Big 2-way Speaker guidance

Hello,

I've started sketching on a new speaker project and looking for some initial guidance. This will be the first time I design, design a crossover, etc as I dont plan on following any set design for this project.

The design requirements I've been following:

  • 2x 8" woofers
  • 1 tweeter
  • I want a big box, so around 100-140L (picture below)

I've been looking at these woofers:
HiVi M8N-1B 8" Damped Aluminum/Magnesium Cone Woofer 45 euro (if buying 4)
Maybe the HiVi-M8N-1 version though 🙂
https://www.soundimports.eu/en/hivi-m8n-1b.html
https://doc.soundimports.nl/pdf/brands/HiVi/M8N-B/M8N-B-manual.pdf

or

Dayton Audio DS215-8 8" Designer Series Woofer Speaker - 72 Euro (if buying 4)
https://www.soundimports.eu/en/dayton-audio-ds215-8.html

And I'm leaning towards the HiVi M8N, they look nicer to me and are a bit cheaper.

For tweeter I'm looking at:

SB acoustics SB26STWGC-4 85 Euro
https://www.soundimports.eu/en/sb-acoustics-sb26stwgc-4.html

Looking at the crossover, based on the very little experience of designing this myself, I'd assume it would be somewhere around ~1500Hz. I have bought a calibrated mic, so I will measure and simulate before I actually buy components for crossover.

I don't have a hard budget, but aiming for like 350-400 tops. This current setup would be 350 euro. I was also considering building a 2.5 or 3way, but seeing how it's my first time doing crossovers and things myself maybe its better to limit myself to a 2-way?

But I have a few questions:

  • Does anything stick out as unreasonable?
  • I get the feeling that 100-140L is big for WinISD, but since I have two woofers maybe that is ok? WinISD didn't seem to change that much depending if I selected 1-2 drivers.
  • Comparing the two drivers in WinISD, HiVi seems to perform better with lower freqs despite being 50% cheaper. Does this make sense? How much should I trust WinISD?
  • Should I do some optimizations to make front baffle more narrow? Currently I have it maybe 1-2 cm wider than the woofer element so it has some "breathing room" inside.
  • Any issues with the design of enclosure other than it will be more difficult building than a normal box due to angles?

Picture is of inner dimensions of the enclosure. Will add 15-18mm of material.

1743321754265.png
1743322363359.png


Just pic of how the two different drivers look in WinISD. HiVi 140L, RS225-P around 120L.

1743321967801.png


Thanks 🙂
 
It is for these woofers. It may be fine though, just more than you need.


..or more wide, depends what you're into.
Ok, good to hear 🙂.

I think I'm too new to know how "depends what you're into" would affect the sound. But I guess I will notice once I start playing around with baffle edge diffraction simulator.


If the cabinet is too big you can quite easily baffle the box to reduce the internal volume used by your woofers
How would you determine if the cabinet is too big? Is it after listening, or looking at graphs?
Because to me I can alter the volume of the enclosure by +/- 40L in WinISD and the graph changes but not by a big margin.

Just as an example this is a double 8" woofer 2.5 way project:

https://www.donhighend.de/?page_id=8971
Ok 🙂 I'll take a look, thanks.
 
How would you determine if the cabinet is too big? Is it after listening, or looking at graphs?
Because to me I can alter the volume of the enclosure by +/- 40L in WinISD and the graph changes but not by a big margin.
Hi!

I usually create several box simulations with different volumes (properly tuning each one) and use the same speaker.
Then, I compare the frequency response. See below - same speaker in 40 liters, 70 liters and 120liters.

1743367435545.png
 
Look at the SPL graphs of the M8N and the DS215. The M8N has about a 10dB peak at 2 kHz where the cone breakup begins, whereas the DS215 has about a 7-8 dB peak around 2.4 kHz. The DS215 is going to be easier to work with.

Take a look at the Dayton Audio SIG225-4 also. (Although it is 4 ohms, if that matters to you.)
 
  • Like
Reactions: Inductor and stv
Hi!

I usually create several box simulations with different volumes (properly tuning each one) and use the same speaker.
Then, I compare the frequency response. See below - same speaker in 40 liters, 70 liters and 120liters.
Yeah, I did that for the dayton at first at first, but I removed them, I felt like I could eyeball it better haha. And I didn't really see how the low volume boxes would be better with WinISDs automatic optimization. Even though it says it optimizes for a flat line, to me it felt like optimizing for getting low F3 is better regardless if the line got a tiny bit away from 0 above 50Hz.


Look at the SPL graphs of the M8N and the DS215. The M8N has about a 10dB peak at 2 kHz where the cone breakup begins, whereas the DS215 has about a 7-8 dB peak around 2.4 kHz. The DS215 is going to be easier to work with.

Take a look at the Dayton Audio SIG225-4 also. (Although it is 4 ohms, if that matters to you.)
I was thinking to do crossover around 1.5Hz. before that happens, or do you mean that it would still influence it more than I'd like?
And I was leaning for not buying Dayton Audio drivers, because I read another thread and it felt like everyone was bashing them for being bad, and that the page were just pitching them because it was their own brand.
 
Yeah, I did that for the dayton at first at first, but I removed them, I felt like I could eyeball it better haha. And I didn't really see how the low volume boxes would be better with WinISDs automatic optimization. Even though it says it optimizes for a flat line, to me it felt like optimizing for getting low F3 is better regardless if the line got a tiny bit away from 0 above 50Hz.
Ok! I disregard all automatic optimization. I do my own tuning, since I always want the tuning frequency to be way below the normal music range.
I always tune below 30Hz. In addition, for small volumes we need to consider the length of the port, maximum air velocity in the music range etc.
I prefer a "side effect" of port gain instead of its full gain. A kind of "last mile" help to have good frequency response below 30Hz.
 
  • Like
Reactions: Freedom666
...I was thinking to do crossover around 1.5Hz. before that happens, or do you mean that it would still influence it more than I'd like?

If you crossover at 1.5kHz then it will be extremely hard (maybe impossible) to bring 10 dB peaks of cone breakup at 2 kHZ down to an acceptable level.

Look at the SPL graphs of the MN8 and the SIG225 and the difference will be clear.

And I was leaning for not buying Dayton Audio drivers, because I read another thread and it felt like everyone was bashing them for being bad, and that the page were just pitching them because it was their own brand.
Dayton Audio drivers are a great value in the US. Some of the drivers are better than others and this is pretty well indicated in the price. The Reference Series drivers are very good for the money. However, they are not as inexpensive relative to SB Acoustics and even Scanspeak in Europe. In the US the Dayton drivers are significantly less expensive.
 
Ok! I disregard all automatic optimization. I do my own tuning, since I always want the tuning frequency to be way below the normal music range.
I always tune below 30Hz. In addition, for small volumes we need to consider the length of the port, maximum air velocity in the music range etc.
I prefer a "side effect" of port gain instead of its full gain. A kind of "last mile" help to have good frequency response below 30Hz.
Ok, good to hear someone else also moves it graph around 🙂
Am I understanding you correctly that you're intentionally aim for a bit of a bump in the low end like your picture of the black line?
 
If you crossover at 1.5kHz then it will be extremely hard (maybe impossible) to bring 10 dB peaks of cone breakup at 2 kHZ down to an acceptable level.
Ah okay. If I look at the tweeter I have been eye:ing at, I could potentially start crossover around as low as 1000Hz? And avoid the 2k peak? I guess I for some reason wanted to give the woofers more of the frequency spectrum, but I might as well just let the tweeter do the job?

https://doc.soundimports.nl/pdf/brands/SB Acoustics/SB26STWGC-4/SB26STWGC-4 Prototyperev-1.pdf
 
I think crossing over a woofer at 1.5kHz is better than crossing a tweeter at 1kHz.

I'm just suggesting that the M8N is not a good choice for an 8 inch 2-way. I have nothing against the M8N, I have a pair but I intend to use them in a 3-way.

The DS215 or SIG225 are better choices than the M8N.
 
  • Like
Reactions: stv
Given the aspect ratio of the enclosure, simukating with WinISD will nnottell you btheb story, since the box will be an ML-TL and WinISD knows nothing about them.

And given the inevitable driver centre-to-centre i’d want to bring the XO point way down. 250 Hz woud be good, 500 sorta OK.

dave
 
Ok, good to hear someone else also moves it graph around 🙂
Am I understanding you correctly that you're intentionally aim for a bit of a bump in the low end like your picture of the black line?
Hi!

As the box gets too big related to the speaker T/S parameters, we see this behavior. The rolloff starts and if we try to tune to hold it flat, it's not possible. If we tune that black curve a bit above (30Hz and up), you will see a real bump above 0dB which can make the response boomy. I don't like it.

But tuning at 27Hz, in this case, there is just a 1dB drop around 50Hz and the FR returns to 0dB at 30Hz.
We can say that we have flat response down to 30Hz which is good.
In addition, we cross -3dB at 25Hz which is super cool.

If we use automatic WinISD optmization for this speaker, the volume is 66 liters cause it tries to make the curve super flat (nothing above, nothing below) with just a roll-off. But, with 120-liter it gets better bass extension below 30Hz. We get 5dB more at 25Hz and 3dB more at 30Hz.

This is the kind of analysis I do, instead of using the auto optmization.

If I had room for 120liter, I'd use this option. I'm using the speaker I'm working with just an example.
I'm planning a 40-liter studio loudspeaker and another version of 100-liter floor standing with this 10" woofer.
 
I think crossing over a woofer at 1.5kHz is better than crossing a tweeter at 1kHz.

I'm just suggesting that the M8N is not a good choice for an 8 inch 2-way. I have nothing against the M8N, I have a pair but I intend to use them in a 3-way.

The DS215 or SIG225 are better choices than the M8N.
Okay. What I guess I don't understand just looking at the SPL graphs is why it would be an issue to crossover a tweeter down at 1k. Assuming I can create a crossover that starts and ends where both woofer and tweeter are flat.

Yeah, point taken about them, but SIG225 is like 2x the price. And while I guess one would be enough to offset the price, I just really like the look of 2 woofers haha (probably at the cost of sound quality 😛).


Given the aspect ratio of the enclosure, simukating with WinISD will nnottell you btheb story, since the box will be an ML-TL and WinISD knows nothing about them.

And given the inevitable driver centre-to-centre i’d want to bring the XO point way down. 250 Hz woud be good, 500 sorta OK.

dave
I planned on not going for a ML-TL, just have a port vent in the back.

But I assume there are basically zero tweeters that would do 250 or even 500? So I would need to have another midrange driver then, and make it a 3-way?
 
I'm a starting to consider adding a mid woofer since everyone makes the case that it would be easier for the crossover. Would I need to create a separate enclosure in the box if I add a mid range driver? Assuming I would have woofer woofer mid tweeter.

If I added for instance: https://www.soundimports.eu/en/sb-acoustics-sb15mfc30-4.html

DriverFromTo
2x HiVi-M8N-120300
SB Acoustics SB15MFC30-43002-3k
SB Acoustics SB26STWGC-42-3k 20k
 
  • Like
Reactions: Arez