Whenever i've considered something like the Clairaudient, my thots always come around to using a big PVC pipe cut in half (it helps to have 6 & 10" pipe lying around unused) tacked onto a front & rear baffle.
dave
Notch filter
Bill,
I beleive this notch filter will have to be in series with BSC sircuit and not in parallel.
I've designed both circuits for single driver application, but not for an array. There is thread about A3 driver on Audio Circles, where I posted it. Look here: Have you heard the Audience A3 drivers?.
I've also created a small gallery there with measurements for a single-driver version. It's located here: VadimB's Audience A3
VadimB
Bill,
I beleive this notch filter will have to be in series with BSC sircuit and not in parallel.
I've designed both circuits for single driver application, but not for an array. There is thread about A3 driver on Audio Circles, where I posted it. Look here: Have you heard the Audience A3 drivers?.
I've also created a small gallery there with measurements for a single-driver version. It's located here: VadimB's Audience A3
VadimB
I thought it was rather nice. So did many other people. There didn't seem to be any lack of bass. IIRC each driver was in a separate chamber with its own port. And I think the tuning of each port was slightly different.
VadimB,
Why not?
A series-notch filter can be added as a parallel leg in that BSC circuit (essentially a series RLC across amp's + & - inputs).
The circuit load of the notch filter leg only acts in a narrow band well outside the BSC range of frequencies (notch only acts to cancel high-frequency, high Q, driver resonance).
Actually can place a notch either in a series or parallel leg - it depends whether the notch is a parallel-resonant, or a series-resonant type. Selection of a series or parallel notch would depend on what you want to accomplish with the overall compensation circuit, & the choice of associated component values (& interactions), available voltage gain levels, etc.
Placing a series-resonant R-L-C notch filter in the line network (at amp's + & -input) as a parallel leg, is analogous to placing a series-resonant notch filter across the driver's + & - terminals.
Bill
Why not?
A series-notch filter can be added as a parallel leg in that BSC circuit (essentially a series RLC across amp's + & - inputs).
The circuit load of the notch filter leg only acts in a narrow band well outside the BSC range of frequencies (notch only acts to cancel high-frequency, high Q, driver resonance).
Actually can place a notch either in a series or parallel leg - it depends whether the notch is a parallel-resonant, or a series-resonant type. Selection of a series or parallel notch would depend on what you want to accomplish with the overall compensation circuit, & the choice of associated component values (& interactions), available voltage gain levels, etc.
Placing a series-resonant R-L-C notch filter in the line network (at amp's + & -input) as a parallel leg, is analogous to placing a series-resonant notch filter across the driver's + & - terminals.
Bill
notch filter
Bill, you are right and I am wrong - I stay corrected. What happend is that I was in a hurry and did not look the reference in the original post, so did not realize it was about a line-level circuit. I was under impression you were talking about speaker-level design.
I have used the parallel resonance traps for the tweeter crossovers (outside of the passband), but have not thought about using this as a notch. Have you ever tried that? What are the benefits and drawbacks of one vs. the other?
Thanks, VadimB
Bill, you are right and I am wrong - I stay corrected. What happend is that I was in a hurry and did not look the reference in the original post, so did not realize it was about a line-level circuit. I was under impression you were talking about speaker-level design.
I have used the parallel resonance traps for the tweeter crossovers (outside of the passband), but have not thought about using this as a notch. Have you ever tried that? What are the benefits and drawbacks of one vs. the other?
Thanks, VadimB
Notch filters/traps
VadimB,
So far, I've only experimented with series-resonance traps for passive mid-woofer crossovers (connect an R-L-C leg across the mid-woofer driver's terminals). Usually, it's the mid-woofer that presents the high-frequency resonant peak(s) that must be suppressed by the passive crossover.
IMO, a very good website to see several examples of incorporating series, or parallel, resonant notch filters in passive speaker crossovers is Zaph Audio ("trap" is just another word for a notch filter, generally speaking).
Vance Dickason's "Loudspeaker Design Cookbook" covers notch filters too.:
Bill
VadimB,
So far, I've only experimented with series-resonance traps for passive mid-woofer crossovers (connect an R-L-C leg across the mid-woofer driver's terminals). Usually, it's the mid-woofer that presents the high-frequency resonant peak(s) that must be suppressed by the passive crossover.
IMO, a very good website to see several examples of incorporating series, or parallel, resonant notch filters in passive speaker crossovers is Zaph Audio ("trap" is just another word for a notch filter, generally speaking).
Vance Dickason's "Loudspeaker Design Cookbook" covers notch filters too.:
Bill
To clarify my post above:
By "mid-woofer" crossover, I meant a passive crossover for a typical two-way speaker. One driver in the two-way is the "mid-woofer", and the other driver is the tweeter.
The mid-woofer usually is the culprit that gives the most trouble needing the notch filter (trap) attention.
I wish I could be a better writer for posts like this.
Bill
By "mid-woofer" crossover, I meant a passive crossover for a typical two-way speaker. One driver in the two-way is the "mid-woofer", and the other driver is the tweeter.
The mid-woofer usually is the culprit that gives the most trouble needing the notch filter (trap) attention.
I wish I could be a better writer for posts like this.
Bill
Audience now has a dual voice coil version of the A3 to allow the user to select 4,8, 0r 16 ohms. I will be testing them in a QWTL this spring.
Audience A3 information
I am a distributor for Audience's A3 drivers and we have made some design recommendations of late.
We've added the 2+2, 4+4, 8+8 & 16+16 driver configuration in our new ClairAudient loudspeakers to elimate the BSC problems.
When building your A3 cabinet, use no more than 4 drivers per divided cabinet within the cabinet ie partitions. If using the front and rear drivers for BSC like a square box design, be sure to separate the rear drivers from the front with a simple divider for delay. Still wire them for your 4 or 8 ohm load.
We are not using a ported enclosure anymore. The drivers unload too quickly at the lower frequencies and tend to reach xmax quicker and lower the dB output.
Regarding standing waves, the reason we use only four drivers per cabinet is to eliminate them. You need to add some real wool fill for dampening as well.
We're hoping to have our new drivers in by mid month. We have an order coming in then, and hopefully they will be with the newer voice coils.
I made recommendation to our engineers that we try the passive radiator design because of the low Q. We got one finished just in time for the Las Vegas SHOW and displayed it there. Here is a link to the photo. Mr. McDonald is proud of the new prototype:
Daily Coverage - January 10 (Part 2)
I am a distributor for Audience's A3 drivers and we have made some design recommendations of late.
We've added the 2+2, 4+4, 8+8 & 16+16 driver configuration in our new ClairAudient loudspeakers to elimate the BSC problems.
When building your A3 cabinet, use no more than 4 drivers per divided cabinet within the cabinet ie partitions. If using the front and rear drivers for BSC like a square box design, be sure to separate the rear drivers from the front with a simple divider for delay. Still wire them for your 4 or 8 ohm load.
We are not using a ported enclosure anymore. The drivers unload too quickly at the lower frequencies and tend to reach xmax quicker and lower the dB output.
Regarding standing waves, the reason we use only four drivers per cabinet is to eliminate them. You need to add some real wool fill for dampening as well.
We're hoping to have our new drivers in by mid month. We have an order coming in then, and hopefully they will be with the newer voice coils.
I made recommendation to our engineers that we try the passive radiator design because of the low Q. We got one finished just in time for the Las Vegas SHOW and displayed it there. Here is a link to the photo. Mr. McDonald is proud of the new prototype:
Daily Coverage - January 10 (Part 2)
Last edited:
He is correct about trying to use the A3 in a ported enclosure. They will bottom out easily, yet to their credit, mine did not suffer any mechanical damage. They are very well engineered driver.
Ken,
Thanks for the nice email you sent to Audience. We always want to hear from the DIYer about their builds. These ares such extremely fine drivers and their uses and design are almost endless.
Mike
I've build a ported box for a single driver - 3.0 l, tuned to about 60 Hz. A second order in-line hi-pass filter at 45 Hz, installed in front of the power amp keeps the driver from exceeding Xmax at up to 20W. The -3dB point is at 57 Hz and -6dB is at 50Hz. Not bad at all!
For 4 drivers use 4 times the volume with the same tunig and the same in-line filter.
VadimB
For 4 drivers use 4 times the volume with the same tunig and the same in-line filter.
VadimB
Very clever indeed Vadim!
I built a four driver ported and it sounded great, but I was cautious to not overdrive them. I sold them as I didn't use or demo them. I needed the build experience and had to learn about their performance. Quite the speaker for only 3"!! Good job, and smart too....
I avoid filters if possible, but that is one way to reduce the unloading and maintain a more efficient performance.
WHY NOT APERIODIC for A3 array - like 'variovent' for cabinets?
An alternative might be to use aperiodic loading for an array of A3 drivers, instead of vented or a passive radiator.
Can get a "variovent" (e.g. Dynaudio or Seas - avail at Parts Express) to get aperiodic loading which reduces low freq impedance rise compared to sealed cabinet. However, I would prefer a DIY implementation (not a Variovent per se) to get the aperiodic bass tuning.
For a small "2+2" A3 array configuration you might try the SEAS Prestige SP18R 6.5" Passive Radiator which allows you to add mass for adjusting the Helmholtz resonant frequency of the passive radiator.
Bill
An alternative might be to use aperiodic loading for an array of A3 drivers, instead of vented or a passive radiator.
Can get a "variovent" (e.g. Dynaudio or Seas - avail at Parts Express) to get aperiodic loading which reduces low freq impedance rise compared to sealed cabinet. However, I would prefer a DIY implementation (not a Variovent per se) to get the aperiodic bass tuning.
For a small "2+2" A3 array configuration you might try the SEAS Prestige SP18R 6.5" Passive Radiator which allows you to add mass for adjusting the Helmholtz resonant frequency of the passive radiator.
Bill
I built a few of these back in the 60s and they worked well. They are easier to tune than trying to stuff a QWTL. There are a number of materials that can be experimented
with. I used different densities of what we used to call foram rubber in layers.
with. I used different densities of what we used to call foram rubber in layers.
Bill,
The aperiodic loading certainly can be used. The question, however, is: what will that acheive. The driver has relatively low Q=.4. In 3.4 liter box it results in my favourite Q=.5 tuning with -3 dB point of about 140-145 Hz. Yuo may also opt for Q=.7 in a 1 liter box with -3 dB point of about 135 Hz. Using variovent will allow you to acheive the same in "slightly" smaller box, but in this case it will not allow to acheive lower -3 point. At least not significantly. IMO the box is small already, so why bother?
Ported or passive, on the other hand, produces significant benefits in this case, so much that using these driver alone (no sub) becomes a possibility.
Don't get me wrong, my main speakers have closed alignment for the woofers...
Vadim
The aperiodic loading certainly can be used. The question, however, is: what will that acheive. The driver has relatively low Q=.4. In 3.4 liter box it results in my favourite Q=.5 tuning with -3 dB point of about 140-145 Hz. Yuo may also opt for Q=.7 in a 1 liter box with -3 dB point of about 135 Hz. Using variovent will allow you to acheive the same in "slightly" smaller box, but in this case it will not allow to acheive lower -3 point. At least not significantly. IMO the box is small already, so why bother?
Ported or passive, on the other hand, produces significant benefits in this case, so much that using these driver alone (no sub) becomes a possibility.
Don't get me wrong, my main speakers have closed alignment for the woofers...
Vadim
The advantage of an aperiodic enclosure is that it is lower pressure, and can be used to flatten the impedance peak at resonance.
dave
dave
Update on application for Audience A3 drivers: DIY aperiodic '4+4' configuration
Update on what I'm doing with Audience A3 drivers -- DIY '4+4' style.
Building two pair of '4+4' configurations, A3 drivers mounted in inexpensive Dayton rectangular 1/2 cu ft cabinets, for 1. bedroom TV system, & 2. for computer monitor system. Both small systems with subwoofers for below say, 70Hz.
Originally considered saving money by building '2+2' speakers (use half as many A3 drivers) in Dayton 1/2 cu ft rectangular cabinets & with 6.5" passive radiator(s), like the picture of the Audience Clairaudient prototypes at the 2010 Show(?).
IMO, the sonically preferred '4+4' configuration offers a boost in acoustic radiation efficiency and power handling/lower distortion -- and also, the bass output might be higher quality (passive radiators, and vented alignments, shift phase, etc.). Rear-mounted array of drivers to help with BSC.
Plan is to try a very-different DIY quasi-aperiodic loading in my DIY '4+4' Dayton cabinets -- idea is to reduce low-frequency speaker impedance rise compared to typical sealed-cabinet loading of drivers.
Might also use a simple RC circuit at input to power amps for freq resp contouring (also, might add an RLC leg in passive ckt before the amp input for 8 kHz resonance suppression) -- this will be a custom arrangement compatible with my 'IcePower' based power amps.
IMO, IcePower, a 'class D' technology, sounds great with the A3s -- incidentally, Audience's class D amp reportedly sounded better than another amp in the recent March Stereophile-sponsored Florida Show.
BTW, I understand Audience Inc. has recently discontinued their original 16-ohm VC version of the A3, and in the future will be selling their dual-voice-coil version of the A3.
Reportedly, none of the A3 parameters will change when both VCs are series-connected to get 16 ohm impedance (I hope). So, will be adding ("mix 'n match") some of the new dual VC drivers in with my older 16-ohm A3s for DIY '4+4' speakers.
Bill
Update on what I'm doing with Audience A3 drivers -- DIY '4+4' style.
Building two pair of '4+4' configurations, A3 drivers mounted in inexpensive Dayton rectangular 1/2 cu ft cabinets, for 1. bedroom TV system, & 2. for computer monitor system. Both small systems with subwoofers for below say, 70Hz.
Originally considered saving money by building '2+2' speakers (use half as many A3 drivers) in Dayton 1/2 cu ft rectangular cabinets & with 6.5" passive radiator(s), like the picture of the Audience Clairaudient prototypes at the 2010 Show(?).
IMO, the sonically preferred '4+4' configuration offers a boost in acoustic radiation efficiency and power handling/lower distortion -- and also, the bass output might be higher quality (passive radiators, and vented alignments, shift phase, etc.). Rear-mounted array of drivers to help with BSC.
Plan is to try a very-different DIY quasi-aperiodic loading in my DIY '4+4' Dayton cabinets -- idea is to reduce low-frequency speaker impedance rise compared to typical sealed-cabinet loading of drivers.
Might also use a simple RC circuit at input to power amps for freq resp contouring (also, might add an RLC leg in passive ckt before the amp input for 8 kHz resonance suppression) -- this will be a custom arrangement compatible with my 'IcePower' based power amps.
IMO, IcePower, a 'class D' technology, sounds great with the A3s -- incidentally, Audience's class D amp reportedly sounded better than another amp in the recent March Stereophile-sponsored Florida Show.
BTW, I understand Audience Inc. has recently discontinued their original 16-ohm VC version of the A3, and in the future will be selling their dual-voice-coil version of the A3.
Reportedly, none of the A3 parameters will change when both VCs are series-connected to get 16 ohm impedance (I hope). So, will be adding ("mix 'n match") some of the new dual VC drivers in with my older 16-ohm A3s for DIY '4+4' speakers.
Bill
Yes, the A3 DVC model is now available. All the other specs are the same. I am in the process of evaluating a pair for a future project where they will be used as wooofers in a two way system.
- Status
- Not open for further replies.