VituixCAD's Optimizer does this with several choices of what you intend to "optimize".
Paul's crossovers are designed to get the best out of his chosen drivers, get the sound he wants and deal with any driver issues such as resonances and peaks. Sometimes, with his two way designs, his crossovers are relatively simple.
As I understand it, 3 way crossovers are much harder to design properly than two ways because you have to reduce the high and low frequencies going to the mid range driver, so you need more parts. The mid range has to integrate perfectly with the woofer and tweeter otherwise the speaker won't sound right.
Geoff
As I understand it, 3 way crossovers are much harder to design properly than two ways because you have to reduce the high and low frequencies going to the mid range driver, so you need more parts. The mid range has to integrate perfectly with the woofer and tweeter otherwise the speaker won't sound right.
Geoff
So, if you have one more question regarding midrange filters...
If, for example, I use a 1.6mH inductor for the bass, should I use exactly the same in the midrange filter?
Also, the capacitor I use in the tweeter should I use the same in the midrange filter? the famous LC
You can answer your own question by using a 1st order, 3-way crossover calculator.
https://goodcalculators.com/crossover-calculator/
This is a the result of inputting 8 ohm drivers, an fL of 500 Hz and an fH of 4,000 Hz.
EDIT: It won't suit real world drivers very well though as they do not necessarily measure 8 ohm at the crossover frequencies as Dave Bullet explains below.
https://goodcalculators.com/crossover-calculator/
This is a the result of inputting 8 ohm drivers, an fL of 500 Hz and an fH of 4,000 Hz.
EDIT: It won't suit real world drivers very well though as they do not necessarily measure 8 ohm at the crossover frequencies as Dave Bullet explains below.
Last edited:
Unlikely.
The on-line calculators take a too simplistic view. They assume both drivers have the same impedance at the crossover frequency, and are the same sound SPL output (sensitivity) and they have the same frequency response (that is neither driver is already tailing off).
Often the woofer inductor will be larger in any case, as it is used to manage baffle step compensation.
Basically - you need to start with as I said, the acoustic responses of the raw drivers, and what parts are added to the crossover to shape that drivers response to align to the intended acoustic rolloff and crossover points you want. then you add more parts (or change topology) to improve the phase alignment between the drivers, checking overall impedance as you go to make sure you don't cause any nasty minimum impedance issues.
I find crossover design to be an interative process.
Ideally it would come down to one's lobing choices.. However when there are less known factors in a non-precision design, such as driver responses, the fact that on average 1st order is half way between one polarity and the other, and not wanting to go over 90 degrees... On the other hand of course, once one goes beyond the slope rate of the other it's anyone's game, and that's easy to do.
But using one does answer Pedroga's question. He's on a steep learning curve when it comes to crossovers!
Last edited:
Yes it does answer his question, however I'm trying to steer him away from on-line calculators. There may be examples - most likely 2 ways with carefully matched drivers where on-line calculators provide an acceptable response.
With this 3 way planned, Pedroga needs to throw these calculators away and start with a ground up design using actual driver measurements. If he accepts this approach , we can help him with either existing or self measurements.
With this 3 way planned, Pedroga needs to throw these calculators away and start with a ground up design using actual driver measurements. If he accepts this approach , we can help him with either existing or self measurements.
@Pedroga are you just exploring crossover concepts? or intending to build a speaker?
If you are looking at concepts, I'd recommend learning about the following topics:
If you are looking at concepts, I'd recommend learning about the following topics:
- Various "Stock" filter types and orders (butterworth, LR etc...)
- Driver spacing / placement on baffle and effect on lobing given "stock" filters
- Driver spacing / placement on baffle and effect on diffraction
- Baffle shape / profile and effect on diffraction
- Baffle step and compensation circuits
- Impedance compensation circuits
- Resonant peak circuits
- Notch filters
- Ladder delay networks
- Parallel vs. series topologies
- Cascaded 1/2 way filters
"If, for example, I use a 1.6mH inductor for the bass, should I use exactly the same in the midrange filter?
Also, the capacitor I use in the tweeter should I use the same in the midrange filter?"
I can't think of an example I've looked at, either two or three way, which does that: for example, look at Paul's crossovers for his Tarkus, Pit Vipers and Dayton Classic Revivals. Have a look at the values of the parts: no way would a calculator come up with anything as good.
I recommend getting Xsim or Vituix, and get some FRD and ZMA files from, say Dayton Audio, and start learning how to use them, it's all free. Load some drivers into the program, then experiment with how parts values shape the frequency response and impedance. This will not give you an accurate crossover, as the drivers aren't measured in your cabinet. But using Xsim is a good way to get a feel for what different parts and values do.
NB I'm just using his designs as examples, there are many great designers and websites out there, but Paul puts up info about driver choices, crossover thinking and cabinet design. Curt Campbell at Speaker Design Works does the same, I just haven't referenced his excellent designs so as not to confuse things.
Geoff
Also, the capacitor I use in the tweeter should I use the same in the midrange filter?"
I can't think of an example I've looked at, either two or three way, which does that: for example, look at Paul's crossovers for his Tarkus, Pit Vipers and Dayton Classic Revivals. Have a look at the values of the parts: no way would a calculator come up with anything as good.
I recommend getting Xsim or Vituix, and get some FRD and ZMA files from, say Dayton Audio, and start learning how to use them, it's all free. Load some drivers into the program, then experiment with how parts values shape the frequency response and impedance. This will not give you an accurate crossover, as the drivers aren't measured in your cabinet. But using Xsim is a good way to get a feel for what different parts and values do.
NB I'm just using his designs as examples, there are many great designers and websites out there, but Paul puts up info about driver choices, crossover thinking and cabinet design. Curt Campbell at Speaker Design Works does the same, I just haven't referenced his excellent designs so as not to confuse things.
Geoff
To start you off, I've chosen some drivers I've worked with. I recommend starting with a two way design.
https://www.daytonaudio.com/product/29/dc28f-8-1-1-8-silk-dome-tweeter-8-ohm
https://www.daytonaudio.com/product/1206/rs180p-8-7-reference-paper-woofer-8-ohm
You'll find the FRD (frequency response) and ZMA (impedance) files under the "Data Files" heading on the product page, when you click, they'll download and you can use in Xsim or Vituix.
Many manufacturers only provide graphs, rather than the FRD and ZMA files, so you have to generate those with a graph tracing program, which makes the process more complex and you don't want that just yet.
https://xsim.software.informer.com/
Geoff
https://www.daytonaudio.com/product/29/dc28f-8-1-1-8-silk-dome-tweeter-8-ohm
https://www.daytonaudio.com/product/1206/rs180p-8-7-reference-paper-woofer-8-ohm
You'll find the FRD (frequency response) and ZMA (impedance) files under the "Data Files" heading on the product page, when you click, they'll download and you can use in Xsim or Vituix.
Many manufacturers only provide graphs, rather than the FRD and ZMA files, so you have to generate those with a graph tracing program, which makes the process more complex and you don't want that just yet.
https://xsim.software.informer.com/
Geoff
Last edited:
First, you may benefit from looking at the basic operation of a 3-way loudspeaker's band pass filter:
The series LC filter on the midrange is called a 'band pass" filter because it only passes a central band of frequencies.
In the above circuit, L1 opposes high frequencies while C2 opposes low frequencies, allowing only a band of central frequencies to reach the midrange driver.
The effects of the low pass, high pass and band pass filter sections are shown below. Note that the frequency cuts are not sudden, but slope off gradually.
Let me know if information like the above is helpful in aiding your understanding. If it is not, I will cease supplying it.
I read about bandpass, the issue is that I can't reason or find something that teaches me how to find the values for the LC circuit.
In my head, I would remove the treble and bass that went to the other speakers, leaving only the midrange for the midrange, so I would use the same capacitors and inductors that I used in the other speakers and use it in the midrange.
In my head, I would remove the treble and bass that went to the other speakers, leaving only the midrange for the midrange, so I would use the same capacitors and inductors that I used in the other speakers and use it in the midrange.
For the tweeter and subwoofer, there are even tables on the internet that show the usual filters for each type of frequency desired.
L-pads too and everything.
But when it comes to the midranges it's like a mystery, there aren't any of those things (at least I've never seen them)
Yes, what you are looking for does not appear to be out there, at least in simple terms.
The mathematics is obviously difficult and that's why people use a computer program to do the calculations.
I can only offer this calculator (shock horror!) for you to play with: https://www.claredot.net/en/sec-Sound/band-pass-cross-over-6dB.php
It is - just like everything you start to dig into!
Once you know just a little more it's getting even more mysterious.
But once you start using simulation tools it will get much more intuitive!