2-way passive crossover

Discussion in 'Car Stereo Speakers' started by kcox669live, Oct 22, 2011.

  1. kcox669live

    kcox669live Full Member

    i have a set of alpine spr-17c comps. I ran a frequency sweep through the entire system, subs, mids and highs, dis-connecting different speakers at times, and noticed holes and rises in a few spots; at 100hz-200hz the system takes a poop. So i took a look at the crossover that came with the speakers.

    I drew a diagram of the circuit, ran the values of the components of the crossover through some calculators and determined that i could do better.

    What i want to do is breadboard my own xover so i can try different values in caps, inductors and resisters.

    The problem I am having is that, again, there are lots of calculators and formulas on the internet and none seem to agree.

    I'd like to try a band-pass for the mids and a high-pass for the tweeters.

    If anyone knows a link to a good calculator or formula or any good advice on this subject I would love to hear it. I'm in a tinkering kind of mood so the more complicated the better, lol.

    Btw, I can't afford xover pro 3 :(
     
  2. pedro quiroga

    pedro quiroga Well-Known Member

    i had a link to something like this a while back.... let me look for it.
     
  3. pedro quiroga

    pedro quiroga Well-Known Member

  4. Ranger SVO

    Ranger SVO Full Member

    I've built alot of crossovers. And its really cool when you build one that works well.

    Example 1

    Example 2

    Example 3

    The different equations is due to the different types of networks. The Butterworth, Bessel, Linkwitz-Riley and Chebychev are the most common types and they have different characteristics and therefore use different calculations even if they have the same slope and crossover point. Before you start building crossovers do a little more research on the different types of crossovers. My opinion, look at the Butterworth and the Bessel. They have the smoothest responses.

    Typically if your missing the 100-200Hz range, make sure the speaker is mounted properly. And I usually would not recommend a band pass filter for your midbass or midrange speaker. In fact I don't even use bass blockers on my midbass/midrange speakers. They just are not necessary unless your using a small 4-inch or 3-1/2-inch speaker.
     
  5. kcox669live

    kcox669live Full Member

    http://www.allaboutcircuits.com/vol_2/chpt_8/4.html

    i found these a while ago and kept them for when i too was in the mood to tinker..never happened. good luck.[/QUOTE]
    Thanx pedro. This second one is new to me. I had to look around a little on it to find out why they like resistors instead of inductors, very interesting.
     
  6. kcox669live

    kcox669live Full Member

    Incredible work ranger. I hope mine turn out that nice. It gave me some really good ideas. Now where can i get some Piezos, ha ha.

    Yep, I've been studying Butterworth and Bessel, I'm on the right track.

    I'll def check the mids again but I mounted them behind the door panel just for that reason, I didn't think they would have been real solid in the panel.

    Here's where it gets weird. When I did my sweep I started at 25hz, being careful at that frequency. At 40hz the mids were picking it up, maybe only at the same db as 100hz but I definitely didn't want to turn it up to loud. I fudged with the amp crossover but could not get rid of it. Problem with my active crossover? So that's really where I came up with the band-pass idea, well... that and when I sketched out the factory crossover I got into a tinkering mood.

    So I'm thinking of a band-pass with the high-pass at maybe 110hz. I still have to call alpine to get a good low-pass frequency for it and set the high-pass for the tweeter to that frequency also. Low-pass to the subs at 100hz.

    Btw, I did search for passive crossover in the little box with the ping-pong paddle next to it but I didn't see your threads.
     
  7. Ranger SVO

    Ranger SVO Full Member

    I want my midrange/midbass drivers to play as low as possible. I use plenty of dynomat on the door, especially around where the speaker is mounted. To strengthen it even more, mine also have a 1/4-inch piece of MDF between the door and the speaker. It stiffens up the thin speaker door mount alot more.

    Set you high pass lower. My recommendation, use a 6dB on the high pass side. A shallow slope lets the sub and the midrange/midbass drivers sound blend better.
    A 100Hz slope would take 2 199uf caps in parallel. The only problem is that High value caps is that they don't perform well. To fix their slow response place a 1.0uf polypropylene cap in parallel.

    These caps are cheap so for experimental purposes buy a couple 199uf, a couple 99uf and a couple of 33 or 47uf. Change the values around and see what happens. Listen to the result for a day, or at least a few hours before you change it again. This is only for the lowpass side.

    Start playing
     
  8. Ranger SVO

    Ranger SVO Full Member

    Resistors allow us to build cheap crossovers. High value caps cost a few cents more than a low value cap and a resistor. So first you choose a cheap low value cap, any value, then figure out the frequency you want and then determine the speaker impedence necessary to get to that frequency. Put a resistor in series or parallel to get that frequency.

    But you can also use it in really good crossovers. You can slow down or speed up the response of a capacitor with resistors. You can tailor the way your crossover works.

    In a piezo crossover the resistor is a must. As frequency goes up, the resistance in the piezo goes down. In fact it will approach zero, a short. The resistor provides a stable load for the amp. Also in the piezo crossover, I did not know the exact capacitance value of the piezo. We changed the Attenuation cap (it was the 1.0uf) a couple of times and settled on a 68uf cap. It sounds amazing, I wanna put one in my truck.
     
    Last edited: Oct 23, 2011
  9. kcox669live

    kcox669live Full Member

  10. kcox669live

    kcox669live Full Member

    These values came from the calculator that pedro posted. Checking these values with a chart that I've been looking at I found the values to be very close.

    Like I posted earlier I am putting this circuit on a bread board for experimentation.

    One of the things I will be looking at is the phase shift of the tweeter. I didn't compensate for it because I heard it might be a good thing.

    What I would like to learn more about is how to use and calculate resistors in the circuit.

    Anyway, I think this is a pretty good place to start.
     
  11. kcox669live

    kcox669live Full Member

    this one.jpg

    corrections so far
     
  12. kcox669live

    kcox669live Full Member

    So, I checked the comp mounts for vibration. They are very solid. I'm always hearing horror stories about installs and it makes me grateful I drive a traiblazer, lol. this one.jpg
     
  13. Ranger SVO

    Ranger SVO Full Member

    I'll crunch the numbers a little later on and see where your at, but it looks good. The only other thing your gonna run into is something called standard values. Its unlikely you'll find a .18mH coil, but a .15mH will be readily available. And that slight a difference won't make any difference in what you hear. I don't think that 24uf is a standard value but I know that 22uf is. But again the difference in nothing we can hear.

    Build it and lets see how it goes
     
  14. kcox669live

    kcox669live Full Member

    Sweet!!! I haven't sat this down, I've just been obsessing on it.

    I went to the electronics store I had high hopes for. I was only able to leave with a bread board and some good info, such as standard values, lol. But since then I found a real nice site that has all the caps, inductors and resisters I can stand.

    Here is one thing that I can't seem to get. I can't remember which site told me this but I saw it right before I posted the sketch. It said that filters 1 and 2 should be separated by 2 octaves. Why? At 2khz the slope will be -12db, 93.7% down from 0.

    The only question I have about the "low" high-pass filter is, is it low enough. I really want to get rid of the noise I heard at 40hz though.

    I think what I will do is buy the components for this circuit and another one that has a high-pass of 4khz and a low-pass at 2khz. That's a 1 octave split, which means that at 3khz they will both be down 6db. I'll also get the components you suggested, but the "low" high-pass filter will have to wait, 3 100uf caps in parallel is a little expensive.

    I'll have something up and running very soon. Thanks ranger.
     
  15. Ranger SVO

    Ranger SVO Full Member

    OK I'm finally on Christmas break, and I looked at your crossover. Here is the problem, your down 12dB at 2500Hz on both the HP and the LP parts. In other words you will be missing a huge range of sound. You cannot have a gap that big, it won't sum flat. If you cross your tweeter at 5000, then you need to cross your midbss at the same thing.

    If you think you might have to much gain at the crossover frequency, you won't. The highpass filter will start to roll off before 5000Hz. Its just the way caps and coils work. Same for the lowpass part. But they will sum flat.

    But if you insist on a seperation (I know your exprimenting and learning) then no more than an octave. If you have 5000 on the tweeter then try 2500 on the midbass.

    Let me know how it goes, I really want to know

    Let me remind you that on the 56 Chevy I used no crossover on the 5-1/4-inch speakers. They are run full range. Only the horns have a crossover on them.

    [​IMG]
     
  16. kcox669live

    kcox669live Full Member

    Going to be doing some testing today.

    The split on this circuit is 2 octaves, makes no sense to me either. I saw another site that said 3 octaves. The people who wrote these sites seem like very smart people, so I guess what they are doing is over my head. I'll stick to the more simple sites on the next one.

    So what I have going on here is, pos and neg in is at the bottom left, pos and neg to tweeter at top left and pos and neg to woofer at top right.

    High pass on the left, low pass on the right.

    I did not remember full range on woofer but good point. I'll do some tests on that set up also.

    Back to resisters, lol. If I use a resister of a certain value and put it in parallel to the tweeter, then if I use a calculator that asks me, "speaker impedance, "would I then put the resisters value, thus bring down a capacitors value?

    Thanks for all the help ranger.

    this one.jpg
     
  17. Ranger SVO

    Ranger SVO Full Member

    Its almost that simple, but remember that the tweeters resistance (called impedence because it changes with frequency) changes. As frequency goes up so does the resistance. As frequency does down, so does resistance. So we do not place a resistor in parallel. Part of the current will pass through the resistor and part will pass through the tweeter. We want all the current to go through the tweeter

    The only exception to the parallel resistor is the piezo tweeter. Its properties is opposite of a normal speaker and we need the resistor to protect the radio or the amp driving it.

    If you wanna use a resistor in series with a cap thats fine. Calculate the caps resistance at the crossover frequency. Then choose the cap your gonna use and calculate its resistance at your frequency. Add a resistor to make up the resistance difference.

    R= 1/(2*pi*C*F) where C is the capacitor value and F is the frequency.
     
  18. kcox669live

    kcox669live Full Member

    Did some testing today :D Way too much to explain here but I'll be uploading something to that other site soon, you know, the one with all the videos?

    This bread boarding stuff is incredibly interesting and very confusing at times.


    this one 2.jpg
     
  19. kcox669live

    kcox669live Full Member

    Resistance of capacitor needed for 90hz
    R=1/(2*pi*c*f)
    R=1/(2*3.14*.000313*90)
    R=1/.1769076
    R=5.65 ohms

    Resistance of 100uf capacitor at 90hz
    R=1/(2*pi*c*f)
    R=1/(2*3.14*.0001*90)
    R=1/.05652
    R=17.69 ohms

    Difference, in resistance, between the two values.
    17.69-5.65=12.04 ohms

    Is my math correct? Do I need to put a 12 ohm resister in series with a 100uf cap to get a 90hz crossover point? It's been a while since I've done anything like this, so I had to get my son to help.
     
  20. Ranger SVO

    Ranger SVO Full Member

    Your math is correct, but looking at the math suggests that might be for your low pass? If so no resistor on the low pass.

    If its for your high pass, then we need more capacitor and less resistance. Lets get our resistor to 5-ohms or less. We don't want to lose audio and a 12-ohm resister will cause us to lose a little.

    Thats how I learned crossovers. Alot of reading, but nothing like experimenting and actually hearing the difference. Its like that piezo crossover, never built anything like it. Read everything I could find, built it, put it in the car. Made three changes once it was installed. But I made it work, nothing like that kind of satisfaction.
     
    Last edited: Dec 25, 2011