Crossover Help - Epique & BZ Project

[Wolf]

FYI, Dezzar and JohnH are advocating for different style models.
Dezzar will want all drivers on their axes as well as off axis to model for a directivity index in VituixCad.
John is going to use PCD or Xsim to find relative offset at the listening position and requires all measurements initially to be on the main listening axis. This is for one axis only, but the most important one, and once developed, off axis measurements or models will be further verified.

Neither is wrong.

Also, I'm uncertain about your woofer wiring. These are DVC woofers, so you have 8 voice coils. Are the "bottom 2 in series" wired with all 4 coils in series to 16 ohms net?

Wolf

[jhollander]

I was assuming he wanted to continue to use XSim.

[Wolf]

Good assumption, should he want to learn more as he goes. I'm with you here...

[jml]

All:

My apologies for the silence over the past couple of weeks. Between work and family, I haven't had much time to spend on this project recently.

Wolf, I appreciate your clarification around the different approaches being suggested for measuring. I looked a bit at VituixCAD, and it sounds like something worth transitioning to over time, but for now, I think I'll stick with Xsim.

All of my initial measurements have been taken at what will approximately be ear height of the speaker when seated - on what I consider an average couch for an average person anyway. If this is "acceptable" when modeling in xsim, then I'll plan to continue to do so.

As for the wiring - with the 8 VCs in each speaker, they're wired series-series- parallel - each DVS woofer wired in series up to 8 ohms, then the top and bottom woofers wired in series up to 16 ohms, then the two pairs wired in parallel down to 8 ohms.

Cheers!

Joe

[jml]

Good Evening!

I felt like it was time to try something new with the crossover model. I've never designed a series crossover before, but figured that it was worth a shot. What do y'all think of the below?

- Woofer crossover point is probably still a bit high.
- The midrange breakup is mostly tamed, but there is still some noise around 4k.
- Component count is getting high, but feels manageable.
- The series crossover seems to have significantly improved the low impedance issues I was having previously

Thanks in advance for the feedback!

Joe

[Steve Lee]

That looks great!

I'm watching the progress.

[jml]

Good Evening!

I had enough parts laying around to get pretty close to building the crossover above - not perfect on all values, but very close. I haven't had much time to listen or test, but below is the first FR measurement.

Cheers!

Joe

[4thtry]

Can't help with series crossovers, as I do not know much about them. However, a few years ago I put together an XSim step by step guide listing my steps for designing crossovers using OmniMic, XSim, and DATS V2 (attached PDF). On your latest crossover CAD screen, it does not appear as if you have included XSim's "mod delay" in your model. When the proper "mod delay" is included, as small "m" will appear after two of the three driver names. For instance, "S1", "S2m", and "S3m" instead of just "S1", "S2", and "S3".

XSim step by step2.pdf

[jml]

Can't help with series crossovers, as I do not know much about them. However, a few years ago I put together an XSim step by step guide listing my steps for designing crossovers using OmniMic, XSim, and DATS V2 (attached PDF). On your latest crossover CAD screen, it does not appear as if you have included XSim's "mod delay" in your model. When the proper "mod delay" is included, as small "m" will appear after two of the three driver names. For instance, "S1", "S2m", and "S3m" instead of just "S1", "S2", and "S3".

[ATTACH]n1495645[/ATTACH]

Thanks! What does the mode delay do?

[jml]

In looking more at the response I posted above, I was confused why the midrange breakup was coming through so much more than in the model. It turns out that if you drop a 22 ohm resister into the notch filter instead of a 2.2 ohm resister, it's not going to work right haha! Here is the updated graph with the rectified resister value.

Cheers!

Joe

[4thtry]

Thanks! What does the mode delay do?

You ask a very difficult question, but I will try to explain this as best I can. The "mod delay" is a small amount time, either positive or negative, that is used by XSim to synchronize time reference (or phase) differences between each driver. When you are using a single channel measurement microphone, such as OmniMic, the phase curve for each measurement begins, or is time referenced to, the peak of the impulse response. This time reference is located slightly in front of each driver flange (quasi-minimum phase). When making measurements with OmniMic, the mic tip always stays in the same position for each measurement. But the drivers are all located at different distances to the tip. As a result, the "time of flight" from the mic tip to the starting point of each measurement is different. The "mod delay" that you enter in XSim is needed to make a "time of flight" correction to synchronize the phase of each dirver. Without this correction, OmniMic measured phase and magnitude response will not sum properly in the XSim model.
​

[4thtry]

Mod delay is entered as the equivalent number of inches on the Select Driver screen: