crossover gurus / component slight value changes…

tajanes

Ok, I’ve been playing with this crossover calculator/simulator (https://speakerwizard.co.uk/calcs/Crossover_Calc_v2.php) as it graphs anticipated results from component value changes.  That being seen however (i.e in a 2nd order LR crossover) I’d be interested in comments (and/or link to articles) as to in what situations would it be beneficial to either increase or decrease a capacitor (or an inductor’s value) from what would be initially calculated as ideal (for a targeted Fc point), and how this approach would differ in the HP vs LP section (as the cap and inductors are in switched positions (parallel / series) relative to the tw and woofer. 

Thanks in advance for your comments / insights

Andy

 

jhollander

The drivers don't have flat frequency responses or impedance so you would adjust the theoretical values so the frequency responses matches the desired response curve.

You would also shift the theoretical valves and the maybe the FC point to align the driver's phase.  This can be seen in programs like PCD where you can create a targeted response curve. 

I also shift the theoretical values to raise the impedance or to use the parts I have on hand.

Wolf

You shouldn't be using a xover calculator, nor forcing a driver to follow a certain roll-off. It's easier to allow drivers to show you where they prefer to go, and not focus on the number Fc. There are times where wobbling response dictates using more parts to blend better at the xover point, but setting a number before you know what is going to work is a bit early in the design process.
Xover points in a second order or higher xover network do not shift with different impedances, but the Q of the filter is altered and the FR is affected. The Q can be readjusted to compensate by increasing or decreasing the Q by increasing the cap or coil value and doing the opposite to the other component in a second order network. This will change the shape of the knee of the roll-off, and help with the blending.

tajanes

Thx for the input.

I'm always interested in the 'why' beyond the calculations, and the potential for tweaking designs for desired results.  Having opportunity to get input from those with experience (and boxes of caps and inductors…) is much appreciated!  

One area I have found where targeting roll-off can/may be beneficial (all else being equal) is dealing with baffle step issues in designing crossover Fo and slopes/orders btwn W and Mid when crossed around the B step (such as in a 3 way, or a Woofer with an extended driver design).  ‘Course its important to design a speaker comprehensively and select drivers with appropriate parameters.  That being said, certainly agrees that crossover point and orders are more dictated by drivers in two-ways and mid/tw.

Kornbread

IIRC, Wogg has links to both WinPCD and XSim on his site, http://woggmusic.com/advanced-speaker-design-simulation/