I received a tip about looking into the Quadratic-Throat Waveguide. It seems like a simple and effective solution for this design.
In the design shown in the whitepaper, the throat entry angle is straight (i.e., it doesn’t match the compression driver’s exit angle). However, I’m curious if matching the compression driver’s exit angle would give a better result.
I’ve attached an illustration showing the difference between matching and not matching the CD’s exit angle. The horn has an 80x60 coverage angle. The left illustration shows the top view and the right shows the side view. On each one, I’ve matched one side to the compression driver’s exit angle of 31°, while the other side remains unmatched. The transition ends up quite a bit longer when matching the CD’s angle.
Which one would you go for? My gut feeling says matching the CD’s exit angle would be a good idea.
In the design shown in the whitepaper, the throat entry angle is straight (i.e., it doesn’t match the compression driver’s exit angle). However, I’m curious if matching the compression driver’s exit angle would give a better result.
I’ve attached an illustration showing the difference between matching and not matching the CD’s exit angle. The horn has an 80x60 coverage angle. The left illustration shows the top view and the right shows the side view. On each one, I’ve matched one side to the compression driver’s exit angle of 31°, while the other side remains unmatched. The transition ends up quite a bit longer when matching the CD’s angle.
Which one would you go for? My gut feeling says matching the CD’s exit angle would be a good idea.
The final design of the first prototype is ready! I’ve started 3D printing some of the parts and plan to build the cabinet after the weekend.
I went with the horn profile that matches the CD’s opening of 31 degrees. The coverage angle is 80x60 degrees, and the crossover is set at 1200 Hz, meaning the distance between the mid ports is ¼ wavelength of 1200 Hz.
I’ve added a rear chamber for the mids that will be 3D printed. I initially wanted to go with a simple cylinder, but it would have been a bit too small, so I extended it downward with a cube. The combined rear chamber volume is 1 liter.
The cabinet dimensions are 47.5 cm (height) x 31 cm (width) x 28 cm (depth).
In the last design, the outer flare was made of two layers of 19 mm MDF. The ratio in Synergy Calc spreadsheet was around 0.75.
However, after running more Hornresp tests, I noticed the response would be more even if I lengthened the L12 (distance between the mid-plane and the beginning of the second flare) and made S3 larger (plane of the beginning of the second flare). I had to compromise on the second flare to keep it within my WAF-friendly size constraints. The ratio according to Synergy Calc is 0.85 now.
The width is slightly larger than before, 31 cm vs. 30 cm, and the 3D-printed part of the horn is much larger. I was worried I’d have to print it in multiple parts, but then I realized I could rotate it 45 degrees, and it fits perfectly on the printing plane, which is 256x256 mm.
3D Printing:
At the moment, I’m considering filling the 3D-printed part of the horn with epoxy. I’d print it with a gyroid infill at around 10%, then drill a hole to pour epoxy into the horn. That way, I can make it solid all the way through while saving filament and printing time.
My colleague suggested that the carbon fiber particles in the PETG-CF filament might be dangerous, so I’m planning to paint it after printing.
L-Pad:
The compression driver is going to be much louder than what I need, so I’m planning to add an L-Pad. I'm using a Topping PA5 II amp for the compression driver. Here’s a test in XSim:
DAC/DSP Module:
I’ve been assembling the DAC modules (following the tutorial mentioned in the first post). It turned out to be much more work than I expected. I initially missed soldering the jumpers on the back of the modules, so I spent a whole day troubleshooting before finding the issue. However, if you follow the instructions carefully, it shouldn’t be too complicated.
I’ve been following this tutorial to add a trigger output and remote, but I haven’t gotten it working properly yet. Not the most important feature at the moments, so I'll figure it out later.
I got a tip from dptucunduva to isolate the DAC modules and power them separately with a linear PSU. So, I’ve added a galvanic isolator (SI8660BA).
I’m still waiting for some parts to arrive in the mail, including the jack sockets for the audio and the linear PSU.
I can’t wait to build this and take some measurements! This project has taken so much time, but I’ve also learned so much in the process. Before this, I had no idea how to use Fusion360 or Hornresp. It's a good feeling to see the idea slowly turn into reality.
I’ll keep you guys updated!
I went with the horn profile that matches the CD’s opening of 31 degrees. The coverage angle is 80x60 degrees, and the crossover is set at 1200 Hz, meaning the distance between the mid ports is ¼ wavelength of 1200 Hz.
I’ve added a rear chamber for the mids that will be 3D printed. I initially wanted to go with a simple cylinder, but it would have been a bit too small, so I extended it downward with a cube. The combined rear chamber volume is 1 liter.
The cabinet dimensions are 47.5 cm (height) x 31 cm (width) x 28 cm (depth).
In the last design, the outer flare was made of two layers of 19 mm MDF. The ratio in Synergy Calc spreadsheet was around 0.75.
However, after running more Hornresp tests, I noticed the response would be more even if I lengthened the L12 (distance between the mid-plane and the beginning of the second flare) and made S3 larger (plane of the beginning of the second flare). I had to compromise on the second flare to keep it within my WAF-friendly size constraints. The ratio according to Synergy Calc is 0.85 now.
The width is slightly larger than before, 31 cm vs. 30 cm, and the 3D-printed part of the horn is much larger. I was worried I’d have to print it in multiple parts, but then I realized I could rotate it 45 degrees, and it fits perfectly on the printing plane, which is 256x256 mm.
3D Printing:
At the moment, I’m considering filling the 3D-printed part of the horn with epoxy. I’d print it with a gyroid infill at around 10%, then drill a hole to pour epoxy into the horn. That way, I can make it solid all the way through while saving filament and printing time.
My colleague suggested that the carbon fiber particles in the PETG-CF filament might be dangerous, so I’m planning to paint it after printing.
L-Pad:
The compression driver is going to be much louder than what I need, so I’m planning to add an L-Pad. I'm using a Topping PA5 II amp for the compression driver. Here’s a test in XSim:
DAC/DSP Module:
I’ve been assembling the DAC modules (following the tutorial mentioned in the first post). It turned out to be much more work than I expected. I initially missed soldering the jumpers on the back of the modules, so I spent a whole day troubleshooting before finding the issue. However, if you follow the instructions carefully, it shouldn’t be too complicated.
I’ve been following this tutorial to add a trigger output and remote, but I haven’t gotten it working properly yet. Not the most important feature at the moments, so I'll figure it out later.
I got a tip from dptucunduva to isolate the DAC modules and power them separately with a linear PSU. So, I’ve added a galvanic isolator (SI8660BA).
I’m still waiting for some parts to arrive in the mail, including the jack sockets for the audio and the linear PSU.
I can’t wait to build this and take some measurements! This project has taken so much time, but I’ve also learned so much in the process. Before this, I had no idea how to use Fusion360 or Hornresp. It's a good feeling to see the idea slowly turn into reality.
I’ll keep you guys updated!
The first prototype speaker is ready. The build has taken longer than expected, but it has been a good learning experience. I was hoping to be able to show you measurements, but I haven’t gotten to that yet. So this post will mostly be about the lessons learnt during the build.
The Speaker:
The horn took about 2.5 days to 3D print. It was printed with PETG CF, with a small layer height of 0.18mm, resulting in a super smooth finish. The print was almost perfect! The only issue was some minor stringing in one spot on the back, underneath the compression driver, but since it’s not visible, it wasn’t a big issue. A light sanding with 180-grit sandpaper fixed the issue.
Before painting, I read some posts online which suggested that paint wouldn't stick very well to PETG, but I did it anyway! I applied two coats of woodprimer, followed by a layer of green wood paint I had on hand. I haven’t had any issues with the paint so far, it seems to stick just as well to the PETG as to wood.
To make sure the 3D-printed horn would stick to the wood cabinet, I added screw mounts to the top and bottom of the 3D-printed horn.
I had a bit too much trust in the measurement system on the table saw I used, which resulted in all the wood pieces being about 1mm too large. This left slight gaps between the wood and the 3D-printed horn. I used wood filler to fill in the gaps, so now the cabinet is airtight. Next time, I’ll double-check the measurements before gluing.
Pictures from when I was fitting the pieces together… without the horn.
The screw mounts:
Cabinet assembled. The gaps being filled with wood filler:
Once the glue dried, I used a router to cut out the holes on the sides of the cabinet for the sub drivers. Then it was time to install the drivers. The compression driver and sub drivers were easy to mount, but the mids proved more challenging.
The 3D-printed mid rear chambers seemed like a smart design, but they had two major issues. First, the small cabinet size made it nearly impossible to reach the screws furthest inside. I managed to get them in place, but they’re not as tight as I’d like, so I’m slightly worried that the mid drivers aren’t secured well enough.
The second issue was the sharp edge on the back of the mid rear chamber. When designing it in Fusion 360, I didn’t account for the inner edge where the backplate screws in. So when mounting the mid drivers with the chamber attached, it was too tight. I had to remove some of the wood to make space for the sharp edge.
I used threaded inserts melted into the horn to screw the mids in place. However, the holes for the inserts were slightly too small, causing some to loosen. A bit of glue solved the problem, but this will be adjusted in future designs.
I’m planning to update the mid rear chamber.
The DAC/DSP:
The DAC/DSP has been a challenge. The jack plugs I ordered from AliExpress never showed up, so I had to order another set, which finally arrived a few days ago. Because of that, I haven’t been able to connect the DAC/DSP to all three amps yet.
Here are some of the mistakes I made during the DAC/DSP build. After wiring up the DAC chips according to the tutorial, I got no sound. I spent a whole day troubleshooting before finally rereading the tutorial and realizing I had missed a crucial detail, there were solder jumpers on the back that had to be shorted. Once that was sorted, the DAC finally made a sound. Lesson learned, RTFM!
I wanted a simple way to power on the system using the 12V trigger input on the amps, so I followed this tutorial to add a trigger output and a remote. I decided to skip the display since I don’t really need it. The scripts didn’t work, though, which I later found out was because they were designed for a Raspberry Pi 4, and I have a Pi 5. I was so eager to get it working that I spent two days writing my own script with help from ChatGPT. I managed to get it working after a while, but it doesn’t have all the functions of the script in the tutorial. The trigger output activates when music is playing, and I can control the volume with the remote. The other functions, like mute and config selection, don’t work with my script. But melomane13, who wrote the tutorial, has made an updated version, which I’ll install as soon as I get the DAC working again…
…because I also managed to fry one of the DAC boards. I was planning to use a 5V-to-12V converter for the trigger output, but in my impatience, I didn’t double-check the wiring. So I got a short circuit, and one DAC board was dead. Lesson learned!
I got a recommendation from dptucunduva to isolate the DAC boards from the Raspberry Pi using an SI8660BA galvanic isolator and power them separately with a linear PSU. Wired everything up and… nothing. No sound. I’ve done some troubleshooting, and it seems like some of the wiring is getting loose.
As you can see, the DAC/DSP is a bit of a mess right now. But hopefully, it’ll work beautifully once I fix those issues. I’ll keep you updated.
Here are pictures of the 3D model of the DAC/DSP enclosure and how it looks at the moment:
The Speaker:
The horn took about 2.5 days to 3D print. It was printed with PETG CF, with a small layer height of 0.18mm, resulting in a super smooth finish. The print was almost perfect! The only issue was some minor stringing in one spot on the back, underneath the compression driver, but since it’s not visible, it wasn’t a big issue. A light sanding with 180-grit sandpaper fixed the issue.
Before painting, I read some posts online which suggested that paint wouldn't stick very well to PETG, but I did it anyway! I applied two coats of woodprimer, followed by a layer of green wood paint I had on hand. I haven’t had any issues with the paint so far, it seems to stick just as well to the PETG as to wood.
To make sure the 3D-printed horn would stick to the wood cabinet, I added screw mounts to the top and bottom of the 3D-printed horn.
I had a bit too much trust in the measurement system on the table saw I used, which resulted in all the wood pieces being about 1mm too large. This left slight gaps between the wood and the 3D-printed horn. I used wood filler to fill in the gaps, so now the cabinet is airtight. Next time, I’ll double-check the measurements before gluing.
Pictures from when I was fitting the pieces together… without the horn.
The screw mounts:
Cabinet assembled. The gaps being filled with wood filler:
Once the glue dried, I used a router to cut out the holes on the sides of the cabinet for the sub drivers. Then it was time to install the drivers. The compression driver and sub drivers were easy to mount, but the mids proved more challenging.
The 3D-printed mid rear chambers seemed like a smart design, but they had two major issues. First, the small cabinet size made it nearly impossible to reach the screws furthest inside. I managed to get them in place, but they’re not as tight as I’d like, so I’m slightly worried that the mid drivers aren’t secured well enough.
The second issue was the sharp edge on the back of the mid rear chamber. When designing it in Fusion 360, I didn’t account for the inner edge where the backplate screws in. So when mounting the mid drivers with the chamber attached, it was too tight. I had to remove some of the wood to make space for the sharp edge.
I used threaded inserts melted into the horn to screw the mids in place. However, the holes for the inserts were slightly too small, causing some to loosen. A bit of glue solved the problem, but this will be adjusted in future designs.
I’m planning to update the mid rear chamber.
The DAC/DSP:
The DAC/DSP has been a challenge. The jack plugs I ordered from AliExpress never showed up, so I had to order another set, which finally arrived a few days ago. Because of that, I haven’t been able to connect the DAC/DSP to all three amps yet.
Here are some of the mistakes I made during the DAC/DSP build. After wiring up the DAC chips according to the tutorial, I got no sound. I spent a whole day troubleshooting before finally rereading the tutorial and realizing I had missed a crucial detail, there were solder jumpers on the back that had to be shorted. Once that was sorted, the DAC finally made a sound. Lesson learned, RTFM!
I wanted a simple way to power on the system using the 12V trigger input on the amps, so I followed this tutorial to add a trigger output and a remote. I decided to skip the display since I don’t really need it. The scripts didn’t work, though, which I later found out was because they were designed for a Raspberry Pi 4, and I have a Pi 5. I was so eager to get it working that I spent two days writing my own script with help from ChatGPT. I managed to get it working after a while, but it doesn’t have all the functions of the script in the tutorial. The trigger output activates when music is playing, and I can control the volume with the remote. The other functions, like mute and config selection, don’t work with my script. But melomane13, who wrote the tutorial, has made an updated version, which I’ll install as soon as I get the DAC working again…
…because I also managed to fry one of the DAC boards. I was planning to use a 5V-to-12V converter for the trigger output, but in my impatience, I didn’t double-check the wiring. So I got a short circuit, and one DAC board was dead. Lesson learned!
I got a recommendation from dptucunduva to isolate the DAC boards from the Raspberry Pi using an SI8660BA galvanic isolator and power them separately with a linear PSU. Wired everything up and… nothing. No sound. I’ve done some troubleshooting, and it seems like some of the wiring is getting loose.
As you can see, the DAC/DSP is a bit of a mess right now. But hopefully, it’ll work beautifully once I fix those issues. I’ll keep you updated.
Here are pictures of the 3D model of the DAC/DSP enclosure and how it looks at the moment:
Thank you! It’s like a baby HRE1, 48 × 31 × 28.5 cm compared to the HRE1’s 139 × 70 × 70 cm. A colleague who was working in the woodworking shop while I was building it said it looked like an owl, so now it’s called The Owl.^^ That should have said,
Reminiscent looks wise of Danley HRE1!
Damn swype / fingers 🙃
I finally managed to calibrate it. This is my first time doing this, so hopefully it's all correct. I followed some tutorials to learn how to take measurements using REW (quasi-anechoic speaker measurements, Room Correction Tutorial, How to Use Room EQ Wizard).
I also included the Hornresp simulations for comparison (in grey). The mid response is weaker in the lower frequencies than Hornresp suggests. I’m wondering if it’s:
A: Because the mid drivers and rear cabinets might be loose.
B: Because the mid drivers are not completely centered above the mid entry point. Since the mid entry points are so close to the horn throat, the mid drivers are somewhat off-axis to the mid entry port. See 3D renderings here for a visual explanation (link).
Last edited:
So, here’s the mid project frustration rant. I’m starting to think the DAC/DSP is cursed. Nothing has worked as expected, and I’ve spent countless hours, days, weeks troubleshooting various issues.
I was hoping to finish the first version of the active crossover so I could finally listen to the speakers properly. But the DAC/DSP has been giving me endless issues. During measurements, it frequently stalls and does weird things, like changing the channel mapping. So out of the blue, the sub channel end up going to the mid output etc. It wouldn't be fun if the amps are turned up and there are big volume differences between the channels.
I’m not sure if the issue is hardware or software related. I did a fresh install on the Raspberry Pi, but the problem persists. I’ve ordered new I2S jumpers to see if that helps. The previous ones were cheap AliExpress cables, and these are higher-quality silver-plated ones from audiophonics.
On top of that I still haven’t been able to get the other parts of the DAC working, the isolator chip/circuit, the remote, and the trigger. The isolator chip has been especially frustrating; when it’s wired up, the DAC either doesn’t produce sound at all or works with distorted sound, only at max volume. Resoldering everything didn’t solve the issue.
The display and remote script aren’t working yet either. melomane13 has been kind enough to adjust the script for the Raspberry Pi 5, so hopefully, it’ll work in the end.
I’m starting to think if I should just order a Motu Ultralite Mk5 and give up on this cheap DAC idea. But something tells me it’s worth investing a little more time in it.
I was hoping to finish the first version of the active crossover so I could finally listen to the speakers properly. But the DAC/DSP has been giving me endless issues. During measurements, it frequently stalls and does weird things, like changing the channel mapping. So out of the blue, the sub channel end up going to the mid output etc. It wouldn't be fun if the amps are turned up and there are big volume differences between the channels.
I’m not sure if the issue is hardware or software related. I did a fresh install on the Raspberry Pi, but the problem persists. I’ve ordered new I2S jumpers to see if that helps. The previous ones were cheap AliExpress cables, and these are higher-quality silver-plated ones from audiophonics.
On top of that I still haven’t been able to get the other parts of the DAC working, the isolator chip/circuit, the remote, and the trigger. The isolator chip has been especially frustrating; when it’s wired up, the DAC either doesn’t produce sound at all or works with distorted sound, only at max volume. Resoldering everything didn’t solve the issue.
The display and remote script aren’t working yet either. melomane13 has been kind enough to adjust the script for the Raspberry Pi 5, so hopefully, it’ll work in the end.
I’m starting to think if I should just order a Motu Ultralite Mk5 and give up on this cheap DAC idea. But something tells me it’s worth investing a little more time in it.
You really should make a mechanical anchor between the basses - will reduce vibration in the cabinet quite substantial. It might be enough to just stick a pin between them that forces the just slightly apart... but a 2-way fix is probably better...
Nice porject!!
Use CamillaDSP on the Pi5 - it's superb.
//
Nice porject!!
Use CamillaDSP on the Pi5 - it's superb.
//
I feel your pain!So, here’s the mid project frustration rant
I've no experience of Camilla with RPi.
RPi with Volumio streamer was enough for me 🙂
I'm sticking to my older but golder Najda DSP dac solution.
Thank you all! I really appreciate your support. There aren’t many people around me who understand what I’m really doing, so seeing your support keeps me motivated to push through the challenges!
Even though the DAC/DSP isn’t working properly at the moment, I’m actually very happy I went this route. CamillaDSP on a Raspberry Pi is such a capable DSP, and I love the flexibility of being able to use various DACs with it. I’ll keep working on it. If I give up on the DIY DAC, I can always buy a MOTU UltraLite.
Great point! That's something I have thought about after reading the push push thread but haven't really understood how to do it properly. It looks like in some cases there are rods going through the screw holes on the drivers. That should be easy to implement.
In other cases it looks like the magnets (?) are coupled together. That seems more complicated (figuring out how to mount a rod to the magnet without affecting the driver too much) and I can't really wrap my head around how it works if the magnets are moving away from each other when in phase.
Or should I use something to push the back of the sub drivers away from each other (maybe a jack similar to this)?
My feeling is that the rods through the screw holes is the easiest solution. Would love to hear your thoughts on this.
Even though the DAC/DSP isn’t working properly at the moment, I’m actually very happy I went this route. CamillaDSP on a Raspberry Pi is such a capable DSP, and I love the flexibility of being able to use various DACs with it. I’ll keep working on it. If I give up on the DIY DAC, I can always buy a MOTU UltraLite.
You really should make a mechanical anchor between the basses - will reduce vibration in the cabinet quite substantial. It might be enough to just stick a pin between them that forces the just slightly apart... but a 2-way fix is probably better...
Great point! That's something I have thought about after reading the push push thread but haven't really understood how to do it properly. It looks like in some cases there are rods going through the screw holes on the drivers. That should be easy to implement.
In other cases it looks like the magnets (?) are coupled together. That seems more complicated (figuring out how to mount a rod to the magnet without affecting the driver too much) and I can't really wrap my head around how it works if the magnets are moving away from each other when in phase.
Or should I use something to push the back of the sub drivers away from each other (maybe a jack similar to this)?
My feeling is that the rods through the screw holes is the easiest solution. Would love to hear your thoughts on this.
Try a "jack" first? Its a 20/80 solution (get 80% results form 20% effort ;-)) Magnets will move away if in a close box as the cones go inwards this would try to bulge the volume. There is always an opposing force so when cone goes inwards, the rest of the driver will try to go out.
//
//
The dac is finally behaving as it should. It was a simple fix in CamillaDSP, disabling the silencer. I spent way too much time trying to problem solve it myself before asking in the CamillaDSP thread.
Now it's working great and I can finally listen to it properly! So far I'm super happy with the sound, even though I know there are so many potential fixes.
I have been following this workflow for the crossover.
Still not perfect, I need to fix the crossover between the mid and CD and get rid of that notch. And I'd also want to learn more about phase correction and see if I can improve the phase.
Here is a measurement from ca 1m away with XO and eq applied.
Now it's working great and I can finally listen to it properly! So far I'm super happy with the sound, even though I know there are so many potential fixes.
I have been following this workflow for the crossover.
Building a CamillaDSP config for modified K-Horns.
The aim of this post is to show the process I followed to build a config file for CamillaDSP to tri-amp a pair of modified Klipschorns. A secondary aim is to provide enough information for someone to follow the procedure to build a CamillaDSP config for their own speakers.
CamillaDSP is running on a RPi4 and using a Motu Ultralight Mk5 audio interface.
This is not the only way to build a config, the methods and options I have chosen work for me.
This is not going to be another tutorial on using REW and rePhase, but there will be...
The aim of this post is to show the process I followed to build a config file for CamillaDSP to tri-amp a pair of modified Klipschorns. A secondary aim is to provide enough information for someone to follow the procedure to build a CamillaDSP config for their own speakers.
CamillaDSP is running on a RPi4 and using a Motu Ultralight Mk5 audio interface.
This is not the only way to build a config, the methods and options I have chosen work for me.
This is not going to be another tutorial on using REW and rePhase, but there will be...
Still not perfect, I need to fix the crossover between the mid and CD and get rid of that notch. And I'd also want to learn more about phase correction and see if I can improve the phase.
Here is a measurement from ca 1m away with XO and eq applied.
here the version working on rpi 4& 5The display and remote script aren’t working yet either. melomane13 has been kind enough to adjust the script for the Raspberry Pi 5, so hopefully, it’ll work in the end.
i have changed the code relate to GPIO, using lgpio which is compatible with rpi 5 & 4 and TM1637.
script work now as service, was a root rights problem.
from now we can use all type of remote command act as keyboard ( FLIRC, Bluetooth , 2.4g ) , actually is mapped on the proposed bluetooth remote.
install is automatized, SSH to pi , paste this:
touch setup-cdsp.sh && chmod +x setup-cdsp.sh && nano setup-cdsp.sh && ./setup-cdsp.sh
in nano window paste the code attached, save and exit nano;
you get this screen:
choose option 5 and after reboot
or make fresh install and apply all desired options.
Attachments
here the version working on rpi 4& 5
Thanks again melomane1963! Love the new setup menu! I had some issues with it which I messaged you about (Unable to determine board revision from /proc/cpuinfo). Hopefully it'll be working soon.
I have been fine tuning the crossover. I used the guide I mentioned in a previous post. I first EQ the drivers flat, then made linear phase XO in rephase and then adjusted the levels so they match.
Here is the current crossovers:
4th order LR between sub and mid at 150Hz
8th order LR at 1400Hz on the mid
8th order LR at 1100 HZ on the high
I was first using the same 1250Hz XO frequency for both mid and high but got quite a deep notch and didn’t really understand why. The polarity and delay have been adjusted, so that’s wasn't the issue.
It got much better when I moved the mid XO to 1400Hz and the high XO to 1100Hz.
I’m not sure if that’s good practice, but it seems to work.
What I still don't understand is the how and what to do with the phase. Does anyone have good tips regarding phase and FIR filters?
Here is the current measurement with no phase adjustments.
Should I be aiming for a specific target phase response or just try to get it flat?
Here is the current crossovers:
4th order LR between sub and mid at 150Hz
8th order LR at 1400Hz on the mid
8th order LR at 1100 HZ on the high
I was first using the same 1250Hz XO frequency for both mid and high but got quite a deep notch and didn’t really understand why. The polarity and delay have been adjusted, so that’s wasn't the issue.
It got much better when I moved the mid XO to 1400Hz and the high XO to 1100Hz.
I’m not sure if that’s good practice, but it seems to work.
What I still don't understand is the how and what to do with the phase. Does anyone have good tips regarding phase and FIR filters?
Here is the current measurement with no phase adjustments.
Should I be aiming for a specific target phase response or just try to get it flat?
Don't be afraid to show the RAW FR, we've all seen them before.........................................................................................................................
Incredible progress in such a short period of time m-a!
I was going to ask if you had dabbled with ATH to create the horn but recently you'd mentioned using Fusion360. Was the learning curve steep to make mounting surfaces / port holes for the 4" drivers? I'm diving into my own 3 way & wondering about how you tackled this step!
Did you have experience with Fusion before starting this project?
I was going to ask if you had dabbled with ATH to create the horn but recently you'd mentioned using Fusion360. Was the learning curve steep to make mounting surfaces / port holes for the 4" drivers? I'm diving into my own 3 way & wondering about how you tackled this step!
Did you have experience with Fusion before starting this project?
Haha! I allready posted them earlier in this thread, but realised that the mid response was inacurate in the lower part of the response. Here is an updated RAW response along with the Hornresp estimations for the sub and mid drivers as well as HF108 response from the datasheet (in grey)Don't be afraid to show the RAW FR, we've all seen them before.........................................................................................................................
Incredible progress in such a short period of time m-a!
I was going to ask if you had dabbled with ATH to create the horn but recently you'd mentioned using Fusion360. Was the learning curve steep to make mounting surfaces / port holes for the 4" drivers? I'm diving into my own 3 way & wondering about how you tackled this step!
Did you have experience with Fusion before starting this project?
Thank you!
I haven’t used ATH and Akabak yet, but I’d like to learn to use them for future projects. This is my first time designing, building, and measuring a speaker, so there have been so many things I’ve had to learn... and it's taken a lot of time! Over the past few days, I’ve been diving into PCB design in EasyEDA to create a PCB for the DAC/DSP, more on that in a future post.
I didn’t have any experience with Fusion 360, just learned by watching tutorials and experimenting. It was a bit confusing at first, but once I understood the basics, it became fairly straightforward.
When I was designing in Fusion 360, I was constantly switching between calculating parameters in the Synergy Calc spreadsheet, simulating the frequency response in Hornresp and manually inputting the size parameters into Fusion 360.
Eventually, I got tired of all these manual transfers, so I integrated the calculations from the spreadsheet directly into the Fusion360 file. Thanks to Fusion 360’s parametric features, I can now easily adjust the size and coverage angle of the 3D model just by modifying input values. It's like the the spreadsheet, but the model updates right away.
Since the design isn’t finished yet and my Fusion 360 file is still a bit messy, I probably won’t release it publicly until I’m happy with the final version. However, if anyone wants to play around with the current file, feel free to message me, and I’ll send over a link. I’d love to get feedback, and you’re welcome to modify and use it as you like.
- Home
- Loudspeakers
- Multi-Way
- 2-Way MEH build