Flak for this idea and the alignment from the online German diy community. His findings where the same as Robert C.White's in the Soundwest hosts articles: improved vented speaker performance, lower group delay, extended headroom, lower excursion belowe Fb.
Simulations in AJ-horn suggest a large bandwidth of possible QB5 allignments.
. Thanks Don for the excel diysub.org suggestion, much appreciated! Will try the alignment by trail and error, hope to make some sawdust today.
Bernd Timmermans from Hobby Hifi published formulas for QB5 vented alignment. His approach is the other way around compared to the soundwesthost article. Driver provides the Q peak, not the enclosure or an opamp / circuit. These formulas work for drivers with Qts between 0.5 and 0.7 and stiff suspension.
Cabinet volume Vas * 3
Fb = 0.6 *fs
Fb = -3dB
The series capacitor is heavily influenced by the inductance of the voice coil. The value must be found emperical. This is a ballpark starting point for the value.
200.000
C = _______
Fb * Z
Z is driver Re, value of the capacitor (C) is in uF. The formula for the portlenght (L) in mm. d = diameter port in mm.
2.340 * d *d
L = ________- 0,8 d
Fb * Fb * Vb
Apologies for the strange Fb * Fb notation, read to the power of two. Layout limitations.
Sadly he got an enormous amount of Flak for this idea and the alignment from the online German diy community. His findings where the same as Robert C.White's in the Soundwest hosts articles: improved vented speaker performance, lower group delay, extended headroom, lower excursion belowe Fb.
Simulations in AJ-horn suggest a large bandwidth of possible QB5 allignments.
Last edited:
AJ-horn 4 & 6, v.6 on the old computer.
In theory you add an order to the alignment with QB5 compared with normal vented enclosures, so group delay goes up, in practice you lower the group delay peak in frequency. Would love to play around with different alignments and see what happens with group delay.
I am big fan of 3rd order closed box subs.
This is the alignment range for a Fane 12-200, Qts 0.56, vas 73,26 liter, fs 62.85 Hz cabinet volume 140 liter 3.600 uF cap (red) upto 200 liters (black) 1.300 uF cap.
In theory you add an order to the alignment with QB5 compared with normal vented enclosures, so group delay goes up, in practice you lower the group delay peak in frequency. Would love to play around with different alignments and see what happens with group delay.
I am big fan of 3rd order closed box subs.
This is the alignment range for a Fane 12-200, Qts 0.56, vas 73,26 liter, fs 62.85 Hz cabinet volume 140 liter 3.600 uF cap (red) upto 200 liters (black) 1.300 uF cap.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
(I've got 600uF poly caps for closed box) - seems like actively eq-ed 6th order reflex can sound good if tuned low enough - I've used the old EV Interface eq box which can be rescaled for other fb.
An externally hosted image should be here but it was not working when we last tested it.
That would almost work. Filter wizard lets you use an arbitrary cutoff point and slope, but won't calculate what that *would* be with the driver/box impedance and cap value. You need to export the hornresp response and impedance data, and put that into a program for designing crossovers. Filter wizard is really for helping you set up your DSP or active crossover.
Such assisted alignments work well, particularly if implemented via DSP. I won't suggest doing it via passive filters if designing for pro audio use.
For example, the Fb for my Blastoramas cabinets is 60 Hz, which is pretty low for a vented alignment based on the Eminence Beta 8 drivers. Typically I use them with a subwoofer, but for those occasions when I don't I enable the DSP in the amp to add 12dB of boost @60 Hz with a Q of 1 and a 48dB HP filter @55 Hz to address everything below Fb. Also, the boost is done via DEQ so it's decreased as the signal level increases so the driver is not driven much out of its linear excursion limits. The result sounds great, quite punchy, and at lower levels you wouldn't believe that you're listening to 8-inch drivers designed for pro audio midrange duty.
???
The attached screenprint shows the power response for the default record with a 100uF series capacitor (C1) inserted between the amplifier and the loudspeaker.
Attachments
Thanks! I can slide or adjust the uF value until QB5 works with passive components? Good to know, something to look forward to when learning Hornresp!
It's for an Fane 12-200 Onken QB5 idea. Small SE tube amp...no DSP, no intention for PA use. I can simulate the enclosure and cap, but have no insight in the groupdelay and that's just the one parameter that will make it a fridge size's failure or win.
So passive QB5, would like to see group delay. Suggestions welcome, if i am sure it works, i will even take the trouble to switch my computer to DOS if needed. Just remember an ancient TL simulation program in DOS which had the possibility to add a series cap... Can't remember if it showed group delay, pre internet/ or maybe just at the brink of internet. By Bullock III.
I don't know Jack about DSP, so this is highly speculative, doesn't DSP x-crossover add group delay to? just like any other x-over? So simple minded; if Brian's Blastorama works with such an amount of x-over "power" in such a small bandwidth, i would be save too?
It's for an Fane 12-200 Onken QB5 idea. Small SE tube amp...no DSP, no intention for PA use. I can simulate the enclosure and cap, but have no insight in the groupdelay and that's just the one parameter that will make it a fridge size's failure or win.
An externally hosted image should be here but it was not working when we last tested it.
So passive QB5, would like to see group delay. Suggestions welcome, if i am sure it works, i will even take the trouble to switch my computer to DOS if needed. Just remember an ancient TL simulation program in DOS which had the possibility to add a series cap... Can't remember if it showed group delay, pre internet/ or maybe just at the brink of internet. By Bullock III.
I don't know Jack about DSP, so this is highly speculative, doesn't DSP x-crossover add group delay to? just like any other x-over? So simple minded; if Brian's Blastorama works with such an amount of x-over "power" in such a small bandwidth, i would be save too?
Last edited:
Obviously I need a new version. Again. Built-in circuit simulator is nice to have for designing mid/tops.
Yes.
The AJHorn filter shown in Attachment 1 would be specified in the Hornresp Filter Wizard as shown in Attachment 2.
Attachments
I don't know how much GD the DSP adds, but I can tell you that the approach works quite well, much better IMO than trying to do it passively (mainly because with DSP you have much more control available, particularly at different drive levels).
True story: Yesterday, my 15-rear old daughter's class had their end-of-term class party. She was asked to provide the "music system". I asked her what she was planning to do. She said she was going to use her little Aiwa boombox (two 3" full-range drivers driven with about 10W). I asked her how many students were going to attend this party. She said it could be as much as 45 if one or two of the other classes joined in. I told her that perhaps it might be better to use the Blastoramas instead. Hauling the "big" sub that I typically use with the Blastoramas into her class would be a bit of overkill though, so I switched the amp (an iNuke 3000DSP) into one of the saved configs that used DSP as I previously outlined to extend the response of the Blastoramas, configured the amp to limit output to 200W, then locked its configuration. Finally, I added an adapter cable to the amp to allow my daughter or any of her friends to use their iPod or whatever to play music through it, and showed my daughter how to adjust the amp's gain and what the red "limiter" indicator meant and what she had to do if it was on continuously. I then set up the amp and speakers and left them to it.
When I came back a few hours later to pick her up, they'd basically converted the classroom into a club and were charging admission, LOL. It was a great success, to the point that other classes (who were having their separate parties) joined in and even the teachers got involved in the dancing. The class above her had been tasked to raise some money for a cause, hence the impromptu "club" and its corresponding cover charge.
I suspect I might get some flack from someone over the volume level however - maybe I should've set the output limiter to 50W or even 25W instead
Thanks for the responses. I will try the simulation in Hornresp. Lovely strory about the "Club"
I know it's "old fashioned" to use passive x-overs. For this cheap "vintage 3-way hi effiency design idea" tri amping with three stereo single ended amps and DSP would be silly. Depending on which way you look at things. Adding a large cap in series is imho not that complex compared to a full DSP setup...
I know it's "old fashioned" to use passive x-overs. For this cheap "vintage 3-way hi effiency design idea" tri amping with three stereo single ended amps and DSP would be silly. Depending on which way you look at things. Adding a large cap in series is imho not that complex compared to a full DSP setup...
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.