It's nice to see a group project idea expanding to consider including accompanying active processing.
As an avid active enthusiast, I'll though a few cents in...
For DIY speakers, I've found separating DSP from amplification and from speaker to be my best route.
The 3-4 plate amp with DSP builds I've done, while seemingly an elegant design solution, have not been worth the trouble. The plate amps have been a pain to design and build into the speaker cabinets, and they generate at least undesirable if not problematic heat. (I've had no reliability problems yet however.)
If I do use a plate amp, I use it externally like any other amp...and just build a box for it with input and speaker connections.
Perhaps worst of all with plate amps, is that they lock you into their DSP implementation. (This reason alone drives me away from Hypex, no matter how nice their amps are. )
Ime, having amplification and DSP separated is a huge plus for ongoing development and experimentation.
A good multi-channel amp can always be used, for whatever present or future speaker one has in mind.
DSPs are a constantly evolving domain....perhaps as much about users DSP learning growth, as with the hardware/software itself.
For ongoing DIY, I love having separate amps and DSP, external to the speaker. Multi-channel speakon connectors handle 4-ways and lower, with a single speaker cable.
If I was trying to build an "end game DIY" (an idea I consider to be an oxymoron 😉)...I might be willing to consider a plate amp inside a speaker.)
But in no other case would I even consider binding DSP with amplification.
As an avid active enthusiast, I'll though a few cents in...
For DIY speakers, I've found separating DSP from amplification and from speaker to be my best route.
The 3-4 plate amp with DSP builds I've done, while seemingly an elegant design solution, have not been worth the trouble. The plate amps have been a pain to design and build into the speaker cabinets, and they generate at least undesirable if not problematic heat. (I've had no reliability problems yet however.)
If I do use a plate amp, I use it externally like any other amp...and just build a box for it with input and speaker connections.
Perhaps worst of all with plate amps, is that they lock you into their DSP implementation. (This reason alone drives me away from Hypex, no matter how nice their amps are. )
Ime, having amplification and DSP separated is a huge plus for ongoing development and experimentation.
A good multi-channel amp can always be used, for whatever present or future speaker one has in mind.
DSPs are a constantly evolving domain....perhaps as much about users DSP learning growth, as with the hardware/software itself.
For ongoing DIY, I love having separate amps and DSP, external to the speaker. Multi-channel speakon connectors handle 4-ways and lower, with a single speaker cable.
If I was trying to build an "end game DIY" (an idea I consider to be an oxymoron 😉)...I might be willing to consider a plate amp inside a speaker.)
But in no other case would I even consider binding DSP with amplification.
In the past the highest crossover frequency in a passive 3-way was usually 800 Hz. Which limits the woofer to be some bass restricted construction.
If you are able to cross at 200 hz or lower, you can use dedicated, modern "bass only" driver. The price for a working 200 Hz passive crossover is rediculous, even if you are limited to 12 dB/oct.
Active makes such low crossover frequency price independent. We can use a dedicated bass driver and have a high quality mid-bass that is using it's most linear and distortion free range. This is not about some stupid religious active-passive dispute, but about extremely audible sonic differences.
Anyone who ever build such a system will not discuss a passive low frequency crossover any more. The option to use a lower powered mid-high amp and a powerfull bass amp is just too convincing, even with an analog x-over. Anyone who questiones the advantages of a DSP for room correction is simply an audio dinosaur.
If you are able to cross at 200 hz or lower, you can use dedicated, modern "bass only" driver. The price for a working 200 Hz passive crossover is rediculous, even if you are limited to 12 dB/oct.
Active makes such low crossover frequency price independent. We can use a dedicated bass driver and have a high quality mid-bass that is using it's most linear and distortion free range. This is not about some stupid religious active-passive dispute, but about extremely audible sonic differences.
Anyone who ever build such a system will not discuss a passive low frequency crossover any more. The option to use a lower powered mid-high amp and a powerfull bass amp is just too convincing, even with an analog x-over. Anyone who questiones the advantages of a DSP for room correction is simply an audio dinosaur.
I have a house full of speakers of all kinds of shapes/sizes/directivity concepts at the moment, all built over the last 4 years..
So, I have no excuse to build another speaker soon (except refining my current designs). Otherwise, I would have volunteered for at least some prototyping work for this project.
But I will eagerly follow this thread and learn from it (and help out with something or the other if any need arises) and maybe apply the learnings in a speaker design at the next excuse I get to build a speaker.. 🙂
I have been thinking about such a 3-way for quite some time.
In any case there has to be an active woofer/ mid section. Tweeter/ mid could be passive without too much disadvantage. The DSP can iron out some irregularities of the passive part too, if no serious mistakes are made in the passive crossover. Anyway, I see active 3-way still as the better option because of flexability and hardly any extra cost.
My driver choice would be a balance of low cost and high quality. There are some nice constructions around that do nothing wrong in a limited frequency range and are still low cost.
Bass should be a 10" (sub) woofer, up to 100-200 Hz. Such a driver will give more impact than a classical, high reaching 12" classic woofer.
Next I think of a 6" mid woofer, as a smaller driver doesn't deliver the kind of punch I expect. Up to a typical 2-way crossover frequency like 2000-3000 Hz. An alternative could be two 5" in MTM configuration, but this would make the look quite different and double costs. There is no advantage in using high excursion on a small mid driver.
On the top a nice wave guided 25mm/ 1" dome tweeter. The wave guide makes it simpler to deal with the quite large baffle. If no wave guide is used, the upper part of the baffle needs some serious shaping. 30° seem to be the sweet spot to counter edge problems.
Volume should be around 50-70 liter net for the bass, at the upper limit for a vented construction with the choosen driver. That way you can have a shallow, early drop of in the low end, that can nicely be equalized with the DSP.
30 Hz linear should be no problem with such a cabinet. For those that oppose vented cabinets for religious reasons, done right such a construction reduces distortion compared to a closed one with an identical sized driver.
With a crossover of 18dB/oct, lower than 200 Hz, there will be no sonic problems from the vent. A 12" passive radiator would still be an alternative.
So far for my low budged, high quality construction plan.
In any case there has to be an active woofer/ mid section. Tweeter/ mid could be passive without too much disadvantage. The DSP can iron out some irregularities of the passive part too, if no serious mistakes are made in the passive crossover. Anyway, I see active 3-way still as the better option because of flexability and hardly any extra cost.
My driver choice would be a balance of low cost and high quality. There are some nice constructions around that do nothing wrong in a limited frequency range and are still low cost.
Bass should be a 10" (sub) woofer, up to 100-200 Hz. Such a driver will give more impact than a classical, high reaching 12" classic woofer.
Next I think of a 6" mid woofer, as a smaller driver doesn't deliver the kind of punch I expect. Up to a typical 2-way crossover frequency like 2000-3000 Hz. An alternative could be two 5" in MTM configuration, but this would make the look quite different and double costs. There is no advantage in using high excursion on a small mid driver.
On the top a nice wave guided 25mm/ 1" dome tweeter. The wave guide makes it simpler to deal with the quite large baffle. If no wave guide is used, the upper part of the baffle needs some serious shaping. 30° seem to be the sweet spot to counter edge problems.
Volume should be around 50-70 liter net for the bass, at the upper limit for a vented construction with the choosen driver. That way you can have a shallow, early drop of in the low end, that can nicely be equalized with the DSP.
30 Hz linear should be no problem with such a cabinet. For those that oppose vented cabinets for religious reasons, done right such a construction reduces distortion compared to a closed one with an identical sized driver.
With a crossover of 18dB/oct, lower than 200 Hz, there will be no sonic problems from the vent. A 12" passive radiator would still be an alternative.
So far for my low budged, high quality construction plan.
Although there is clearly interest in the subject of the project since dropping the monkey coffin aesthetic a group does not seem to be forming to move it forward. Changing the objectives within a thread with monkey box in the title likely hasn't helped. Unless someone is keen I think we should accept a group project hasn't happened.
Over the next month or two I hope to examine the configuration options for a near the wall active cardioid and how well it integrates with the room. I will likely combine this with developing some software to better address the room integration side of things so it won't happen quickly.
Over the next month or two I hope to examine the configuration options for a near the wall active cardioid and how well it integrates with the room. I will likely combine this with developing some software to better address the room integration side of things so it won't happen quickly.