Hi Kipman! This print took around 8 hours. With some slight modifications of the design we estimate that could decrease by 20% or so.
The material (PIPG) is fairly cheap, around 6-10 euro per kilo depending on specs. That’s the good part. The bad part is that the system is quite steep
For these purposes one could lose the robot arm, and attach the extruder to a gantry system. That would be more economic.
The material (PIPG) is fairly cheap, around 6-10 euro per kilo depending on specs. That’s the good part. The bad part is that the system is quite steep
For these purposes one could lose the robot arm, and attach the extruder to a gantry system. That would be more economic.
Indeed Kipman, when compared to conventional plywood it is indeed more expensive. If compared to filament 3d printing it is orders faster, stronger, bigger and cheaper however.
The bit what interests me most though: it makes large, complex, curved and organically shaped speaker designs possible. And do so without the need for large batches or complicated assembly. I'm excited to see where it could take us!
The bit what interests me most though: it makes large, complex, curved and organically shaped speaker designs possible. And do so without the need for large batches or complicated assembly. I'm excited to see where it could take us!
Mine loaded with 18sw115 drivers and build with Birch ply have the opposite and require a cut at 70 Hz ish. Printed is an entirely different beast though, nice work!
cheers
Martin
Last edited:
Hey Art,
Inspired by one of your posts a while back, among other things, I recently did some pretty extensive numerical modelling (and qualification measurements on scale models) to assess when or if horn subs become 'more directional' than direct radiators. Cutting right to the meat of the results, for an array of cabinets having the same frontal radiating area, there is such a tiny difference in directivity until ka=5.7 that it falls within the margin of error.
I only had enough time and computing resources to run models for full 3D, coupled lumped element and boundary element models of up to 3x2 arrays in ground plane (effectively 12x cabinets, or frontal area of 7.84 m^2), but the data showed the same thing for any number of boxes. Each simulated box was qualified with real-world measurements of electrical impedance, acoustic throat impedance, axial sensitivity and the like.
What was really interesting is that the manufacturer's own 'direct sound' prediction software gave results that became less and less valid beyond this size array, which I found during the validation process. That appears to be because the cabinet geometry isn't included in the modelling; making a simple cuboid and putting theoretical point sources at the locations of the driver centres gave directivity results that matched the more complicated models, without crazy calculation times.
Anyway, the main thing is that for all intents and purposes, if the radiating area is a significant portion of the cabinet front, you can consider subs to be equivalent when making arrays. That means gradient, end-fire, and any other kind of crazy pattern control method should work just fine - that echoes my experiences in the field, doing bass arrays with front-loaded horns and tapped horns. Which is nice
I haven't done an asymmetrical radiating element like a tapped horn sub yet, as my supervisor had a hard enough time keeping me focused as it was. Considering the wavelengths are long compared to the 'plain wood' front part of the box, and changing the port size on the reflex subs made little difference to normalised directivity, I don't expect to see anything crazy occurring. Always nice to check, though... maybe I'll add it to the 'pet projects' pile.
At the current price of BB grade plywood, and probably more than 8 hours of manpower to cut, glue, and assemble a cabinet from scratch, it might actually be closer than you think!
Super cool idea; I'm curious if you've got any data or done any tests on the resonance of the material versus plywood? Or plywood with paint, such as polyurea applied? Stiffness is one thing, but that can lead to ringing; if there are resonances from the enclosure, that may explain the uneven measured response in the upper region. The wind will definitely have affected the data too, even with lots of averaging.
Last edited:
Thanks guys. That’s the abridged version of some four or so months of work on my dissertation. It feels good to be able to do that after some hairy moments during the process!
There’s an ABEC / Akabak thread on here somewhere right? I should post a few of the quirks I found in the process to help others avoid some long periods of head scratching after running a calculation for 26 hours
I do plan to revisit and expand on the topic. A broader set of configurations and cabinet types like tapped horns, deeper statistical analysis of the results, data on the impact of acoustic boundaries like stage curtains on acoustic throat and transfer impedance, same for subs in front in horn-loaded end-fire setups. There are lots of bits I didn’t get around to. COMSOL might be calling me…
Hopefully it’ll get to a point where someone considers it suitable for publishing and can be cited by people out ‘in a field’. I need to do something/anything else for a month or so, first though!
Anyway, for now this hopefully suffices to get some fun arrays deployed by those who already own (or want to build) more efficient subs. Boxes like the Keystone have great external dimensions to allow for physical or digital arc configurations - Merlijn can Veen’s SAD spreadsheet is a great place to start for that.
There’s an ABEC / Akabak thread on here somewhere right? I should post a few of the quirks I found in the process to help others avoid some long periods of head scratching after running a calculation for 26 hours
I do plan to revisit and expand on the topic. A broader set of configurations and cabinet types like tapped horns, deeper statistical analysis of the results, data on the impact of acoustic boundaries like stage curtains on acoustic throat and transfer impedance, same for subs in front in horn-loaded end-fire setups. There are lots of bits I didn’t get around to. COMSOL might be calling me…
Hopefully it’ll get to a point where someone considers it suitable for publishing and can be cited by people out ‘in a field’. I need to do something/anything else for a month or so, first though!
Anyway, for now this hopefully suffices to get some fun arrays deployed by those who already own (or want to build) more efficient subs. Boxes like the Keystone have great external dimensions to allow for physical or digital arc configurations - Merlijn can Veen’s SAD spreadsheet is a great place to start for that.
Hello guys, i'm pretending to build 8 keystone in a 2 time project, this year 4 boxes and next year 4 more.
Here in my country i'm not able to buy the 18sw115, just 2 other models that are similar, the 18sw100 and the 18ds100.
Maybe @weltersys or someone else could help me to determine wich one is the best for the Keystone? I'll be using the 8ohms model for better amplification match in the final project with 8 boxes.
I'll be putting the image of both speakers parameteres.
Thanks!
Here in my country i'm not able to buy the 18sw115, just 2 other models that are similar, the 18sw100 and the 18ds100.
Maybe @weltersys or someone else could help me to determine wich one is the best for the Keystone? I'll be using the 8ohms model for better amplification match in the final project with 8 boxes.
I'll be putting the image of both speakers parameteres.
Thanks!
Attachments
Viniciusv,
The 18sw100 is a bit closer to the 18SW115 parameters than the 18ds100.
I'd suggest ordering one of each and test them in the cabinet to see which performs better, though each could use a different exit size/shape to optimize response.
Either of the speakers could easily use 3000 watts peak, which requires around 155 volts for an "8 ohm" driver, 110 volts at "4 ohm". What amplification are you considering?
Art
I'm tending to go all the way with the 18sw100...
Also i'm considering to go with an handmade amplifier that is build by an friend. It has 14.000wrms tested with continuos sine wave for more than 10 minutes. I know that i'll not get the best audio quality, but is what i have in hands for the best budget. Brazil sucks for buying imported products like powersoft, labgruppen etc...
Thanks Art!
Also i'm considering to go with an handmade amplifier that is build by an friend. It has 14.000wrms tested with continuos sine wave for more than 10 minutes. I know that i'll not get the best audio quality, but is what i have in hands for the best budget. Brazil sucks for buying imported products like powersoft, labgruppen etc...
Thanks Art!
@viniciusv where in Brazil are you based? I have friends in São Paulo and Curitiba that can source Powersoft, Linea Research, and 18SW115 drivers if you’d like to get a quote from them.
I also had good results running TH118 from the Studio R X12 amplifiers, even though the spec on paper doesn’t seem to be big enough.
With regards to the the SW100 range for a tapped horn, I’d suggest getting the SW115 if possible. Due to a mix up, my business partner ordered four of the SW100 for some 21” tapped horns we were using for a run of festivals, and they were clearly struggling compared to the other SW115-loaded boxes.
I’m not sure what music you’re playing, but having witnessed first hand what many Brazilians class as ‘acceptable volume’ I think you’ll need all the headroom you can get
I also had good results running TH118 from the Studio R X12 amplifiers, even though the spec on paper doesn’t seem to be big enough.
With regards to the the SW100 range for a tapped horn, I’d suggest getting the SW115 if possible. Due to a mix up, my business partner ordered four of the SW100 for some 21” tapped horns we were using for a run of festivals, and they were clearly struggling compared to the other SW115-loaded boxes.
I’m not sure what music you’re playing, but having witnessed first hand what many Brazilians class as ‘acceptable volume’ I think you’ll need all the headroom you can get