At sufficient distance, and with drivers that small, the line is short enuff that it pretty much acts as a single driver.
dave
Using a specially designed, very unsensitive 3”driver with an Fs on the order of one-half if a typical 3”.
dave
A phase plug in a driver that small seems to be marketing to me. Looks cool, but a driver that small does not really need it (most times),
dave
With a lot of room to move...
But the distortion graphs don't look encouraging though.
The bass I would cross at baffle step frequency like 2mh inductor.
The highs I would not cut but make with the full ranges an omni on all sides of the box. Dispersion in the highs would be good with it, too.
So my advice 1.5 way. Only one inductor for the "bass" and eq the rest of the response egalized with smartfone eq or better
The highs I would not cut but make with the full ranges an omni on all sides of the box. Dispersion in the highs would be good with it, too.
So my advice 1.5 way. Only one inductor for the "bass" and eq the rest of the response egalized with smartfone eq or better
1.5 way could work. Simple solution and probably better coherent. I am going to keep my expectations in limit. I am now targeting 70hz flat in room response.
To me, excursion = distortion.
Xmax, I don't care as much about.
But I see the point for a mini array of 4 or more if you want to as the highs knocked down from combing.
Pricy drivers though.....
Audience a3.
Xmax, I don't care as much about.
But I see the point for a mini array of 4 or more if you want to as the highs knocked down from combing.
Pricy drivers though.....
Audience a3.
Not in the real world on a tiny baffle and full space losses below 250 Hz
It wont go lower than 100 Hz.
Easier to use a woofer.
And cheaper.
Basic speaker sim is assuming Half Space and not accounting
losses and baffle step.
Its just a basic transfer function to match a speaker to a filter alignment
for rising and extended highs, one could use the Faitals and make a line at a fraction of the A3 line price.
Very good drivers, the Faitals. I'm sure they would give the A3 a run for that kind of money.
Hello @wesayso , need more explanation for above statements.
how is cross over designed to achieve this.
Here's some ideas of using multiple drivers, all producing bass, but only one treble. The one on the left is my traffic light speakers, with one driver facing the rear, the middle was an idea for a floor stander, and the one on the right is the quad cube.
Attachments
It looks like this in a schematic:
So it's still a 5x5 series/parallel but with a few passive parts added in the mix. Due to having one of the driver groups split up (5 center drivers with a group of drivers around it, 3 below and 2 on top) I added some extra notches to keep it mostly symmetrical around 3 -5 KHz.
(above are the filters used in one array, below are the notch filters, two of those for each array)
I started with an unfiltered 5x5 driver series/parallel and added the filters as best I could to make the vertical pattern control as tight as I could. You could do even better when starting fresh with this goal in mind 🙂. I have done a few schematics like that for fun.
The vertical pattern control of the filtered array compared to an unfiltered array:
I wanted a wide enough vertical coverage to be able to listen sitting down and standing up, something an unfiltered array does quite well.
People are always afraid of comb filtering in an array. What they don't realise is how much the room causes combing with even a single driver playing. This, plus the fact that Stereo listening is always causing a comb filtering at the ear due to the left speaker reaching the right ear and vice versa. So the combing that does exist in an array like this, while it is doing something (as you need boost to get it up in level) it's not nearly as detrimental as some will have you believe.
Looking at the in room frequency response of a single driver compared to a (filtered) array with floor and ceiling reflections added makes this clear:
The orange line is predicted "in-room" results with floor and ceiling reflections added at a level of -6 dB (quite a realistic number).
So which one has more comb filtering going on... the magic word with arrays is averaging. Averaging of reflections to be more clear.
Along the same lines of making a multi point measurement to take out the room, each driver has its own position, relative to floor and ceiling.
When you combine them all, you average out the effect of floor and ceiling reflections. The more drivers in the array, the better that result becomes.
Up to a limit of course, as the geometry has to be right to make each frequency sum at the listening spot. The filtered array does a better job at that.
It's still not perfect, but quite a fun compromise. The hard part is to land on the right tonal balance, as you have to draw your own. You can't make it flat anechoic and put it in the room as we do with a "normal" speaker, as in this case the room is part of the mix.
Get it right though and it will be surprisingly clear with a lot of potential.
Setup right (making sure parallel planes are treated) and with the right EQ to back it up, it is a fun but above all, potentially loud and clear speaker.
And with the right driver and positioned so the room helps out, it can make bass too.
with added damping panel behind listener, many more examples in my thread.
The Peerless/Vifa TC9 FD18-08 driver can do this, the Scan Speak 10F 8414G10 I have now? Not so much, the distortion of that one will rise at the low end.
The 10F is a better driver overall (*), but the TC9 can do the whole spectrum if your room supports it. Don't try to do it with the array setup away from walls in a large room.
Know what to expect, with the room helping out and at least one of them in a corner it can do a good job. Baffle about half a meter out in front of the wall.
My average listening level to have fun is around 85 to 87 dB (peak comes on top of that).
(*) above ~300 Hz it's distortion is lower than the TC9, in an array this difference is quite clear.
So it's still a 5x5 series/parallel but with a few passive parts added in the mix. Due to having one of the driver groups split up (5 center drivers with a group of drivers around it, 3 below and 2 on top) I added some extra notches to keep it mostly symmetrical around 3 -5 KHz.
(above are the filters used in one array, below are the notch filters, two of those for each array)
I started with an unfiltered 5x5 driver series/parallel and added the filters as best I could to make the vertical pattern control as tight as I could. You could do even better when starting fresh with this goal in mind 🙂. I have done a few schematics like that for fun.
The vertical pattern control of the filtered array compared to an unfiltered array:
I wanted a wide enough vertical coverage to be able to listen sitting down and standing up, something an unfiltered array does quite well.
People are always afraid of comb filtering in an array. What they don't realise is how much the room causes combing with even a single driver playing. This, plus the fact that Stereo listening is always causing a comb filtering at the ear due to the left speaker reaching the right ear and vice versa. So the combing that does exist in an array like this, while it is doing something (as you need boost to get it up in level) it's not nearly as detrimental as some will have you believe.
Looking at the in room frequency response of a single driver compared to a (filtered) array with floor and ceiling reflections added makes this clear:
The orange line is predicted "in-room" results with floor and ceiling reflections added at a level of -6 dB (quite a realistic number).
So which one has more comb filtering going on... the magic word with arrays is averaging. Averaging of reflections to be more clear.
Along the same lines of making a multi point measurement to take out the room, each driver has its own position, relative to floor and ceiling.
When you combine them all, you average out the effect of floor and ceiling reflections. The more drivers in the array, the better that result becomes.
Up to a limit of course, as the geometry has to be right to make each frequency sum at the listening spot. The filtered array does a better job at that.
It's still not perfect, but quite a fun compromise. The hard part is to land on the right tonal balance, as you have to draw your own. You can't make it flat anechoic and put it in the room as we do with a "normal" speaker, as in this case the room is part of the mix.
Get it right though and it will be surprisingly clear with a lot of potential.
Setup right (making sure parallel planes are treated) and with the right EQ to back it up, it is a fun but above all, potentially loud and clear speaker.
And with the right driver and positioned so the room helps out, it can make bass too.
with added damping panel behind listener, many more examples in my thread.
The Peerless/Vifa TC9 FD18-08 driver can do this, the Scan Speak 10F 8414G10 I have now? Not so much, the distortion of that one will rise at the low end.
The 10F is a better driver overall (*), but the TC9 can do the whole spectrum if your room supports it. Don't try to do it with the array setup away from walls in a large room.
Know what to expect, with the room helping out and at least one of them in a corner it can do a good job. Baffle about half a meter out in front of the wall.
My average listening level to have fun is around 85 to 87 dB (peak comes on top of that).
(*) above ~300 Hz it's distortion is lower than the TC9, in an array this difference is quite clear.
Last edited:
Late to the discussion, but I have built several speakers with using all the same small drivers but not all playing the same range. Here is one example.
In this the center driver is the tweeter, the other four drivers are the woofer. IIRC the crossover was first order, but I don't remember the frequency. I've also done this same array with the "tweeters" as the driver on the ends - either center or outside, depending on how I wanted it to sound. It all blended well enough and the crossover was easy, but there was no way to get it to play below 100Hz, so a sub was used.
In this the center driver is the tweeter, the other four drivers are the woofer. IIRC the crossover was first order, but I don't remember the frequency. I've also done this same array with the "tweeters" as the driver on the ends - either center or outside, depending on how I wanted it to sound. It all blended well enough and the crossover was easy, but there was no way to get it to play below 100Hz, so a sub was used.
Hello All, found another worthy candidate
https://sbacoustics.com/product/3in-sb10pgc21-4/
https://sbacoustics.com/product/3in-sb10pgc21-4/
- Nominal Impedance 4 Ω
- Linear coil travel (p-p) 4.5 mm
- Free air resonance, Fs 89 Hz
- Sensitivity (2.83V/1m) 84 dB
- Rated power handling* 20 W
can some one help me find an online calculator or sw which allows multiple drivers to calculate sealed/vented box. thanks in Advance
Imagine you make two separate boxes and simply put them together, so increase the volume as needed.