Crossover Help - Epique & BZ Project

[DeZZar]

What's the distance between the centre of the mid and the lowest woofer?

If you have all 4 woofers playing the same frequency range I feel they are crossed to high to the mid.

[jml]

It’s 32” C2C from the mid to the lowest woofer.

Cheers!

Joe

[DeZZar]

General rule of thumb with that distance is that your maximum crossover should be no more than 420hz for that lowest woofer. Currently its playing well up into the midrange attempting to cross to the midrange driver at ~900hz.

This is likely to contribute to a sort of less focused sound affecting that desirable 'point source' for the midrange and upper frequencies. Could possibly even create some issues with imaging.

My suggestion would be to:

• Cross lower to the midrange - that midrange should be well and truly able to play a lot lower; AND/OR
• Consider configuring for a 3.5 way. Determine your point of baffle step and apply an additional low pass for the bottom two woofers relegating them well into the omni-directional frequencies and letting only the top two woofers play up further to the midrange

(as mentioned before, to model this accurately you really need individual measurements for each woofer, and put them into the sim as independent drivers with the correct y offset for each from the listening axis. For each woofer you also want to take nearfield measurements and combine them into the FRD used for sim)

[Wolf]

Move your tweeter Lpad parallel resitor to after the xover, or those resistors will get very hot being across the amplifier terminals running full bandwidth.
Wolf

[rpb]

You can see the impedance changed from previous measurements!

[rpb]

For measurements, I would get the speaker higher off the floor. one option might be to support it horizontally with all 4 woofers about 4' off the floor. Maybe higher, if the ceiling is more than 8'.

As suggested by DeZZar, I'd cross the woofers lower.

If you concentrate on the mid, and tweeter xo, you can more easily measure them down to a 300hz roll-off at their current padding. Get the blend between the two as smooth as you can. The woofers are probably difficult to measure. A filter based on one woofer could be used. If there's a way to measure one woofer alone, you could see what a close mic measurement looks like. With 4 woofers, I'd measure from the center of the woofers.

[jml]

Thanks everyone. I'll take some individual measurements of the woofers and work on pulling the XO point down. I'll also move the tweeter pad to avoid the heat issue Wolf mentioned.

Cheers!

Joe

[jml]

Greetings! I haven't had much time to work on this lately, but was able to take a few new measurements, and play around with a revised XO a bit. I did not do individual woofer measurements, but measured them in groups of 2 (top 2 and bottom 2). This version is a 3.5 way below are a few of my thoughts, but I obviously value everyone's input.

1) The XO point with the top woofers is still fairly high
2) The breakup of the mid might still be too high?
3) Impedance on the top end is too low - perhaps rolling offer the mid at a steeper slop addresses 2 and 3?
4) The tweeter is inverted in this version, which cleaned up a bit of noise at the XO point.

Cheers!

Joe

[rpb]

My speculation at a glance. I'm not running sims, I'm just looking at it.

Impedance. The tank filter resistor is too low. Also R4 is too low. Increase R4 to about 4 ohms. Add series resistance somewhere else in the mid filter.

Looking at the mid low-pass, it looks to me like a 3rd order, or maybe a damped 3rd order filter, would work better. Try to make a bit of a knee in the response at about 2.7k.

Just another WAG, but what if you added some impedance compensation to the woofers? A rule of thumb guess would be 8uf, and 8 ohms in series with each other, and placed parallel to the woofers.

What happens if you make C7 much larger? Twice as big, and reduce the coil values?

[DeZZar]

any chance you could attach the current FRD files you're using for this sim?

[Wolf]

I agree with what rpb said.
You really are not lowpassing the midrange much at all. You notched it, but didn't roll it off.
Additionally, you could Pad the tweeter differently to increase impedance too by padding more with the series resistor and less with the parallel value; ie- increase both values, say 5 and 20 ohms.

Wolf

[jhollander]

No offset on the drivers in XSim? You'll need a woofer plus tweeter measurement to get the phase right.
Also Tweeter plus mid.
Just to be clear are the mid and woofer divers in separate enclosures?

[jml]

No offset on the drivers in XSim? You'll need a woofer plus tweeter measurement to get the phase right.
Also Tweeter plus mid.
Just to be clear are the mid and woofer divers in separate enclosures?

Yes, mid is in a sealed enclosure - woofers are in a separate ported enclosure tuned to about 40hz.

[jml]

any chance you could attach the current FRD files you're using for this sim?

No problem. Here are the files. I'll post the ZMAs in another post.

[jhollander]

I'm happy to play along if you want to provide the TW plus mid and then a TW plus woofer top and TW plus woofer bottom
I assume the Woofer Top FRD and Bottom FRD are 4 woofers in separate enclosures of 2 and 2?

[jml]

I am more than happy to have the help! all 4 woofers are in the same enclosure, the top two are wired in series and the bottom two are wired in series. My original measurements were all 4 together (series-parallel), and then I measured them separately to attempt the 3.5 way design. I have not done measurements of all woofers individually, but can as well.

For those additional measurements with the tweeter, would you need FRD and ZMA or just FRD?

[jhollander]

Just FRDs when doing combined measurements TW plus mid, TW plus 4 woofers. We need to find the delay between drivers. If all 4 woofers are in the same enclosure you should measure together. You can't get a accurate measurements if you are disconnecting drivers in the same enclosure. Your impedance measurement will also be a problem if you want to separate the bottom 2 woofers for a 3.5.

[jml]

Here are the ZMA files - the woofer ZMA is a pair of woofers wired in series.

[DeZZar]

Yes, they are all take on-axis at approximately 4 feet from the baffle with the mic height slightly higher than the center of the tweeter.

Just want to check if understand you correctly - are you saying that each driver has been measured with the mic staying in the same spot - pointed just above tweeter? Or have you measured each driver on their respective axis?

The latter is what you need in order to develop an accurate simulation.

This is still probably one of the best guides around for making measurements: https://www.audiosciencereview.com/f...tuixcad.21860/

[DeZZar]

Just looking at your FRD files it certainly seems to be off-axis. This is why the tweeter response seems to tank after 8K. The BZ tweeters have a slight -3db roll off from 10K up, but your files show a measured response dropping over 12db from 7k onwards - this is because the measurement is off-axis and is going to completely mislead you attempting to pad the tweeter correctly.

[jml]

The tweeter measurements were perhaps 1.5” above center, but still in line with the top part of the ribbon. I can try some tomorrow that are dead center. I’ll also thoroughly review the guide you posted. Thanks!

[jhollander]

As long as the mic position doesn't move the measurements will be good. Better to have the mic position on the listening axis.