There was a mini consensus of “established” thinkers in a local audio club forum, that it’s better to have lower sensitivity drivers and a high powered amp, than vice versa. The idea being higher sensitivity drivers have more ragged response.
I think this is often true, eg a club member has a 95(?) db Focal 6.5 inch that has a pretty unpleasant “shouty” quality, apparently a peak around 1 kHz.
I know from a friend’s experiments that the 100 dB PHL 3451 has a lot of peaks starting around 2.2 kHz, so many that they were not possible to tame with notch filters. I believe the 96 dB Fostex 166 full ranger also has a number of peaks 2- 3 kHz.
I think this is a handy generalization
but what are the exceptions?
Eg, I believe the 96-100 dB Aurum Cantus (once ESg) ribbons – yes not cones, and tweeters, are exceptions.
What mid drivers can cover say 100 – say 3000, 95 or more dB, without unwanted peaks (or dips or distortion)?
I think this is often true, eg a club member has a 95(?) db Focal 6.5 inch that has a pretty unpleasant “shouty” quality, apparently a peak around 1 kHz.
I know from a friend’s experiments that the 100 dB PHL 3451 has a lot of peaks starting around 2.2 kHz, so many that they were not possible to tame with notch filters. I believe the 96 dB Fostex 166 full ranger also has a number of peaks 2- 3 kHz.
I think this is a handy generalization
but what are the exceptions?Eg, I believe the 96-100 dB Aurum Cantus (once ESg) ribbons – yes not cones, and tweeters, are exceptions.
What mid drivers can cover say 100 – say 3000, 95 or more dB, without unwanted peaks (or dips or distortion)?
Would a wider range than 100 to 3K be ok? The Eminence Beta 8 drivers are very nice. Fast, clean, transparent and 95dB. They do have a rising response above 1K but that is easily flattened with a .4mh choke in series with the driver. The Beta 8 drivers have good response out to 4.5Khz then roll off nicely.
I’m currently using the Beta 8’s in a U-Baffle system. There are 2 drivers per side with one running full range with the .4mh series choke and the second driver running bass reinforcement with a 10mh series choke. The Beta 8’s are covering the 60Hz to 6Khz range. The 6Khz+ range is covered by Hi-Vi Research RT-8 tweeters.
The 20Hz to 60Hz bass duties are covered by a pair of OB subs each having 2 Dayton IB-15 drivers in W frames. The added mass of the air load from the W frames lowers the Fs of the IB-15’s into the 16Hz range. My room supports 20Hz nicely from the OB subs, drops quickly in the 18Hz range but comes back up again in the 16Hz to 14Hz range.
Here is an image of the frequency response measured on one of the early OB test baffles. The dips at 900 and around 2K were caused by the baffle. These don't show up in the current U-Baffle version. There is no crossover other than the .4mh choke. The Beta 8 has a nice natural roll off without out of band peaks.
Beta 8 Frequency response
Here is a current picture of the OB bass, U-Baffle midbass-mids system.
Current system
Gary
I’m currently using the Beta 8’s in a U-Baffle system. There are 2 drivers per side with one running full range with the .4mh series choke and the second driver running bass reinforcement with a 10mh series choke. The Beta 8’s are covering the 60Hz to 6Khz range. The 6Khz+ range is covered by Hi-Vi Research RT-8 tweeters.
The 20Hz to 60Hz bass duties are covered by a pair of OB subs each having 2 Dayton IB-15 drivers in W frames. The added mass of the air load from the W frames lowers the Fs of the IB-15’s into the 16Hz range. My room supports 20Hz nicely from the OB subs, drops quickly in the 18Hz range but comes back up again in the 16Hz to 14Hz range.
Here is an image of the frequency response measured on one of the early OB test baffles. The dips at 900 and around 2K were caused by the baffle. These don't show up in the current U-Baffle version. There is no crossover other than the .4mh choke. The Beta 8 has a nice natural roll off without out of band peaks.
Beta 8 Frequency response
Here is a current picture of the OB bass, U-Baffle midbass-mids system.
Current system
Gary
rick57 said:
List price is 122 Euro. I`ve seen a kit where it is used BLH, but no reports. The Supravox GMF drivers all have holes.
http://www.supravox.fr
A look at the manufacturers’ frequency response graphs shows the Focal (6WM) peaky around 1 kHz, and the Beta 8 peaky around 3 kHz.
Beta 8 Focal
100 93 92
1000 94 97
3000 98 93
I agree the Beta 8 seems much better value than the (now discontinued) 6WM.
Your chokes seems to have worked very well. Though I guess they’re not the answer for a middle of passband problem like the Focal.
Your U-Baffle system is cool!
What is (approximate) system efficiency?
If I understand chokes' operation correctly, in shaving off the rising response or peaks, they lower sensitivity to the lowest level in the passband. If the lower XO is 100, the Beta 8 becomes about 93 db, the Focal 92 dB; while the Ciare (which per manufacturers’ graph is > 100 dB at 2.2 kHz) would be more like 95 dB. (And 122 Euro seems quite reasonable)
Cheers
Beta 8 Focal
100 93 92
1000 94 97
3000 98 93
I agree the Beta 8 seems much better value than the (now discontinued) 6WM.
Your chokes seems to have worked very well. Though I guess they’re not the answer for a middle of passband problem like the Focal.
Your U-Baffle system is cool!
What is (approximate) system efficiency? If I understand chokes' operation correctly, in shaving off the rising response or peaks, they lower sensitivity to the lowest level in the passband. If the lower XO is 100, the Beta 8 becomes about 93 db, the Focal 92 dB; while the Ciare (which per manufacturers’ graph is > 100 dB at 2.2 kHz) would be more like 95 dB. (And 122 Euro seems quite reasonable)
Cheers
You are correct. The series choke works well with a simple rising response. EQing out a hump in the middle of the response is more difficult and has a larger sonic penalty.
The system sensitivity seems to be in the 95 to 96dB range as it matches up with the Hi-Vi RT-8 tweeter with no padding needed. depending on who measured it the RT-8 is rated between 95 (Madisound) and 96 (Partsexpress).
The dual driver system works out well as you can come up with a compromise that keeps the sensitivity up and helps compensate for the roll off of the U-baffle. The baffle calculates out to ~140Hz but with the 2 drivers the system response is good down to the 60Hz resonance of the Beta 8's. Varying the 10mH choke on the booster driver is a very effective tweak for tuning the bass in the 60Hz to 200Hz range.
The U-baffle as used needs no EQ added before the amplifier.
I've been doing some calculations using the spreadsheet from Linkwitz labs. Looks like converting the OB sub to a U-baffle by adding a 24" extension on the back side effectivly increases the sensitivity by 10dB. I'm going to give it a try tomorow. This will make the 60 watts from the 6550 sub amp go alot further. It will be interesting to hear what the U-baffle sub sounds like compared to the OB sub.
Gary
Gary P said:
Really.. well I guess great minds think a-like.
Check out the "customer systems" and "reference systems" links to the left of the page:
http://www.audio-consulting.ch/
While certainly not the same - there is one attribute that is distinctive that you have also incorporated into your design. (..of course I could have said the same for SL/Phoenix sub - but that was too easy.)
Gary P said:
The best way to acomplish this (the U-baffle) is loading the driver in a tube (like a sonotube) and stuffing it with drinking straws (..or better yet the smaller diameter pro-flow low air resistance tubing that will extend the length of the tube). (..perhaps with a box surrounding the tube and sand filled between the tube and the box.)
BTW, do you have anything to compensate for the reduction in off-axis response (besides loudspeaker "toe-in") as the freq. response rises? (..i.e. that hi-vi tweeter has a rather depressed response off-axis.)
Gary P said:
Gary,
I think you need to double check your formulae. It looks like your U baffle is 20" deep, which is a +/- 60hz Fequal not 160hz. Also keep in mind your current W for the 15's is dipole of what looks to be a 16-17" D. Every inch you add on to the back as a U-baffle is 2" of added D. That's why U baffles get 6db more at the bottom, because D is doubled which lowers everything by an octave. 24" added to the back is adding 48" to "D". It looks to me like 2 full octaves or 12db, so your 10db seems pretty close in terms of Fequal, but your max should be significantly more. You will definitely need bracing and some fiber fill back there. Going from dipole to cardoid response isn't a bad thing, but if you have a solid brick or concrete wall behind you, your listening position can't be near it or the bass response can vary drastically with small changes in distance.
I've got a pair of four 12" driver W/U baffles with the W all the way to the front with the U portion giving me a total depth of 55cm, but a "D" of over 80cm. I'd recommend clamping a 24" extension on before making it permanent. Also, that 1/2" plywood isn't going to cut it. With a pair of IB 15"s per side, you should have bass to spare. Ignoring the baffle, how high do you think they can run and still sound good in OB ? I missed the sale PE had a while back, but won't miss the next one.
It's great to see someone else taking the minimalist approach to OB using U-baffles and not cheating by boxing the bass. Keep experimenting.
454Casull said:
as wide (horizontally) as possible. Granted it likely will not be omni (horizontally).. only a few of those available. Still that particular driver has quite a bit of spl loss off-axis - and not that far off-axis either. Addtitionally it starts relativly low in freq. (due to the width of the radiating area).:
http://www.partsexpress.com/pdf/297-413.pdf
not only will the averaged response likely be depressed higher in freq. at the listening position, but image position will also be altered and reduce apparent stereo seperation.. (..If indeed the manufacturer's graphs are fairly representetive.)
Gary
Was John K’s NaO www.geocities.com/kreskovs/John1.html (which Linkwitz disagrees with the theory of) a help with the design?
Regards
Richard
(BTW, I just bought 2nd had a pair of Eminence Kappa 12s).
Was John K’s NaO www.geocities.com/kreskovs/John1.html (which Linkwitz disagrees with the theory of) a help with the design?
Regards
Richard
(BTW, I just bought 2nd had a pair of Eminence Kappa 12s).
I've read through his stuff. Nicely presented and usefull concepts. Not sure I agree with all his points though.
Your probably right on the formulas. Have to do further checking when the brain is engaged...
Good eyes. The W is 16" front to back. "D" is 25" when you include the distance to the center port on the rear. Path over the top is blocked by the mid unit.
Yesterday I added a 16" extension to the rear just for grins. No need to worry about the cosmetics of a few new screw holes on the rear of the W as this is pure prototype anyway. I did several baseline measurments before adding the extension, then did the same measurments with the extension in place. Interesting results. Not what I was expecting but usefull none the less. I'll have to add a page to my webpage to present all the data.
The first measurment was made with the extension on but no damping. Efficiency increase was ~3dB, not as much as I was expecting for 16". The frequency response was very similar with just minor changes.
Both the W and U baffles showed a strange LF response with a peak at 13.5Hz, a dip at 16.5Hz, followed by the main response starting at 22Hz. The peak at 13.5 was the same ougput as the 22Hz and above response, the dip at 16.5 was 15dB deep. As I added layers of damping (3.5" thick cotton building insulation) inside the extension the response on the bottom smoothed out. With 3 layers of insulation the peak at 13.5Hz had dropped 4dB and the bottom of the dip had risen 10dB. What looked quite ugly before the damping was now quite nice. The response of the subs goes from the crossover region at 60Hz down to 13Hz (with the exception of a room peak at 50Hz) while fitting in 5dB peak to peak window. The peak at 50Hz has been removed by adding another notch filter in the DCX.
Yep, once I heard boxless bass all the way down I was hooked. Using a sealed sub for the very botton did not appeal to me at all...
I don't compensate for the off axis response. I don't feel that better imaging comes with wider horizontal dispersion. I prefer speakers with a narrower dispersion angles to cut down on the clutter of reflections from the side walls. Reflection control is what led me away from true open baffle to the damped U-baffle. To my ears it seemed that the reflections off the rear were limiting the depth of the perceived image to the rear wall. With the damped U-baffle the image extends way beyond the rear wall of my living room. My thoughts on this is the reflected sound from the rear of the OB mids was louder than the embeded clues of depth from the main sound from the speakers. I'm much happier with less room reflections, and the narow dispersion of the Beta 8's at HF and the RT-8 tweeters supports this.
I took a look at the http://www.audio-consulting.ch/
webpage. I see what you mean
Gary
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.