Crossover: The infamous CAPACITORS

That's why so many people suggest building an existing project where someone else has done the hard work. No-one will look down their nose at you for doing that, in fact they will probably tell you what a great job you've done.

You could take Moondog's suggestion in post #75 to make a small 2 way from your tweeters and mids. Do you have the FR and impedance graphs for the mid - if so, we could probably come up with a sim which could get you started on something which you could then tweak.

You might also perhaps consider an existing design which uses your Vifa, it's an excellent, high quality tweeter. I've seen it used with the Peerless 830657; the Dayton RS180P and R180 and in 2 and 3 way designs and there are many more out there.

Geoff
 
Pedroga, you are over thinking things. Use all electrolytics for this project and don’t worry about anything greater than 50v and 10% tolerance. There are more of them in speakers around the world than all others combined by a massive margin. The drivers you have are fine, so are electrolytic for this system. You have a first order xo of the most basic design. As mentioned, you will notice a great deal of difference in getting the values correct more than you ever will in what type of cap or coil you use. You have no resistors nor shunting happening so you are at the very beginning of your journey. I hope it goes well. Your ears will tell you the truth. That’s the point at which this crew will be able to help with things like notches, zobels, L-pads, BSC, etc.
I basically agree with all above > except, unless used in series, I NEVER buy 50V bipolars anymore - I buy 100V.
Using small value PolyCaps as a bypass can save money & work, especially for tweeters.
PS.
I don't see how the efficiency of the early mentioned 4 inch mid can match the output of the Pioneer 'woofer'.
Reducing output level of tweeters is easy, but reducing the output of woofers is just not a practical thing to do passively.
But then again, you can use a woofer's inductor that has high DCR. ( and expect heat )
 
Last edited:
But even if I build everything beforehand (box, etc.) I don't have the equipment like a microphone to measure. It would be an even bigger problem
Get a cheap electret capsule. It will be sufficiently precise at the frequencies where your crossover points are likely to be. While far from perfect, it will be magnitudes better than poking around in the dark.
Just as an example: aliexpress.com/item/1005004281678104.html


Good luck!
 
  • Like
Reactions: Audio>X and JMFahey
That's why so many people suggest building an existing project where someone else has done the hard work. No-one will look down their nose at you for doing that, in fact they will probably tell you what a great job you've done.

You could take Moondog's suggestion in post #75 to make a small 2 way from your tweeters and mids. Do you have the FR and impedance graphs for the mid - if so, we could probably come up with a sim which could get you started on something which you could then tweak.

You might also perhaps consider an existing design which uses your Vifa, it's an excellent, high quality tweeter. I've seen it used with the Peerless 830657; the Dayton RS180P and R180 and in 2 and 3 way designs and there are many more out there.

Geoff
I have the graphs of all the speakers.
Why shouldn't we trust sellers' information when cutting frequencies?
I basically agree with all above > except, unless used in series, I NEVER buy 50V bipolars anymore - I buy 100V.
Using small value PolyCaps as a bypass can save money & work, especially for tweeters.
PS.
I don't see how the efficiency of the early mentioned 4 inch mid can match the output of the Pioneer 'woofer'.
Reducing output level of tweeters is easy, but reducing the output of woofers is just not a practical thing to do passively.
But then again, you can use a woofer's inductor that has high DCR. ( and expect heat )
This was an error that I only saw recently.
I noticed their low sensitivity but when I connected it to my amplifier I found their volume to be quite high and the quality, to be honest, of all the Brazilian speakers I had the opportunity to see, they beat them in terms of quality and performance.
I've always had terrible luck with originally Brazilian speakers, from muddy sound to poorly sized boxes described in the speaker's own manual.
And for the volume of sound it gave, I wouldn't mind attenuating the woofer even though it's a huge waste of energy.
 
Get a cheap electret capsule. It will be sufficiently precise at the frequencies where your crossover points are likely to be. While far from perfect, it will be magnitudes better than poking around in the dark.
Just as an example: aliexpress.com/item/1005004281678104.html


Good luck!
I'm going to take a look at this microphone.
The ones I have available at home are all condenser microphones for voice recordings.
 
You build the box, according to sensible principles: midrange close to tweeter, placement of both simulated beforehand with some software (e.g. Vituixcad has a module for that) for smoothest results. If possible give the box a generous roundover.
Then you measure the drivers individually, in the box, on- and off-axis. You feed those results into a simulator - again, e.g. Vituixcad, but there are several around, many shareware.
Then you virtually cable it together to simulate crossover variants. No need to buy any components until you settle on a design that makes you happy. That's one great advantage of using computer tools - another is better loudspeakers!

There is a learning curve, but it's sure worth it.
 
I have the graphs of all the speakers.
Why shouldn't we trust sellers' information when cutting frequencies?
You need to measure the drivers in the box you build for them, because the variations in response of different box geometries are not subtle. You need to incorporate that result into the crossover - that's why everyone here is so insistent that super-low part count is unlikely to work well.


That cheap mic is just an example, but usually electrets are pretty linear between 200-6000 Hz, which is all you need to get you into the ballpark. There are quite a few super-affordable options on Ali. The bass you will need to calculate via T/S parameters anyway. Did I mention a learning curve? 😉
 
if not obvious for beginner, sound reacts with physical objects reflecting, diffracting and transmitting. A driver attached to different objects would make the response differ some. Although the driver itself emits sound like it does, the immediate surroundigns, the physical structure the driver is attached to, reflects, diffracts and transmits sound, so affects the response! A driver datasheet measurement is usually taken in a standard IEC measurement baffle, but you are building something else, like bookshelf speaker so the measurement would look different.
 
  • Like
Reactions: JMFahey and stv
You need to measure the drivers in the box you build for them, because the variations in response of different box geometries are not subtle. You need to incorporate that result into the crossover - that's why everyone here is so insistent that super-low part count is unlikely to work well.


That cheap mic is just an example, but usually electrets are pretty linear between 200-6000 Hz, which is all you need to get you into the ballpark. There are quite a few super-affordable options on Ali. The bass you will need to calculate via T/S parameters anyway. Did I mention a learning curve? 😉

So will I need to do the box before the crossover?
 
  • Like
Reactions: tmuikku