I agree, that's a good, easy place to start learning.
What you'll need to know:
Impedance of Tweeter at Xover freq
Impedance of Woofer at Xover freq
an online calculator:
http://ccs.exl.info/calc_cr.html#second
But first read about choosing xover frequency, there's a good article or 2 about on the web
Pete McK
What you'll need to know:
Impedance of Tweeter at Xover freq
Impedance of Woofer at Xover freq
an online calculator:
http://ccs.exl.info/calc_cr.html#second
But first read about choosing xover frequency, there's a good article or 2 about on the web
Pete McK
A way to get some magic fast is..
Take a good 6-8"wideranger of good rep & one that will run OB down to 100hz or so, then go active with with the bass augmentation to 30-40hz. stick a decent ribbon or what ever way up high...& maybe a notch filter if you have a nasty peak were you don't want it. I in 6 months of DIY speakers have beat all my commercial stuff & only keep one set of good monitors for reference. They collect nothing but dust now.Basicaly a 3 way with 95% covered with the wideranger.. 1- 3 part notch & a single .22uf cap on the ribbon is all I use. It is very, very good. .Ugly ,......but I am only after the sonics. I think I will let my daughter paint the baffles.Study till your eyes bleed & do it again . DON'T buy expensive wire & ****.....there is much more imporatnt things to address. getting the commercial restraints off me was the best thing I ever did in this hobby, bar none.
Take a good 6-8"wideranger of good rep & one that will run OB down to 100hz or so, then go active with with the bass augmentation to 30-40hz. stick a decent ribbon or what ever way up high...& maybe a notch filter if you have a nasty peak were you don't want it. I in 6 months of DIY speakers have beat all my commercial stuff & only keep one set of good monitors for reference. They collect nothing but dust now.Basicaly a 3 way with 95% covered with the wideranger.. 1- 3 part notch & a single .22uf cap on the ribbon is all I use. It is very, very good. .Ugly ,......but I am only after the sonics. I think I will let my daughter paint the baffles.Study till your eyes bleed & do it again . DON'T buy expensive wire & ****.....there is much more imporatnt things to address. getting the commercial restraints off me was the best thing I ever did in this hobby, bar none.
legendaryfrog said:
If you want to truly "design" a crossover you need to measure and simulate your drivers with various components.
In lieu of measuring, you can "scan" a published frequency response and impedance using a program called SPL Trace. Then you would place the measurements in a crossover design program like the excellent and free Passive Crossover Designer (actually a very sophisticated spreadsheet for Excel) and adjust the capacitor/coil/resistor values until you arrive at a good combined response. Then you build the crossover and listen to it. If something is not quite right you tweak a couple of parts until it sounds right.
Forget textbook calculators. No real world drivers have textbook response, therefore textbook crossovers don't work for them.
Go to FRD Consortium http://www.pvconsultants.com/audio/frdgroup.htm and read some of the papers posted there. It will open your eyes to the task at hand. Then try the spreadsheets. Once you try those tools and have some success you might want to move on to other software and real measurement systems.
Another good guide to designing crossovers is posted on IJData.com the makers of lspCAD. It shows how to use their program step by step, but even without owning lspCAD the procedure outlined is pretty much universal regardless of the tools used.
I'm surprised that nobody said get Vance Dickison's book on loudspeakers. It's not exactly a simple read but if you can read and re-read it until it starts sinking in, you'll have a good idea of what is involved.
If you are new to electronics, get Horowitz and Hill's "Art of Electronics" It's the most accessible book on electronics I've seen.
Read, and listen to other threads. The problem with trying to learn on this forum is there is so much information flowing at you, and we often go off on tangents that you won't or shouldn't be concerned about at this stage. Knowing when to move on is a trick.
There's the madisound board as well.
Figuring out a crossover is as much about what drivers are picked for a design as it is picking the component values. Picking a woofer and tweet is about knowing what works well and what does not, what you like and what you don't.
For example, the Seas P17REX is a great poly driver that doesn't sound like a plastic driver. It's got a hell of a motor on it, which makes for a low Q (a good thing in many designs. Most importantly, it's well behaved, the frequency response is nice and smooth at both ends with no nasty resonant peaks. This makes crossing it over easy, and gives the designer some flexibility to choose different filter orders depending on what he's trying to accomplish, what tweeter it's mated with, what sort of box it's in, etc. Contrast that with the Seas excel magnesium cone woofers. They are amazing down low, but up in the midrange they have a resonant peak that will knock your teeth out if you don't do something with it. So your crossover needs to absorb the energy at that frequency area. And you really don't want to use anything above that. This forces a fairly steep crossover, or crossing lower and using a midrange driver.
Some tweeters are a dream to work with. the Seas 27TFFC has a nice natural well behaved rolloff at low frequencies, and can be crossed over a lot of different ways and sound good. It doesn't need much up top in the way of frequency response corrections. On the other hand, the vifa XT25, which is an amazing tweeter from 2500 Hz up, thinks it's really a midbass driver and tries to play down to about 500Hz without a whole lot of natural rolloff. It really shouldn't play much of anything below 2 KHZ which means that you really have to smack it down pretty aggressively in the crossover.
You should be starting to notice that the desired response from the driver is really a combination of it's natural response curve and the imposed curve by the crossover. So a tweeter that is naturally rolling off at say 12 dB per octave when combined with a second order crossover (12 dB/octave) will produce a tweeter response that has a 24 dB/oct response. The same is true of all drivers.
Now there's also phase to consider, in a typical small 2 way design, the tweeter and woofer are mounted to the same front baffle with the mounting flanges more or less flush with each other. The acoustic center of the tweeter is at the face plate or a little to either side depending on design. The woofer acoustical center is actually quite a bit behind the baffle front. This means that if they are both playing the same note, there is a phase mismatch and some cancelation going on when listening on axis. So the crossover also has to compensate for the phase difference due to mechanical offset of the drivers.
Notice I haven't said a word about capacitors, inductors, and resistors. Most crossovers look pretty similar. They are made up of voltage dividers (tweeter pads), low pass and high pass RC and LC filters, RLC notch filters, and RL shaping networks and such. There aren't a whole lot of building blocks. the art is in knowing how to apply the building blocks and what is needed for good sound. That's the practice part.
I find that a crossover breadboard, a soundcard, microphone, and speaker measurement software are the key to building good crossovers. I do some paper modeling, but the final crossover is measured in real time, then listened to.
I wrote a lot of words, and never answered your question really. Because the answer exists in many very comprehensive textbooks. It's a subject worthy of many hundreds of pages, and my fingers will fall off first.
Sheldon
If you are new to electronics, get Horowitz and Hill's "Art of Electronics" It's the most accessible book on electronics I've seen.
Read, and listen to other threads. The problem with trying to learn on this forum is there is so much information flowing at you, and we often go off on tangents that you won't or shouldn't be concerned about at this stage. Knowing when to move on is a trick.
There's the madisound board as well.
Figuring out a crossover is as much about what drivers are picked for a design as it is picking the component values. Picking a woofer and tweet is about knowing what works well and what does not, what you like and what you don't.
For example, the Seas P17REX is a great poly driver that doesn't sound like a plastic driver. It's got a hell of a motor on it, which makes for a low Q (a good thing in many designs. Most importantly, it's well behaved, the frequency response is nice and smooth at both ends with no nasty resonant peaks. This makes crossing it over easy, and gives the designer some flexibility to choose different filter orders depending on what he's trying to accomplish, what tweeter it's mated with, what sort of box it's in, etc. Contrast that with the Seas excel magnesium cone woofers. They are amazing down low, but up in the midrange they have a resonant peak that will knock your teeth out if you don't do something with it. So your crossover needs to absorb the energy at that frequency area. And you really don't want to use anything above that. This forces a fairly steep crossover, or crossing lower and using a midrange driver.
Some tweeters are a dream to work with. the Seas 27TFFC has a nice natural well behaved rolloff at low frequencies, and can be crossed over a lot of different ways and sound good. It doesn't need much up top in the way of frequency response corrections. On the other hand, the vifa XT25, which is an amazing tweeter from 2500 Hz up, thinks it's really a midbass driver and tries to play down to about 500Hz without a whole lot of natural rolloff. It really shouldn't play much of anything below 2 KHZ which means that you really have to smack it down pretty aggressively in the crossover.
You should be starting to notice that the desired response from the driver is really a combination of it's natural response curve and the imposed curve by the crossover. So a tweeter that is naturally rolling off at say 12 dB per octave when combined with a second order crossover (12 dB/octave) will produce a tweeter response that has a 24 dB/oct response. The same is true of all drivers.
Now there's also phase to consider, in a typical small 2 way design, the tweeter and woofer are mounted to the same front baffle with the mounting flanges more or less flush with each other. The acoustic center of the tweeter is at the face plate or a little to either side depending on design. The woofer acoustical center is actually quite a bit behind the baffle front. This means that if they are both playing the same note, there is a phase mismatch and some cancelation going on when listening on axis. So the crossover also has to compensate for the phase difference due to mechanical offset of the drivers.
Notice I haven't said a word about capacitors, inductors, and resistors. Most crossovers look pretty similar. They are made up of voltage dividers (tweeter pads), low pass and high pass RC and LC filters, RLC notch filters, and RL shaping networks and such. There aren't a whole lot of building blocks. the art is in knowing how to apply the building blocks and what is needed for good sound. That's the practice part.
I find that a crossover breadboard, a soundcard, microphone, and speaker measurement software are the key to building good crossovers. I do some paper modeling, but the final crossover is measured in real time, then listened to.
I wrote a lot of words, and never answered your question really. Because the answer exists in many very comprehensive textbooks. It's a subject worthy of many hundreds of pages, and my fingers will fall off first.
Sheldon
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.