God job earl! Is the design from the beginning passive? and how did you get all channels out from a 2x4 dsp?
If fully active do one use high efficiency resistors to get the ohm's right? I mean if one uses 4ohm drivers it may be a problem for the amps and speakers do not do well in series.
Cheers!
If fully active do one use high efficiency resistors to get the ohm's right? I mean if one uses 4ohm drivers it may be a problem for the amps and speakers do not do well in series.
Cheers!
retearl sounds like you have had some success. I am on a bit of a budget as well so please keep posting your updates. Thats very interesting that you used those little LM3886 amps to run your speakers. I was going to go something bigger but I will get the design sorted first, I have help from another member on here and I am so very grateful for his help, I look at amps after everything else is nailed down. I hope I end up as happy as you are with yours. To have only spent $400 and produce something akin to an iconic speaker is simply amazing.
pbn audio continues to build large WMmTmMW speakers. pbn KAS-2 pics
Cabinet includes diffraction control baffle bevels.
==========
Suggestion for good value (1", 4", 8", 12" ) 7-speaker set for one WMmTmMW =$800
1" dome tweeter 4-ohms
sb26stcn-04
Xover 1500Hz with BW3 or LR4
4" midrange. 2-parallel = 4-ohms
sb12nrx25-08
Xover 1500Hz with BW3 or LR4
Xover 400Hz with LR4
8" woofer. 2-parallel = 4-ohms
sb23nrx45-08
Xover 400 with LR4
Xover 100Hz with LR4
12" woofer. 2-parallel = 4-ohms
sb34nrxl75-8
Xover 100Hz with LR4
All drivers in sealed enclosures.
Cabinet includes diffraction control baffle bevels.
==========
Suggestion for good value (1", 4", 8", 12" ) 7-speaker set for one WMmTmMW =$800
1" dome tweeter 4-ohms
sb26stcn-04
Xover 1500Hz with BW3 or LR4
4" midrange. 2-parallel = 4-ohms
sb12nrx25-08
Xover 1500Hz with BW3 or LR4
Xover 400Hz with LR4
8" woofer. 2-parallel = 4-ohms
sb23nrx45-08
Xover 400 with LR4
Xover 100Hz with LR4
12" woofer. 2-parallel = 4-ohms
sb34nrxl75-8
Xover 100Hz with LR4
All drivers in sealed enclosures.
Attachments
The original Dunlavy speakers were passive. I modeled mine after the SC-IV model but instead of passive, I used 1 input with 3 outputs from the miniDSP 2x4. I connect each speaker driver directly to an amplifier, so to balance them, I used a Umic from miniDSP with REW software to set the crossover slopes, frequencies & accurately smooth the response & balance the levels. no resistors required.
Many folks get caught up in the "more power" game. If you model the speaker drivers in the box, you'll likely discover that you'll reach xmax at or below the rated power level of the drivers. On my plans to model after the SC-V, which uses a pair each of 12" woofers, 6.5" lower mids, 3.5" upper mids, and a single tweeter. If you double the number of drivers & each has it's own amplifier, that provides 6dB increase in output which is equiv. to 4x the power of a typical single woofer speaker. So if each woofer gets 120w from a class D amp, that's 240 watts for just the bass thru 2 drivers, so for a typical single woofer speaker, it'd be like 500w per channel. the LM3886 amps make 75w available to each of the lower & upper mids & tweeter & provides outstanding dynamic range capability, but would burn our most mids & tweeters since it exceeds their long-term power rating. So, for ~$100 per speaker, I get the equiv. of 500w per channel output with this speaker design. Dunlavy's designs were really well thought out, because doubling the driver count allowed him to successfully utilize 1st order crossover slopes, because of the overall speaker efficiency, you didn't need to use a lot of power to play them loudly.
These are nicer (more expensive, at least) drivers that I'm using. I'm using all Peerless (vifa) drivers, & GRS 12" woofers. Not in the same league as these you've selected, but mine are about 1/10 the cost. The miniDSP gets the most out of any speaker driver and the drivers I've chosen are similar to what he used back then. I surely can't match them as closely as he did, but i can smooth the response well with the DSP using the cal'd mic. Before Dunlavy sold the company for health reasons, he designed the Magnus, a prototype DSP/active speaker because of the profound advantages a DSP offers for setting up a speaker. He was pretty excited about it, but they were never manufactured.
I modeled the sb34nrxl75-8 woofers & they are really designed for a vented box & don't too well in a sealed box. I couldn't get f3 below 50hz in a sealed box. In a room they may go lower.
In the SC-IV, Dunlavy used 2 Scanspeak Classic 25W/8565-00 10" Woofers in 75L each to lower the Q to about .6 for a milder rolloff of the 1st octave below fs, which allowed him to get into the low 20s in a typical room. I haven't been able to determine who's 12" woofers he used in the SC-V. Since I'm using the miniDSP, I can really flatten the GRS 12SW woofer response in an 82L box for f3 of 30hz & even use a Linkwitz Transform biquad in the miniDSP for flat response to 20hz lower in a room. I give up some dynamic range, since the woofer has xmax of 8.5, but there are two of them per channel, 4-12 woofers in total for a stereo pair. The modeled output still easily exceeds my needs for SPL.
I modeled the sb34nrxl75-8 woofers & they are really designed for a vented box & don't too well in a sealed box. I couldn't get f3 below 50hz in a sealed box. In a room they may go lower.
In the SC-IV, Dunlavy used 2 Scanspeak Classic 25W/8565-00 10" Woofers in 75L each to lower the Q to about .6 for a milder rolloff of the 1st octave below fs, which allowed him to get into the low 20s in a typical room. I haven't been able to determine who's 12" woofers he used in the SC-V. Since I'm using the miniDSP, I can really flatten the GRS 12SW woofer response in an 82L box for f3 of 30hz & even use a Linkwitz Transform biquad in the miniDSP for flat response to 20hz lower in a room. I give up some dynamic range, since the woofer has xmax of 8.5, but there are two of them per channel, 4-12 woofers in total for a stereo pair. The modeled output still easily exceeds my needs for SPL.
"
John Dunlavy claimed his speakers did not follow the D'Appolito design theory, rather he considered his design to be SDA (symmetrical driver array. He was an antenna engineer/inventor). The D'Appolito's MTM were designed with very close driver placement using 18dB/octave crossovers to minimize lobing.
Mr. Dunlavy claimed (and backed it up with actual measurements) that his speakers were designed to be listened to at least 10 ft away. At that distance, the lobing effects were gone due to: his time-aligned driver mounting.; the SDA arrangement; the diffraction absorbing felt & 1st order crossover slopes. His design required all these attributes to perform well. He seemed to space his drivers to allow room for the F13 felt between & around all midrange & tweeters. I've followed all these aspects of the design and got great results, but when I experimented with steeper slopes, they didn't image near as well. I'm not done experimenting though!
He was also fanatical that the box had to be sealed, because vented boxes alters the "Q" drastically the screwed up the phase. Never tested this, just built the sealed boxes, appropriate for the drivers I chose.
retearl...Your analysis is correct.....
A $145 subwoofer like the Dayton RSS390HF-8 models well in a modest volume sealed enclosure, and is one example of a better 12" for this design. Heavier cone 12" Dayton Ultimax UM12-22 subwoofers will produce deeper bass from an acceptable volume sealed enclosure, but their high Mms and stiff suspensions require much higher power from common Class-D plate amps with built-in LR4 crossovers. In the USA, the $170 Dayton Ultimax UM12-22 is a common HT SUBwoofer.
==========
For an all SB-Acoustics design, the lower cost ($175) SB34NRX75-6 has better T/S parameters for a sealed box woofer alignment. Highly recommended by top designers like Zaph.
I look at the SB34NRXL75-8 as a very musical woofer which can produce low bass with a Linkwitz Transform and extra watts of power. With four of these woofers in this modest volume listening room, the extra power demands are also modest. The SB34NRXL75-8 is an expensive($230) "until death do us part" woofer with great transients and flexible future usage in higher frequency crossover 3-way configurations, and it also has high enough efficiency to put bass under a Pro audio midrange plus horn tweeter design. Google: Tweek Geeks BMF-1 speaker. When you discussed separate woofer boxes, I thought you might also desire SB34NRXL75-8 flexibility.
=========
The pbn KAS-2 speaker is very well designed. Great cabinet to MIMIC.
The single large cabinet provides both the weight for low vibration bass, and also the total box volume for deep bass with modest equalization demands. Attractive bevel cuts reduce diffraction and give the big box a "waist".
A $145 subwoofer like the Dayton RSS390HF-8 models well in a modest volume sealed enclosure, and is one example of a better 12" for this design. Heavier cone 12" Dayton Ultimax UM12-22 subwoofers will produce deeper bass from an acceptable volume sealed enclosure, but their high Mms and stiff suspensions require much higher power from common Class-D plate amps with built-in LR4 crossovers. In the USA, the $170 Dayton Ultimax UM12-22 is a common HT SUBwoofer.
==========
For an all SB-Acoustics design, the lower cost ($175) SB34NRX75-6 has better T/S parameters for a sealed box woofer alignment. Highly recommended by top designers like Zaph.
I look at the SB34NRXL75-8 as a very musical woofer which can produce low bass with a Linkwitz Transform and extra watts of power. With four of these woofers in this modest volume listening room, the extra power demands are also modest. The SB34NRXL75-8 is an expensive($230) "until death do us part" woofer with great transients and flexible future usage in higher frequency crossover 3-way configurations, and it also has high enough efficiency to put bass under a Pro audio midrange plus horn tweeter design. Google: Tweek Geeks BMF-1 speaker. When you discussed separate woofer boxes, I thought you might also desire SB34NRXL75-8 flexibility.
=========
The pbn KAS-2 speaker is very well designed. Great cabinet to MIMIC.
The single large cabinet provides both the weight for low vibration bass, and also the total box volume for deep bass with modest equalization demands. Attractive bevel cuts reduce diffraction and give the big box a "waist".
retearl,
Nice to hear about your success with minidsp and multiway! I agree with all your findings and considerations, based on my experience with my 4-way dipoles with minidsp and hearing some friends' systems.
I too prefer closed box for bass, with some room-adjusted boost for lowest octave! The result is an amazingly smooth, low-distortion, snappy sound, naturalness and easiness, transparency etc. finesses included. The benefit of high-end drivers, dsp or amplifiers is negotiable because the genius lies in operating at each drive unit's comfort zone.
Nice to hear about your success with minidsp and multiway! I agree with all your findings and considerations, based on my experience with my 4-way dipoles with minidsp and hearing some friends' systems.
I too prefer closed box for bass, with some room-adjusted boost for lowest octave! The result is an amazingly smooth, low-distortion, snappy sound, naturalness and easiness, transparency etc. finesses included. The benefit of high-end drivers, dsp or amplifiers is negotiable because the genius lies in operating at each drive unit's comfort zone.
No.
Two drivers fed with the same signal (no matter wether from individual amps or not) means displacement volume will double ==> 6dB more acoustic output in terms of SPL.
DSP controlled 4 way active speakers
Hiya,
Here are a few suggestions for your DSP controlled 4 way active design:
(1) DriveRack VENU360 | dbx Professional Audio
This is a great example of modern audio DSP which will be a significant upgrade compared to any passive crossover John (or anyone else) could have designed then or now. Modern pro audio gear offers higher performance and far greater flexibility compared to mini DSP audiophile / DIY type kit.
(2) Use 8 identical mono block or 4 stereo amplifiers to power your 8 drivers.... Matching power output is not critical, but maintaining identical phase / time domain performance is!
(3) Use three separate boxes per channel: 1 small box for your dome midrange and tweeter. Volt dome mid (600Hz to around 4KHz) with Beyma TPL 150 makes a killer combo.
(4) 2nd box for a Beyma 12MC500 (80Hz to 600Hz) and a third box for an RCF 15 inch sub.
(5) I have attached a couple of simulations showing box dimensions using 24mm birch ply laminated layer construction similar to my studio monitors here CIA ? New SM 15 monitors | Custom Install Audio
Hope this helps and all the very best
Derek.
Custom Install Audio | Advanced audio equipment
Hiya,
Here are a few suggestions for your DSP controlled 4 way active design:
(1) DriveRack VENU360 | dbx Professional Audio
This is a great example of modern audio DSP which will be a significant upgrade compared to any passive crossover John (or anyone else) could have designed then or now. Modern pro audio gear offers higher performance and far greater flexibility compared to mini DSP audiophile / DIY type kit.
(2) Use 8 identical mono block or 4 stereo amplifiers to power your 8 drivers.... Matching power output is not critical, but maintaining identical phase / time domain performance is!
(3) Use three separate boxes per channel: 1 small box for your dome midrange and tweeter. Volt dome mid (600Hz to around 4KHz) with Beyma TPL 150 makes a killer combo.
(4) 2nd box for a Beyma 12MC500 (80Hz to 600Hz) and a third box for an RCF 15 inch sub.
(5) I have attached a couple of simulations showing box dimensions using 24mm birch ply laminated layer construction similar to my studio monitors here CIA ? New SM 15 monitors | Custom Install Audio
Hope this helps and all the very best
Derek.
Custom Install Audio | Advanced audio equipment
double the power and you get +3dB.
For the special condition where the doubling of area there can be upto +3dB available due to the improved coupling from moving diaphram to air.
But neither he nor you have attempted to specify the special conditions for which any of that improved coupling will exist.
This is entirely correct.
Double the power (into one single driver) and you get +3dB more acoustic pressure level. Correct, since the signal (input voltage) has only increased by 3dB (factor of sqrt(2)) and SPL scales with input voltage, not input power.
Double the power (by using two woofers getting the same input voltage) and you get +6dB more acoustic pressure level. Correct as well and that's what has been under discussion here.
Coupling might add a few dB's more but that's another topic.
you have still not specified the special conditions that need to apply to gain any of that extra +3dB.
Work outwith those conditions and you'll not get it. Get close and you get some of it.
Work outwith those conditions and you'll not get it. Get close and you get some of it.
C'mon! You brought up that irrelevant acoustic coupling topic, not me (nor did anyone else).
dB increase for driver count & power
Sorry folks, I didn't mean to throw y'all in the ring over the +3dB vs 6dB thing. I didn't actually measure the increase, but I thought if I double the driver count & doubled the power via each driver being driven by matching amplifier channels since they each receive the same drive signal that that would equate to +6dB.
I know that with the miniDSP, as I adjust levels for each of the outputs, I can't say that I can discern a 1dB change, but I can discern a 2 dB change, a 3db change is easy for me to notice.
I guess the easy test is to set up my digital dB meter at listening position, then play only left or right channel, measure, then play both channels, all on a mono signal.
I first built just one speaker to test "Proof of Concept". Was all the discussion by John Dunlavy, of why he did, what he did regarding his design, something that I could benefit from? Answer: Wow, yes! When I finally finished (still ugly unfinished boxes) the second speaker, it is MARKEDLY louder, than just the one speaker playing, not subtly. A stereo pair has twice the driver area & twice the power of a single speaker.
I will pick a fight by saying that active vs passive crossovers do provide an improvement in detail, I think due to the voice coil of the speaker directly connected to the amplifier output. I have a pair of MTM speakers with passive crossovers I build years ago using Via drivers. I used a spare pair of bass/mid drivers for the MTM pair as the mid pair on one of my prototypes of this "Dunlavy" speaker. Obviously more clarity with the miniDSP active crossover. I've checked the freq response of both with my Umic & with very similar response measured with REW, the active just sounds cleaner. Placebo effect??
Sorry folks, I didn't mean to throw y'all in the ring over the +3dB vs 6dB thing. I didn't actually measure the increase, but I thought if I double the driver count & doubled the power via each driver being driven by matching amplifier channels since they each receive the same drive signal that that would equate to +6dB.
I know that with the miniDSP, as I adjust levels for each of the outputs, I can't say that I can discern a 1dB change, but I can discern a 2 dB change, a 3db change is easy for me to notice.
I guess the easy test is to set up my digital dB meter at listening position, then play only left or right channel, measure, then play both channels, all on a mono signal.
I first built just one speaker to test "Proof of Concept". Was all the discussion by John Dunlavy, of why he did, what he did regarding his design, something that I could benefit from? Answer: Wow, yes! When I finally finished (still ugly unfinished boxes) the second speaker, it is MARKEDLY louder, than just the one speaker playing, not subtly. A stereo pair has twice the driver area & twice the power of a single speaker.
I will pick a fight by saying that active vs passive crossovers do provide an improvement in detail, I think due to the voice coil of the speaker directly connected to the amplifier output. I have a pair of MTM speakers with passive crossovers I build years ago using Via drivers. I used a spare pair of bass/mid drivers for the MTM pair as the mid pair on one of my prototypes of this "Dunlavy" speaker. Obviously more clarity with the miniDSP active crossover. I've checked the freq response of both with my Umic & with very similar response measured with REW, the active just sounds cleaner. Placebo effect??
earl,
Have you tried 24 db LR? You will get minimal loobing in the sound stage and the driver seems to shoot the sound straight out.
In the best of worlds you only need the dsp EQ on the bass to remove the rooms main standing wave. Try eliminate nasty frequency problems by sliding the x-over points and volumes in each band. EQ in the upper area will prevent the sound from "pop out".
Well worth a preset!
If you still have problems in the upper area you may look for another driver for that band.
Check your rooms fundamental bass problem.
https://www.youtube.com/watch?v=qNf9nzvnd1k
Have you tried 24 db LR? You will get minimal loobing in the sound stage and the driver seems to shoot the sound straight out.
In the best of worlds you only need the dsp EQ on the bass to remove the rooms main standing wave. Try eliminate nasty frequency problems by sliding the x-over points and volumes in each band. EQ in the upper area will prevent the sound from "pop out".
Well worth a preset!
If you still have problems in the upper area you may look for another driver for that band.
Check your rooms fundamental bass problem.
https://www.youtube.com/watch?v=qNf9nzvnd1k
I've experimented quite a bit with the miniDSP crossovers. I think I've probably tried every slope, and actually only recently switched to Bessel. So far, it's my favorite. Outstanding sound stage, the best so far with this setting. The speakers are still located in my shop, which has NO sound absorption, so I use REW to identify bad things in the shop & miniDSP to fix those problems with equalization. Sound bouncing off welders & machining equip are probably not optimal for diffusion! The only saving grace out there is that it's big, ceiling is up 20 ft & back wall is 45 ft, so few early reflections besides the floor.
- Status
- Not open for further replies.