Greetings!
I started to tinker with loudspeaker design and kind of set myself to build a two way around an exsisting enclousore, because i already have wood cut out form something
left over. Its plywood 10 mm thick and the plate dimensions are as follows:
300x300 mm
300x600 mm
300x550 mm, which results in about 42 liters or about 1,48 ft^3 of internal space.
I read a few books on loudspeaker design and started with WINISD and proceeded with VituixCAD as it seems like a wonderful tool.
I decided to go two way, as i plan to use the speakers for general listening and would like the frequency response to be as flat as possible. Bass reflex resonator cabinet with passive crossover design.
After some research and of course tinkering with simulation software, i decided to go with the folowing woofer (of course if the decision is stupid feel free to yell in my face 🙂)
Dayton Audio RS225-8, 8" Reference series woofer 8 ohm.
With the enclosure simluator and enclosure tuned to about 30 Hz i get good SPL response, port dimensions are so at 50 W RMS i do not exced Xmax nor port velocity 17 m/s as i read that is audible. The lower part of freqency response is not peaked and gently tapered down so i think the driver fits well enough inside my enclosure.
Then i started simulating and of course i dont have measurements as i did not buy the drivers yet, but i have ZMA and FRD from the manufacturer. Now, as I understand there is "baffle step" loss, when the driver stops radiating from half space to full space and this needs to be taken into account.
My steps in VituixCAD have been:
0. Check in enclosure tool if i am "happy" with the woofer in my cabinet
1. Name driver
2. Open Diffraction tool and simulate diffraction on front baffle, mic position 3 m away, and in the middle of baffle, import halfspace response and i select 0 deg frd file from Dayton Audio and check "full space" and export this as "Diffraction woofer".
3. I load frequency responses of step 2 for particular driver
4. I load impedance response of a particular driver
5. Repeat process for tweeter
6. Start designing Crossover
Now as i understand with this i am able to simulate diffraction and come just a little closer to what it would actually be like. Correct me if i am wrong in all this.
For the tweeter i initially tried to go with: Dayton Audio DC28-F which is 8 ohms and coincides with the woofer, but when simulating i found that it was about 10 dB lower than the woofer if i take diffraction into acount.
I then went with Dayton Audio RST28F-4, which is 4 ohms so that is not perfect but i gotta start somehow. With this tweeter i get a better resoult at least IMO. Now
i now very little about xover design i know orders and how to arange parts and i know which orders should be better regarding phase so i know "something".
Also whats up with driver position on the baffle board? Is this relative to listening position? If i imagine head height at tweeters then i need to ofset woofers by Y value?
I will put screenshots of simulation in thread. I would kindly ask all you qualified and skilled designers how many things i got wrong, and i would like to know how close are these simulations to what happens when you actually build the thing?
When i decide to buy drivers i can do nearfield measurement in an anechoic chamber at friend of mine and can then with those FRD files design proper crossover. I think that simulating before buying would at least give me "some kind of positive toughts" on driver picking.
Thanks in advance for your time.
I started to tinker with loudspeaker design and kind of set myself to build a two way around an exsisting enclousore, because i already have wood cut out form something
left over. Its plywood 10 mm thick and the plate dimensions are as follows:
300x300 mm
300x600 mm
300x550 mm, which results in about 42 liters or about 1,48 ft^3 of internal space.
I read a few books on loudspeaker design and started with WINISD and proceeded with VituixCAD as it seems like a wonderful tool.
I decided to go two way, as i plan to use the speakers for general listening and would like the frequency response to be as flat as possible. Bass reflex resonator cabinet with passive crossover design.
After some research and of course tinkering with simulation software, i decided to go with the folowing woofer (of course if the decision is stupid feel free to yell in my face 🙂)
Dayton Audio RS225-8, 8" Reference series woofer 8 ohm.
With the enclosure simluator and enclosure tuned to about 30 Hz i get good SPL response, port dimensions are so at 50 W RMS i do not exced Xmax nor port velocity 17 m/s as i read that is audible. The lower part of freqency response is not peaked and gently tapered down so i think the driver fits well enough inside my enclosure.
Then i started simulating and of course i dont have measurements as i did not buy the drivers yet, but i have ZMA and FRD from the manufacturer. Now, as I understand there is "baffle step" loss, when the driver stops radiating from half space to full space and this needs to be taken into account.
My steps in VituixCAD have been:
0. Check in enclosure tool if i am "happy" with the woofer in my cabinet
1. Name driver
2. Open Diffraction tool and simulate diffraction on front baffle, mic position 3 m away, and in the middle of baffle, import halfspace response and i select 0 deg frd file from Dayton Audio and check "full space" and export this as "Diffraction woofer".
3. I load frequency responses of step 2 for particular driver
4. I load impedance response of a particular driver
5. Repeat process for tweeter
6. Start designing Crossover
Now as i understand with this i am able to simulate diffraction and come just a little closer to what it would actually be like. Correct me if i am wrong in all this.
For the tweeter i initially tried to go with: Dayton Audio DC28-F which is 8 ohms and coincides with the woofer, but when simulating i found that it was about 10 dB lower than the woofer if i take diffraction into acount.
I then went with Dayton Audio RST28F-4, which is 4 ohms so that is not perfect but i gotta start somehow. With this tweeter i get a better resoult at least IMO. Now
i now very little about xover design i know orders and how to arange parts and i know which orders should be better regarding phase so i know "something".
Also whats up with driver position on the baffle board? Is this relative to listening position? If i imagine head height at tweeters then i need to ofset woofers by Y value?
I will put screenshots of simulation in thread. I would kindly ask all you qualified and skilled designers how many things i got wrong, and i would like to know how close are these simulations to what happens when you actually build the thing?
When i decide to buy drivers i can do nearfield measurement in an anechoic chamber at friend of mine and can then with those FRD files design proper crossover. I think that simulating before buying would at least give me "some kind of positive toughts" on driver picking.
Thanks in advance for your time.
And to add this:
This is an example of a simulation with the previously mentioned Dayton Audio DC28-F, which in theory should play louder than woofer because of higher sensitivity. When i simulate without baffle step i get first picture, but when i do the proces described in previous post and include baffle the tweeter is so quiet 🙂. Either this is normal or I am doing something wrong.
This is an example of a simulation with the previously mentioned Dayton Audio DC28-F, which in theory should play louder than woofer because of higher sensitivity. When i simulate without baffle step i get first picture, but when i do the proces described in previous post and include baffle the tweeter is so quiet 🙂. Either this is normal or I am doing something wrong.
You're doing something wrong. Almost any tweeter is sensitive enough for this woofer.
Last edited:
To get started, forget the computer estimation for baffle step.. Put a 2mH, or larger coil on the woofer, until the response at 1k has decreased by 6dB. Almost everything I build seems to need a 2.5mH. Now shape the roll-off smoothly to be down 6dB more at your desired x-over frequency. Now match the tweeter to that response at approximately the 1k SPL level. .
Last edited:
The DCF28F-8 is approximately as sensitive as the woofer. When you put them in a box, the treble should stay near the same and the woofer lower frequencies should be 6dB less than before. You want to bring the woofer upper and treble down by 6dB as well.
If you can't see that then I'd be reviewing your steps in the baffle simulator.
If you can't see that then I'd be reviewing your steps in the baffle simulator.
I just looked at your second sim. Woofer looks good. Pad the tweeter down to the 85dB line, and you will be doing what I just suggested. Baffle losses will bring the bass down to the 85dB line as well.
Thank you for tips. So as i understand its best to buy drivers, make the box, mount everything and mesaure nearfield response of woofer and tweeter in anechoic chamber (we have one near me) and then use vituix for crossover.
I would like to simulate as much as possible so i wont buy wrong drivers and be stuck with them. I would like to get as flat as possible response.
Is baffle step simulation without nearfield measurements in anechoic chamber or gated measurements and then using baffle step tool really that bad idea?
How is this woofer /tweeter combo? You think i could get approx. flat response out of cabinet of said size?
Thank you for your time.
I would like to simulate as much as possible so i wont buy wrong drivers and be stuck with them. I would like to get as flat as possible response.
Is baffle step simulation without nearfield measurements in anechoic chamber or gated measurements and then using baffle step tool really that bad idea?
How is this woofer /tweeter combo? You think i could get approx. flat response out of cabinet of said size?
Thank you for your time.
That is i presume the one where i have not considered baffle loss. In that sim i used frequency response files of manufacturer 0 to 60 degres i think and the impedance file for both drivers.
So from what i see, there is something quite horibly wrong in my steps when using baffle tool, as the tweeter gets so low. 🙂
Or i should rather say the woofer gets high. Tweeter stays at 85 dB line. So my baffle step sim is actualy "raising" my woofer almost to 100 dB. When i look at it now, i see its the oposite that should happen. Yikes to me...
I suspect you may be using the word "nearfield" improperly. You don't want nearfield measurements in an anechoic chamber.
You can sim more accurately but I suspect that the end resulting x-over is very close to what you come up with the other way.
I think that you can save your response from your sim as an frd, and then import it into a baffle simulation. If so, you can play with location, and dimensions.
When you actually build a box and measure you measure from a meter or so using gated results. Then add the x-over you came up with in the sim and measure it. Make changes as needed. You can make measurements at home. No need for the chamber.
An 8" 2-way is a challenge. The off axis is not as smooth. Some think that matters. Some not so much..
What is your room setup like.? Me, I use a sub. That changes the speaker requirements greatly.
What is your room setup like.? Me, I use a sub. That changes the speaker requirements greatly.
I generally have 2 speakers (stereo) and an amp. Speakers slightly of the wall, tweeters ear height. I would like the freq response to be as flat as possible so i can then adjust eq to taste. The funny thing is that always no matter what i listen to i seem to end up with a "smiley face" eq - audiophiles will say no my ears say yes!
If you think these to drivers are a good combination i will simulate further tomorrow and try to get baffle step sorted. Will update how it goes. Again thank everybody for their valuable time.
If you think these to drivers are a good combination i will simulate further tomorrow and try to get baffle step sorted. Will update how it goes. Again thank everybody for their valuable time.
Because your taste are a little bit spoiled and ruined by general trends, car audio and so on. There should be smiley face, but with the right side of smile down like this :/
When you will have all the data for simulation blue line on CTA-2034 has to be like that down smile. Also there is Preference Rating tool on Vituix which more or less is good starting point
You cannot simulate something similar to real world without measurements. Totally different results. You need 360 (180 if speaker is symmetrical) degrees measurements on at least hor plane.
For the overall design you need a tweeter which can go low or the woofer that can go high. Or both. Directivity is a thing, because of the real performance of speaker in real room in real conditions.
So, some of the things to think about:
1. Go for smaller RS180?
2. Go for pretty deep waveguide for tweeter. In this way you may win some 300-500Hz of range for your tweeter. There are threads about waveguides. That may work
3. Go for bigger tweeter? 30-35mm? You will loose 15+kHz, but that may be worth it. They are expensive. Still probably will need some waveguide.
4. Go for other tweeter, which can go full 20kHz, but is confirmed to work on ridiculously low freq. Don't know any. Still will need waveguide.
5. Go for small fullrange instead of tweeter. There is a thread for exactly project with RS225 + some various 2,5-3 inch drivers. https://www.diyaudio.com/community/threads/10f-8424-rs225-8-fast-waw-ref-monitor.273524/
6. Go for paper, propylene or other than metal cone woofer. If you are still stubborn to go with RS225 - that thing you will hear past 2kHz from woofer will be not music, it is breakup modes. If you cross it high even with some huge and not necessary slopes you will still hear them. RS225P may be a better option. I cannot comment on RS225 metal or paper, but I have RS100 metal and its high end is on the ugly side. Most if not all metal cones for my ears are ugly on high freq, but YMMV
Pretty bad. You seem obsessed with BSC, but you apparently don't understand it, and are doing it wrong. I can explain it to you.
Last edited:
Can you show the raw woofer response on your sim? Forget the x-over. Forget the tweeter. Forget all the other charrts. Just show the woofer measurement that you are using from Parts Express data.
Will do. First image is with simulated baffle step in intended enclosure, second without, just imported from parts express.
First I apologize. You appear to be on track as far as baffle step goes. What is the size of your baffle? The added effect of the baffle looks far worse than what I usually see with my projects. It may be that your baffle could be greatly improved.
It's OK, if you can't, but getting the other graphs out of the thread will make this simpler. The SPL graph on the top left is the only thing to look at at the moment.
It's OK, if you can't, but getting the other graphs out of the thread will make this simpler. The SPL graph on the top left is the only thing to look at at the moment.
Last edited:
Of course, feel free to corect me in everything yell in my face if you have to 🙂.
My front baffle is 300 x 600 mm (width x height) that is 11,81 x 23,62 inches roughly. The thing is i have these boards to build speakers around. Is baffle step really that drastic? Its possible there si something wrong in my sim process..
My front baffle is 300 x 600 mm (width x height) that is 11,81 x 23,62 inches roughly. The thing is i have these boards to build speakers around. Is baffle step really that drastic? Its possible there si something wrong in my sim process..