xover impedence issues

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
well not knowning the details fo your driver impedance curves and your crossover here are a few general comments. they might or might not apply or might only partially apply. sorry for being vague. if you want to know more you have to get impedance curves of your driverand use a software like LEAP or LSPCAD or MLSSA (gurus call it Melissa).

the impedance of a driver is hardly flat. some crossovers equallise the impedance of drivers others only do it to the woofer and mids and not the tweeter as the Le of teh tweeter is usually small and the rise is often outside the audible spectrum. so you must consider the impedance of the drivers at the XO point. That means if a driver in nominal 8ohms but is 6.5 ohms at the XO point 6.5 ohms is the what yo must consider.

When 2 drivers are XOed lets say you have a 2 way with a woofer and tweeter. the XO is usually not a brick wall and hence for an octave of so both driver are producing significant sond (assuming they are XOed at the same freq.).

This means that both driver are in parallel and if you have a wofoer that is 6.5 ohms at the XO point and a tweeter that is 7 ohms at the XO point you have 3.5 ohms or there abouts at the XO. To avoid this some XOs use assyemtrical XO points or slopes or both.

I can go on but I think you got the drift. Very few systems are purely resistive. In fact a purely resistive 4ohm load will probably be less trouble to an amp than say a very reactive load that varies from 5 ohms to 25 ohms and goes up and down etc.

To build a XO which compensates for driver impedance curve anomalies and is purely resistive over say 50-10k is not easy. thorw in flat phase and frequency response (sensitivity correct) and you got a complex problem.
 
I still am not quite sure I understand... If I connect two 8-ohm speakers in parallel like it seems a crossover would make them, hen I end up with 4-ohms. If I hook up 3 this way, then I get something like 2 2/3 ohms. This is not good. I need 8-ohms... I have seen crossovers like the Dayton's at Parts Express that give 8-ohms with the mid, woofer, and tweeter hooked up-- the mid and tweet are 8-ohm and the woofer can be 4 or 8. How does this work? Is it possible to make a crossover like theirs that will act the same way, except with my values? How is this possible?
 
Administrator
Joined 2001
Paid Member
trespasser_guy said:
If I connect two 8-ohm speakers in parallel like it seems a crossover would make them, hen I end up with 4-ohms. If I hook up 3 this way, then I get something like 2 2/3 ohms. This is not good. I need 8-ohms...

This is true if you hook them in parallel. But if you have an XO in the mix what happens is that in the pass-band of the driver it is nominally 8 ohms (as mentioned a speakers impedance is not constant), but outside its passband its impedance is much, much higher.

So, for instance lets say we have a 250 Hz and 5000 Hz crossover points. Then the woofer is about 8 ohms till about 250 Hz and then its impedance -- because of the crossover -- becomes very large. Similarily the mid has a very large impedance below 250 and above 5k, and about 8 ohms between 250 and 5k. The tweeter is 8 ohms above 5 k and much larger below that.

Now lets look at the impedance at say 1k. The woofer is very large (VL), the mid is 8, the tweeter is very large(VL). Parallel them and you get a total impedance = 1/(1/VL + 1/8 + 1/VL). Now 1/VL is a very small number so we end up with an 8 ohm total impedance.

Because of the crossover, you end up with a nominal impedance of 8 ohms at any frequency.

The above is very much a generalization. No speaker is an 8 ohm load (or 4 or 2). The magnitude is not constant and the phase is not constant. Add in a crossover and it gets even wilder.

dave
 
Disabled Account
Joined 2002
Same with series crossovers? As far as Ohms equallying out.

On an interesting note, in the loudspeaker design cookbook, he said that series crossovers were not very popular nor practical so he would not discuss them, yet I have another speaker book ("How to Design, Build and Test Complete Speaker Systems
", I think) thats not anywhere as detailed as the LCB, yet it does include info and formulas for series. Just thought it was funny.
 
Administrator
Joined 2001
Paid Member
Westrock2000 said:
Same with series crossovers?

Yes

On an interesting note, in the loudspeaker design cookbook, he said that series crossovers were not very popular nor practical so he would not discuss them, yet I have another speaker book ("How to Design, Build and Test Complete Speaker Systems
", I think) thats not anywhere as detailed as the LCB, yet it does include info and formulas for series. Just thought it was funny.

Some are afraid of Series XOs because everything in the circuit affects everything else so they are lot slippery -- not nearly as cut & dried as parallel. Sort of like science (before chaos theory) not bothering with the 95% of problems not covered by linear mathematics.

dave
 
Administrator
Joined 2001
Paid Member
trespasser_guy said:
So, I am assuming that I can just build the crossover and hokk everything up and I have a THEORETICAL impedence of 8-ohms... right? Sounds too easy... :xeye:

In theory yes, but when reality hits the road one finds that the theory is often very much oversimplified. Getting an XO with all its reactive, real world components, and the speaker with its real-world reactive nature to all work together in a way that is ideal is far from easy -- one of the reasons i try to avoid them. :D

dave
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.