I had an attempt at this a little while ago, but Mark's project (on the full-range forum) inspired me to get back to it.
Right now, I'm listening to 12x Bose SoundDock drivers (273488004) in a straight array. They're wired in 3x groups of series to 2x NL4 connectors on the back. Using a Powersoft T604, I've got EQ, delay, level etc control over each bank of three drivers.
I would, however, very much like it if I could get the column working well with just one amp channel - that way, putting them to use in the PA system gets much easier. A single 4-channel amp could do main L/R, sub and a monitor mix.
Couple of images of what I have at the mo:
The 12" "sub" has been dispensed with, though - I'll probably just use my 15"s and put these columns on stands.
So, at the moment I have 12 drivers in a vertical line, but I find the vertical off-axis response to be preferable when the banks are delayed by 0.08ms, 0ms, 0.08ms, 0.2ms (top to bottom). Those correspond to a distance of a few centimetres - well within the bending capabilities of most wood.
The sound is pretty good, and actually gets fairly loud. Despite the sealed box design, I've got bass to 100Hz. I plan on adding a bass reflex port (as Art did with his Dirty Dozen Array - https://www.diyaudio.com/forums/multi-way/236037-dirty-dozen-line-array.html) to provide ventilation for the drivers, and reduce excursion at the bottom of the passband. The drivers are audibly distressed at around 4mm p/p travel, so keeping them out of that danger zone will be important.
I have another four drivers here in a box, and 16 more on their way. The amps I have on-hand deliver most power into 4ohm, which is convenient - the drivers are 4ohm nominal, so a simple series/parallel wiring scheme will get back to 4ohm.
So, the next steps are:
- Measure T/S parameters, come up with a ported box alignment for these drivers
- Design a 16-driver cabinet with a curved front.
- Build another prototype. Test test test. Come up with the correct processing, inc. alignment with subs.
- Build the finished pair.
Stay tuned!
Chris
Right now, I'm listening to 12x Bose SoundDock drivers (273488004) in a straight array. They're wired in 3x groups of series to 2x NL4 connectors on the back. Using a Powersoft T604, I've got EQ, delay, level etc control over each bank of three drivers.
I would, however, very much like it if I could get the column working well with just one amp channel - that way, putting them to use in the PA system gets much easier. A single 4-channel amp could do main L/R, sub and a monitor mix.
Couple of images of what I have at the mo:
The 12" "sub" has been dispensed with, though - I'll probably just use my 15"s and put these columns on stands.
So, at the moment I have 12 drivers in a vertical line, but I find the vertical off-axis response to be preferable when the banks are delayed by 0.08ms, 0ms, 0.08ms, 0.2ms (top to bottom). Those correspond to a distance of a few centimetres - well within the bending capabilities of most wood.
The sound is pretty good, and actually gets fairly loud. Despite the sealed box design, I've got bass to 100Hz. I plan on adding a bass reflex port (as Art did with his Dirty Dozen Array - https://www.diyaudio.com/forums/multi-way/236037-dirty-dozen-line-array.html) to provide ventilation for the drivers, and reduce excursion at the bottom of the passband. The drivers are audibly distressed at around 4mm p/p travel, so keeping them out of that danger zone will be important.
I have another four drivers here in a box, and 16 more on their way. The amps I have on-hand deliver most power into 4ohm, which is convenient - the drivers are 4ohm nominal, so a simple series/parallel wiring scheme will get back to 4ohm.
So, the next steps are:
- Measure T/S parameters, come up with a ported box alignment for these drivers
- Design a 16-driver cabinet with a curved front.
- Build another prototype. Test test test. Come up with the correct processing, inc. alignment with subs.
- Build the finished pair.
Stay tuned!
Chris
The HF response is mostly a function of EQ - I can get them flat to 18kHz, but it needs around 10dB of boost there. Alternatively, flat to 15kHz only needs about 3dB. Since my live vocal mics (Sennheiser e935s) drop like a rock at around 12kHz, I think it'll be fine for live music.
For playback, those with good hearing will spot the difference in a comparison to something that is flat out to 20kHz. But hey, it's unlikely that side-by-side comparisons will be made, so I'm not too worried. If something like that does come up, then I suppose I'll have to hope the design stands up for itself.
For what I'm planning on using these for, I think 15kHz will do just fine.
Edit - cross-post with JMF. You're thinking of the 801s and 901s - they used 4" drivers and are probably the reason people say "no highs, no lows, must be Bose". These drivers do a reasonable amount of treble, but don't have the cone area or linear excursion to do anything useful in a PA system <100Hz.
Chris
PS - a quick test with a vocal mic reveals it's very resistant to feedback. Holding the mic a foot away from my mouth, I could whisper into the mic and hear the sound coming back from the speaker clearly.
For playback, those with good hearing will spot the difference in a comparison to something that is flat out to 20kHz. But hey, it's unlikely that side-by-side comparisons will be made, so I'm not too worried. If something like that does come up, then I suppose I'll have to hope the design stands up for itself.
For what I'm planning on using these for, I think 15kHz will do just fine.
Edit - cross-post with JMF. You're thinking of the 801s and 901s - they used 4" drivers and are probably the reason people say "no highs, no lows, must be Bose". These drivers do a reasonable amount of treble, but don't have the cone area or linear excursion to do anything useful in a PA system <100Hz.
Chris
PS - a quick test with a vocal mic reveals it's very resistant to feedback. Holding the mic a foot away from my mouth, I could whisper into the mic and hear the sound coming back from the speaker clearly.
A difficulty with such arrays is that the array transistions from line source to point source at a certain distance. This distance is frequency dependent and increases with frequency. In practice this means that low mids drop off at 6 dB for each doubling of distance, while high frequencies only do 3 dB. Curving the array somewhat alleviates this problem.
A true CBT solves it but costs a lot of high frequency output. JBL has got some interesting speakers which are marketed to be CBT. I believe the delay / curving is done by low pass filtering the signal at about 20 kHz. Larger delays are obtained by cascading these filters. All drivers seem to run at full power which should not be the case for a CBT. Maybe (just speculating) they only are CBT at midrange frequencies and driver attenuation is implemented in the low pass filters.
JBL: http://www.jblpro.com/ProductAttachments/JBLpro_CBT_brochure_July2012.pdf
Crossover: Passive Loudspeaker Delay - diyAudio
A true CBT solves it but costs a lot of high frequency output. JBL has got some interesting speakers which are marketed to be CBT. I believe the delay / curving is done by low pass filtering the signal at about 20 kHz. Larger delays are obtained by cascading these filters. All drivers seem to run at full power which should not be the case for a CBT. Maybe (just speculating) they only are CBT at midrange frequencies and driver attenuation is implemented in the low pass filters.
JBL: http://www.jblpro.com/ProductAttachments/JBLpro_CBT_brochure_July2012.pdf
Crossover: Passive Loudspeaker Delay - diyAudio
Last edited:
Hi TBTL,
Thanks for the info. When I was using this prototype with no delays, I had a very big problem with the transition from line to point source. The HF balance was only ever correct at one distance - move away, and there's too much treble. Get closer, and there isn't enough.
The vertical off-axis response was also a problem - if you weren't in-line with some part of the array, it sounded very bad.
Those two problems together were enough to banish the project into storage for a few years, and it's only now that I have a good 4-channel amplifier with lots of DSP that I decided to have another go at it.
Having the array curved (using delays for now) seems to have helped both of those problems, but I haven't done any serious outdoor testing yet.
Chris
Thanks for the info. When I was using this prototype with no delays, I had a very big problem with the transition from line to point source. The HF balance was only ever correct at one distance - move away, and there's too much treble. Get closer, and there isn't enough.
The vertical off-axis response was also a problem - if you weren't in-line with some part of the array, it sounded very bad.
Those two problems together were enough to banish the project into storage for a few years, and it's only now that I have a good 4-channel amplifier with lots of DSP that I decided to have another go at it.
Having the array curved (using delays for now) seems to have helped both of those problems, but I haven't done any serious outdoor testing yet.
Chris
What are the specs of an individual 273488004 driver?
I want to build an amp with two 273488004 drivers, and can't figure out what circuit to use.
each driver individualy requires ~18w, 8ohm impedance, and a -18to+18V AC power supply?
am i correct?
Would an amplifier circuit, with such an output and power supply, work with an individual 273488004 driver/speaker?
I want to build an amp with two 273488004 drivers, and can't figure out what circuit to use.
each driver individualy requires ~18w, 8ohm impedance, and a -18to+18V AC power supply?
am i correct?
Would an amplifier circuit, with such an output and power supply, work with an individual 273488004 driver/speaker?
IIRC, they're 4ohm drivers.
Here's the T/S I have for the one I tested:
Fs - 189Hz
Qes - 0.6
Qms - 6.41
Re - 3.28ohm
Sd - 20.4cm^2
Le - 0.3mH
Vas - 0.17L
Xmax - 2mm
Pe - 20w
Should be enough info to get going. FWIW, I've got some in small ported boxes tuned to 70Hz. They do okay.
A 12v class D amp should do fine with these drivers.
Chris
Here's the T/S I have for the one I tested:
Fs - 189Hz
Qes - 0.6
Qms - 6.41
Re - 3.28ohm
Sd - 20.4cm^2
Le - 0.3mH
Vas - 0.17L
Xmax - 2mm
Pe - 20w
Should be enough info to get going. FWIW, I've got some in small ported boxes tuned to 70Hz. They do okay.
A 12v class D amp should do fine with these drivers.
Chris
Thank you, finally someone on a forum actually answered a question, that i asked
So, a TDA8932 ( 35W TDA8932 Digital Amplifier Board ) Class D amp would be good, for two 273488004 drivers, wired in series?
35W mono digital amplifier module
Operating voltage: 10V - 30VDC
Recommended use voltage: 24V output 2A power supply
Recommended speaker impedance: 8 ohm Best (4 ohm or 6 ohm also works)
the two series drivers would be individualy experiencing 4ohm, 12V, 17.5W, so almost exactly what i need, right?
So, a TDA8932 ( 35W TDA8932 Digital Amplifier Board ) Class D amp would be good, for two 273488004 drivers, wired in series?
35W mono digital amplifier module
Operating voltage: 10V - 30VDC
Recommended use voltage: 24V output 2A power supply
Recommended speaker impedance: 8 ohm Best (4 ohm or 6 ohm also works)
the two series drivers would be individualy experiencing 4ohm, 12V, 17.5W, so almost exactly what i need, right?
Keep in mind that ribbon, AMT and planar tweeters are a line source/array for the higher frequencies.
Better not put drivers in series as their differences will be amplified.
Why not use TPA3116/8 instead of TDA8932?
Cheap TPA3118D2 boards, modding them and everything that comes with it
TPA3116D2 Amp
Better not put drivers in series as their differences will be amplified.
Why not use TPA3116/8 instead of TDA8932?
Cheap TPA3118D2 boards, modding them and everything that comes with it
TPA3116D2 Amp
Last edited:
Can you help me figure out the inputs?
strange buzz problem with input stage of home-made chip amp. | All About Circuits
in this thread, i was told to connect the -input terminal to ground. is that correct? i haven't seen any one else do this online, when using this amplifier.
the schematics are realy confusing me, and they do show some connections to ground, but the amplifier chip contains no connections to ground, so should i realy connect the negative input terminal to the ground on my power supply?
i just wan't to connect a simple mono jack socket to the input, and can't figure out how to wire it. i thought i knew how.
strange buzz problem with input stage of home-made chip amp. | All About Circuits
in this thread, i was told to connect the -input terminal to ground. is that correct? i haven't seen any one else do this online, when using this amplifier.
the schematics are realy confusing me, and they do show some connections to ground, but the amplifier chip contains no connections to ground, so should i realy connect the negative input terminal to the ground on my power supply?
i just wan't to connect a simple mono jack socket to the input, and can't figure out how to wire it. i thought i knew how.
I can only see very small pictures in that topic. But it reads to me that you have 2x (l+R) differential/balanced audio inputs: positive, ground negative, and you want to connect an unbalanced stereo audio signal: Left Ground Right.
You can do that by shorting the negatives to ground on the inputs. Connecting the ground audio input cable to both ground and negative pins does the same. So the ground from the audio input connects to ground and negative pins of both (L+R) inputs.
If this is not it, please open a new topic with some pictures we can see in https://www.diyaudio.com/forums/class-d/
You can do that by shorting the negatives to ground on the inputs. Connecting the ground audio input cable to both ground and negative pins does the same. So the ground from the audio input connects to ground and negative pins of both (L+R) inputs.
If this is not it, please open a new topic with some pictures we can see in https://www.diyaudio.com/forums/class-d/
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.