Small speaker design for electronic organ

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have followed DIY speaker design for years, but haven't yet built my own. I'm considering doing so now in order to fit a big electronic church organ into my living room. Allen Organ speakers work very well to reproduce pipe organ sound, but their full-range cabinets are about 18" x 18" x 33", making even a four-channel setup just too big for my living room. I really want to fit an eight-channel or larger instrument.

My difficulty in figuring out a good speaker design is that organ music is probably the most demanding in terms of frequency response. Most channels will need to go down to 32Hz, and the bass-only channel down to 16Hz.

In summary, based on what I know of speaker design, the challenges and simplifying assumptions are:

  1. The speakers have to be as small as possible, since I'll have a bunch of them (e.g., seven plus subwoofer).
  2. Cost is important for the same reason, but I'm more sensitive to size.
  3. Frequency response has to have good usable output down to 32Hz, but a slow rolloff as from a sealed enclosure can be compensated for in voicing adjustments. Usable output is the key; it doesn't have to be flat, although that would certainly be nice.
  4. Efficiency isn't too important. I only need to fill my living room, not a concert hall. The organ already has good amplifiers that can drive a 4-ohm load.
  5. Stereo imaging is not important, as each channel is independent. Broad dispersion of the sound is preferred.
  6. While I'm decent at woodworking, I don't plan to make these cabinets look too fancy, and don't want to tackle anything truly complicated.
What recommendations would you all have for this unusual kind of project? I started off by reading a bunch of articles and threads here, and then posted in the full-range forum, not realizing that those folks want to get every frequency out of one speaker :). I don't mind using two drivers instead of one, so I think here in the multi-way forum is the right place to ask my question.
 

ICG

Disabled Account
Joined 2007
You said stereo imaging isn't important, yet you want a lot of channels. Could you please elaborate on that? I mean, it would be a hell of a lot cheaper and space saving, if you'd reduce the channels to one or two, Subwoofer and sat. I understand that the organ got several output channels and amps but that can be electronically summed-up to one or two channels, even if the organ doesn't have low-level outputs for each channel.

As for the 16Hz, there are 2 solutions I can think of atm:

Strassacker, Komponenten: Lautsprecher, Frequenzweichen, Bauelemente

TL-SUB 30 | Visaton

There are surely more which can reproduce that low frequencies.
 
Your satellite speakers don't need to go down to 35Hz. The sub will take care of that. 80Hz is fine.

On wall speakers can be great, but they need to be designed specifically for on wall/bookshelf operation.

Zaph audio and Troels gravesen have kits that are good sounding and relatively cheap. Although I would stay away from the overpriced crossover components troels uses.
 
You said stereo imaging isn't important, yet you want a lot of channels. Could you please elaborate on that?
The short answer is that for electronic organ sound reproduction, more channels is better, and electronic summing is always detrimental to the sound. So if you don't care about the details, just take it as a given that I need 7 full-range speakers going down to 32Hz.

The long answer is two-fold. First, in real pipe organs, the keyboards each have their own division of pipes, so the sound comes from different locations depending on which stop you pull. Reproducing that in an electronic organ requires many channels, unless you have some sort of sophisticated convolution reverb that can simulate placing the sound sources in several distant locations and then reproducing all their reflections and reverberation in a big room before getting reproduced by a smaller number of speakers.

Even if reason #1 didn't exist, reason #2 is that electronic summing causes interference between the simulated pipe sounds from different notes and different stops. For illustration purposes, suppose that the two sounds for two different stops happen to be the same waveshape, and I'm hitting the same note for each. That means they both will be playing the same pitch and waveform, but the exact phase of the two waves will depend on when I hit each key. If I happen to hit them when they're exactly in phase, then the two will sum to sound like a single note that's twice as loud. If I happen to hit them when they're exactly out of phase, they'll cancel each other out and I'll hear nothing. Neither result sounds anything like having two identical pipes playing next to each other. Acoustical mixing is currently the only practical solution to this problem. I can imagine a convolution reverb simulating the waves bouncing around and then getting reproduced by a pair of speakers, but don't know of any solution that will actually do what's necessary. Let me know if there is one!

As for the 16Hz, there are 2 solutions I can think of atm:

http://www.lautsprechershop.de/hifi/sub320xxl.htm]Strassacker, Komponenten: Lautsprecher, Frequenzweichen, Bauelemente[/url]

TL-SUB 30 | Visaton
Thanks for the sub links. I'll investigate those. Since I will need only one or at most two subs, I'm less focused on that part of the problem right now. I want to figure out the solution for the 7 or 8 full-range speakers for the main channels first. BTW, the organ already has a LPF built into the audio system so the sub channel is routed by itself to its own amp. The sub will only have to worry about the really low bass.
 
What's your budget?

Can you built infinite baffle speakers, can you mount speakers in a wall, floor or ceiling where the front points into the music room and back points into another room?

My past church had a very nice electronic organ with many different speakers around the church. I popped some of the grills to find what I'd consider mid-grade hi-fi parts, Peerless tweeters and Jensen woofers. 70s MAC parts.

maybe start with this woofer for a three way DSA315-8.
 
...What recommendations would you all have for this unusual kind of project?
Allen Organ speakers specs typically tell little about true values, an embelished presentation. The majority of models are PA drivers that are most famous of shallow bass, yet you speak of their performance as being satisfactory so you would probably do as well if you made a clone of these.

On the other hand, if you wanted a truly smaller loudspeaker capable of playing in the 30 Hz (-6dB) region, a specially made drivers with low Fs and of lower sensitivity in free space (about 80-82 dB/2.83V/1m), then a 6.5" class units and a tweeter would make a good combination in a 1 cuft cabinet.
 

ICG

Disabled Account
Joined 2007
The short answer is that for electronic organ sound reproduction, more channels is better, and electronic summing is always detrimental to the sound. So if you don't care about the details, just take it as a given that I need 7 full-range speakers going down to 32Hz.

That is not true because it completely depends on the level it's played at. With high volume the speaker(s) have to make wide excursions, which do have a negative impact on different played notes. But with the introduction of the subwoofer, the other speaker(s) will only have to play with a fraction of the former excursion, especally when you don't need high volume.

The long answer is two-fold. First, in real pipe organs, the keyboards each have their own division of pipes, so the sound comes from different locations depending on which stop you pull. Reproducing that in an electronic organ requires many channels, unless you have some sort of sophisticated convolution reverb that can simulate placing the sound sources in several distant locations and then reproducing all their reflections and reverberation in a big room before getting reproduced by a smaller number of speakers.

That's actually quite easy. The output of each channel can be distributed via a very simple mixer to two channels, each channel 'positioned' via a pan pot, which places them on the 'stage'. For the reverb and amplification buy a cheap, used multichannel home cinema receiver, even older ones have an excellent room simulation by the dsp and you can chose of the room type/size at any time. For the output 2 speakers can be used or, to increase the room impression, 4. The speakers can be much smaller and cheaper because they don't have to go down to 100, let alone 32Hz since the subwoofer can't be located because of the low crossover frequency.

The reduction to 2 or 4 speakers, which don't need to go all the way down and a single subwoofer will make it much, much cheaper than the original 7 channels (which will give you no reveberation at all). The receiver will do nearly all, the amplification, the sub/sat crossover, the reverb, only the small mixer is needed. The mixer and the receiver will be likely less than 300 bucks and you still save a lot on the 2 or 4 speakers instead of full range of 7 speakers.

Even if reason #1 didn't exist, reason #2 is that electronic summing causes interference between the simulated pipe sounds from different notes and different stops. For illustration purposes, suppose that the two sounds for two different stops happen to be the same waveshape, and I'm hitting the same note for each. That means they both will be playing the same pitch and waveform, but the exact phase of the two waves will depend on when I hit each key. If I happen to hit them when they're exactly in phase, then the two will sum to sound like a single note that's twice as loud. If I happen to hit them when they're exactly out of phase, they'll cancel each other out and I'll hear nothing. Neither result sounds anything like having two identical pipes playing next to each other.

That is not true, it will only be cancelled if it's the exact frequency and loudness (amplitude) and a pure sine wave. Since the channels are distributed (unevenly on 2/4 channels with the pan and the room simulation) and the organ not only plays plain sine waves but have a lot of different overtones, your scenario will not happen the way you're imagining it.

Acoustical mixing is currently the only practical solution to this problem. I can imagine a convolution reverb simulating the waves bouncing around and then getting reproduced by a pair of speakers, but don't know of any solution that will actually do what's necessary. Let me know if there is one!

See, I don't want to shove something down your throat but there are other options you haven't thought about or didn't even knew they existed. If 2 or 4 sat speakers don't work for you, you still got the option to extend it to 7 and buy or build more.

BTW, the organ already has a LPF built into the audio system so the sub channel is routed by itself to its own amp. The sub will only have to worry about the really low bass.

Okay, then the other channels don't have to go down to 32Hz anyway. You just need to find out at which frequency the LPF crosses over.
 
What's your budget?
That's TBD. I'm trying to figure out how much it will cost to do the job right, and that cost will factor into how much I can afford to pay for a larger instrument.

Can you built infinite baffle speakers?
No, unfortunately that is not an option.

maybe start with this woofer for a three way DSA315-8.
Thanks for the recommendation. Can you explain to me why you suggested that one? I'd like to learn what parameters you noticed that make it interesting. I am concerned that a 12" woofer is going to make the resulting speaker too big to fit seven of them on my wall though.
 
Allen Organ speakers specs typically tell little about true values, an embelished presentation. The majority of models are PA drivers that are most famous of shallow bass, yet you speak of their performance as being satisfactory so you would probably do as well if you made a clone of these.
For their bass cabinets, they do publish specs:
Allen Organ Company


That SR-5 would do the job, but it obviously would be too large for my house! Even the B-20 is bigger than I want.


On the other hand, if you wanted a truly smaller loudspeaker capable of playing in the 30 Hz (-6dB) region, a specially made drivers with low Fs and of lower sensitivity in free space (about 80-82 dB/2.83V/1m), then a 6.5" class units and a tweeter would make a good combination in a 1 cuft cabinet.
That sounds like exactly what I'm looking for. Do you have any suggestions on where I could find such drivers, or better yet, designs utilizing this kind of driver? The lower sensitivity is not a problem, since I only need to fill a living room with sound, not an auditorium like these organs usually have to fill.
 
That's TBD. I'm trying to figure out how much it will cost to do the job right, and that cost will factor into how much I can afford to pay for a larger instrument.

No, unfortunately that is not an option.

Thanks for the recommendation. Can you explain to me why you suggested that one? I'd like to learn what parameters you noticed that make it interesting. I am concerned that a 12" woofer is going to make the resulting speaker too big to fit seven of them on my wall though.


Surf over to Parts Express and google the woofer. It looks to work well into the 30s and under 2 cubic foot box with good efficiency. It's a good balance of parameters. Wall mount would boast midbass some to give a full sound.


Also look at the RSS265HO-44 in pairs for subwoofers. Tune them to 18 hz with boost and place in a corner.
 
That is not true because it completely depends on the level it's played at. With high volume the speaker(s) have to make wide excursions, which do have a negative impact on different played notes. But with the introduction of the subwoofer, the other speaker(s) will only have to play with a fraction of the former excursion, especally when you don't need high volume.
I think I gave the wrong impression. I need seven full-range cabinets that go down to 32 Hz, plus a subwoofer that goes down to 16 Hz. The seven full-range channels have to be able to stand on their own, since the subwoofer is only going to get bass from the channel that handles the 32' stops (the ones that go down to 16 Hz). Six of the other channels have 16' stops that go down to 32 Hz, and do NOT have their bass frequencies routed to the subwoofer. So technically, I could have one cabinet that rolls off much higher than 32 Hz, but all the rest have to be able to produce 32 Hz on their own without the subwoofer.

The output of each channel can be distributed via a very simple mixer to two channels, each channel 'positioned' via a pan pot, which places them on the 'stage'.
That won't solve the phase cancellation problem completely; it will reduce it somewhat, but it will still be there.

BTW, the organ already has its own amplifiers capable of filling an auditorium with sound, albeit with Allen's own efficient speakers. The organ also has its own 8-channel reverberation system built-in.

[Referring to electronic summing and cancellation artifacts:] That is not true, it will only be cancelled if it's the exact frequency and loudness (amplitude) and a pure sine wave.
You're right that it is unlikely that the waves will completely cancel out, but I can hear the same problem in non-hypothetical situations with my current small organ. It is a real problem. Imagine you're playing back two pipe samples at roughly the same frequency, but not precisely the same frequency, since real pipes are never perfectly in tune. Those two samples will beat against each other as the waves add constructively and cancel destructively depending on their relative phase. It sounds horrible when you play them back through the same channel with electronic mixing instead of acoustical mixing. The perfect (extreme) example is when the two waves are identical and cancel each other out completely, giving 0 volume. Mixing the same signal acoustically just gives you full volume with the two channels out of phase.

Electronic organ builders do all kinds of tricks to try to avoid this problem, like assigning similar stops to different audio channels, and even splitting the same stop into C and C# channels so that different notes from the same stop don't beat against each other. But the bottom line is that there is no complete solution to the problem without more audio channels.

The bottom line is that my goal is to build a small speaker that will produce 32 Hz tone at living room volume levels. "Small" is defined as small enough to fit seven of these into my living room. The space I have is some space to the left and right of the organ console, and I can build a shelf above the console and place speakers across that wall (about 11' wide). I'll probably use the space to the right of the console for the subwoofer that will handle the 16 Hz material.
 
scholls suggestion is excellent -

for the mains and 6.5" woofer you could probably get by for $40 per cabinet for driver cost. FWIW I simed this woofer in a 16 liter box - the vent is pretty long even with 2" ID.


Dayton Audio DC160-8 6-1/2" Classic Woofer

HmkVKLy.png
 
Last edited:
...That's actually quite easy. The output of each channel can be distributed via a very simple mixer to two channels, each channel 'positioned' via a pan pot, which places them on the 'stage'...
Sorry, but that's not the same thing. This method simply creates "phantom" images that move around depending on the listening position, and are still subject to the electrical summing/ cancellation problems.

Mlaird is correct: Discrete speakers are the best approach here, assuming they can be placed at sufficient distance from each other. Note that especially in the lower frequencies, if the speakers are to close to each other you begin to experience significant acoustic cancellation as well, although this never seems quite as pronounced as the electrical kind.
 
Last edited:

ICG

Disabled Account
Joined 2007
I think I gave the wrong impression. I need seven full-range cabinets that go down to 32 Hz, plus a subwoofer that goes down to 16 Hz. The seven full-range channels have to be able to stand on their own, since the subwoofer is only going to get bass from the channel that handles the 32' stops (the ones that go down to 16 Hz). Six of the other channels have 16' stops that go down to 32 Hz, and do NOT have their bass frequencies routed to the subwoofer. So technically, I could have one cabinet that rolls off much higher than 32 Hz, but all the rest have to be able to produce 32 Hz on their own without the subwoofer.

No, I do not. You got a wrong concept of what I said. You do not need 7 speakers go down to 32Hz since the summation of all 7 channels already includes ALL of the complete reproduced channels and ALL frequencies. The separation between sub and the other channels happens in the receiver. The subwoofer plays all of the sub-woofer frequencies of all channels. You cannot locate deep frequencies so it does not matter if it's just one subwoofer or 7 speakers which play low frequencies.

That won't solve the phase cancellation problem completely; it will reduce it somewhat, but it will still be there.

You have the same problem with your 7 speakers, there will be a place where they can actually cancel each other out. They do not have the same phase at all, even if you think you press the keys at the same time. The phase to each other will determine where the addition or subtraction (cancelling) happens, no matter if that's in the mix or 'only' from the speakers.

BTW, the organ already has its own amplifiers capable of filling an auditorium with sound, albeit with Allen's own efficient speakers. The organ also has its own 8-channel reverberation system built-in.

You don't need to use them, it's just the signal of it tapped for the mixer. If you'd rather spend more on more speakers and use all built-in, sure, you can do that but the difference is, one will be much, much cheaper and save a ton of room.

You're right that it is unlikely that the waves will completely cancel out, but I can hear the same problem in non-hypothetical situations with my current small organ. It is a real problem.

Yes, I know exactly what you are talking about. But you don't seem to grasp that exactly that happens acoustically too from different speakers. It depends from the phase of the signals, effectively the distance from the speakers to your ears and on the reflections in your room.

Electronic organ builders do all kinds of tricks to try to avoid this problem, like assigning similar stops to different audio channels, and even splitting the same stop into C and C# channels so that different notes from the same stop don't beat against each other. But the bottom line is that there is no complete solution to the problem without more audio channels.

Well, I wonder how synths do it then, they also create the sounds electronical and most only put out one or two channels. Allen does many things the old way. I don't understand the passive crossover for a subwoofer i.e., that's outdated since at least 30 years. They got digital technology in their organs for a long time but on some things they are way behind. That applies to other organ manufacturers too ofcourse and I fully understand why, they simply don't need to change it in such a traditional-loving market but just because they don't do it, doesn't mean it can't be done.

The bottom line is that my goal is to build a small speaker that will produce 32 Hz tone at living room volume levels. "Small" is defined as small enough to fit seven of these into my living room. The space I have is some space to the left and right of the organ console, and I can build a shelf above the console and place speakers across that wall (about 11' wide). I'll probably use the space to the right of the console for the subwoofer that will handle the 16 Hz material.

See, that's perfectly fine if you want to do it that way, but you should do it because of the right reasons.
 
...You have the same problem with your 7 speakers, there will be a place where they can actually cancel each other out...
This simply doesn't happen the way you've explained it.

..They do not have the same phase at all, even if you think you press the keys at the same time. The phase to each other will determine where the addition or subtraction (cancelling) happens, no matter if that's in the mix or 'only' from the speakers...
This implies that there's no difference between electrical and acoustic cancellation, but that's not the case at all.

When 2 or more signals are electrically summed, the math is pretty simple, and the plus & minus is predictable & effective. By comparison, the acoustic field is impossibly more complex, and the additions & subtractions are (thankfully, in this case) nowhere near as absolute.

Mixing down to a stereo panned spread is simply tweaking the electrical sums; it doesn't really work the same as multiple output channels & speakers.

...Well, I wonder how synths do it then, they also create the sounds electronical and most only put out one or two channels...
Apples & oranges. Synthesizers don't solve this problem at all - it's just less noticeable because they generally produce different types of sounds. If you were to program a synth to recreate the same type of multiple similar sounds (with multiple pitch references) that a digital organ produces, you'd have the same problems Mlaird is describing.

...Allen does many things the old way. I don't understand the passive crossover for a subwoofer i.e., that's outdated since at least 30 years...
Not sure of your point here, since Allen hasn't used a passive sub crossover for at least that long.

They got digital technology in their organs for a long time but on some things they are way behind...
I'll certainly agree with you on that general sentiment. But when is comes to multiple outputs & speakers, that is one area where they're definitely getting it right.
 

ICG

Disabled Account
Joined 2007
This simply doesn't happen the way you've explained it.

This implies that there's no difference between electrical and acoustic cancellation, but that's not the case at all.

When 2 or more signals are electrically summed, the math is pretty simple, and the plus & minus is predictable & effective. By comparison, the acoustic field is impossibly more complex, and the additions & subtractions are (thankfully, in this case) nowhere near as absolute.

To start with it, mathematically it only cancels if the phase is 180° turned and the amplitude is the same. That does not happen here because it's not on the same amplitude but in different channels on different volume/amplitude aswell, hence the mathematically the cancellation to zero cannot ever happen. Since it's also distributed to two channels, there is always a difference between each channel, the room simulation adding delay, reveberation and phase shifts, working that to 2 or 4 speakers, each a different signal and does not sum it up back to one single channel (then there would indeed happen the cancellation). That means, what's reproduced by the speakers has never a 100% cancellation.

Not sure of your point here, since Allen hasn't used a passive sub crossover for at least that long.

Allan hasn't? :eek: Then I must be imagining that at the B-20 bass cabinet. :rolleyes:

allen.png

As you can see, it's still in their program:
Allen Organ Company

:rolleyes:
 
I'm guessing the point of mlaird's exercise is to somewhat emulate the complex interactions, beat-tones and spatial aspect of an organ's pipe array? (I've only heard a pipe organ once and it wasn't being played loud so have no live reference)

(one user of that Dayton 6.5 pointed out its real graph was rougher looking than Dayton's stretched graph - might be usable to 1K5)

in a small room I once ran two of my 4 - 18" RJ and 4 Karlson 12 - that was fun on organ recordings
and a good sub (K16 ala' Exemplar) in a mono system was pretty fine too on theater organ cds

l7OjCYg.jpg
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.