I am looking for a list that Bill published featuring all the steps he follows to create a crossover in XSim. Anyone have this or know how to find it?
Did you mean this post?
Probably (I'm not familiar with ARTA, but I think that's one of its purposes). Just make sure to include phase information and relative delay as well. Here is a short writeup I did (coincidentally just yesterday) outlining how this would be done with a 3-way:
____
For a three-way, measure the following (keeping the microphone position, the baffle position, and the levels the same for all measurements -- WITH phase response included):
**Tweeter response
**Tweeter + Midrange response (both together, no crossover)
**Midrange response
**Midrange + Woofer response (both together)
**Woofer response
Then, make an XSim design model with nothing but three drivers connected in parallel together to the power amp. "Tune" the top one to have the Tweeter's FRD file; the middle one to have the Midrange's FRD file; and the bottom one to have the Woofer's FRD file. (You don't need to bother with impedance files at this point until you do actual crossover designing)
Mute out the woofer to start. You should see in the Frequency Response plot the simulated effect of the tweeter and midrange both playing together, which is the situation of the 2nd measurement you made above. In that Frequency Response plot, use its "Curves" menu to "Get File" and load the actual measured combined "Tweeter + Midrange" response. The idea is to compare this to simulated. "Tune" the midrange driver's "Mod delay" until the simulated response very closely resembles the measured combined response.
Next, unmute the Woofer and mute the Tweeter. Change the "Get File" to the combined "Midrange + Woofer" response FRD file. Tune the WOOFER's "Mod delay" value till the simulated closely matches that curve. Don't touch the midrange's or tweeter's "Mod delay" values, since you previously aligned them together. Now you have all the response files loaded to the proper drivers, with the actual relative delays established between them all. You can load your measured impedance files now, detach the drivers in the model from each other, and begin designing/simulating your crossover. The simulated results you get should essentially EXACTLY agree with what you will get when you build the actual designed crossover.
____
If your ARTA setup measures absolute phase (is "time locked" and uses a soundcard that can do that right), and can produce FRD file format, you could just measure to make files and go. The 'combined' measurements and delay adjustments are when the time isn't locked, as with Omnimic (Omnimic is designed to not need any cables between the measuring system and what is being measured).
Unfortunately, Xsim will not predict (calculate) the values of the components of the crossover network to get a desired target frequency response curve. It will tell you which will be such response with the components you put between the driver and the amp.
Suppose you put a L-C second order filter between them. The expected response will be calculated with the values of L and C and the frd and zma data of the driver you want to use. The calculated response will depend on the values of L and C you give as inputs. Later, you need to compare that response with that you have in mind as a target.
I test the values of L and C, by trial and error.
I share to all the team, an Excel file to calculate the transfer function of the low and high-pass filters, of a 4th order Linkwitz-Riley crossover.
In it, I copy and paste the response plot that Xsim gives me, with the components I put in the schematic, I paste these plots (low or hi-pass) over the plots of Excel of the transfer functions, but I need to "sent to back" the captured image of Xsim, in order I see the Excel plots superimposed over the Xsim plots. Is very probable that I need to adjust the axis of the Excel plots to get the correct alignment on both plots, the one of Xsim and the other of Excel.
I did this exercise with the frd and zma files of the RS150P-8A woofer and the DC28F-8 tweeter (both from Dayton Audio) I´ve downloaded from Parts Express site.
Also, I share the Xsim files of this exercise.
The last step: be patience. To find the right values is time consuming, and suicide is not an option.
Cheers!
Suppose you put a L-C second order filter between them. The expected response will be calculated with the values of L and C and the frd and zma data of the driver you want to use. The calculated response will depend on the values of L and C you give as inputs. Later, you need to compare that response with that you have in mind as a target.
I test the values of L and C, by trial and error.
I share to all the team, an Excel file to calculate the transfer function of the low and high-pass filters, of a 4th order Linkwitz-Riley crossover.
In it, I copy and paste the response plot that Xsim gives me, with the components I put in the schematic, I paste these plots (low or hi-pass) over the plots of Excel of the transfer functions, but I need to "sent to back" the captured image of Xsim, in order I see the Excel plots superimposed over the Xsim plots. Is very probable that I need to adjust the axis of the Excel plots to get the correct alignment on both plots, the one of Xsim and the other of Excel.
I did this exercise with the frd and zma files of the RS150P-8A woofer and the DC28F-8 tweeter (both from Dayton Audio) I´ve downloaded from Parts Express site.
Also, I share the Xsim files of this exercise.
The last step: be patience. To find the right values is time consuming, and suicide is not an option.
Cheers!
https://www.diyaudio.com/community/threads/xsim-free-crossover-designer.259865/post-5260072
(This link should work if the one in post #2 does not.)
(This link should work if the one in post #2 does not.)