ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1] 2 3  All   Go Down

Author Topic: Combining drivers  (Read 7203 times)

Helge A Bentsen

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1753
  • Oslo, Norway.
Combining drivers
« on: March 02, 2021, 05:25:39 PM »

Attached is the phase plot of two drivers post equalization with the same reference delay.

What would be your method for combining these for the most linear phase response?

Logged

Russell Ault

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2484
  • Edmonton, AB
Re: Combining drivers
« Reply #1 on: March 02, 2021, 10:45:30 PM »

Attached is the phase plot of two drivers post equalization with the same reference delay.

What would be your method for combining these for the most linear phase response?

That red plot is barely visible on my screen. Is your coherence blanking jacked way up? Can you attach the two .trf to your post (or dump them out as ascii)?

Also, to make any meaningful recommendations it's important to see the mag trace as well, since the goal is phase alignment within the overlap band (and without the mag traces we don't know what the overlap band is).

-Russ
« Last Edit: March 02, 2021, 10:47:42 PM by Russell Ault »
Logged

Helge A Bentsen

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1753
  • Oslo, Norway.
Re: Combining drivers
« Reply #2 on: March 03, 2021, 04:40:32 AM »

That red plot is barely visible on my screen. Is your coherence blanking jacked way up? Can you attach the two .trf to your post (or dump them out as ascii)?

Also, to make any meaningful recommendations it's important to see the mag trace as well, since the goal is phase alignment within the overlap band (and without the mag traces we don't know what the overlap band is).

-Russ

I've turned off coherence just to clear up the screenshot.
Also forgot to save the two captured traces showing overlap, this is with a Linear Phase BW crossover inserted.
Logged

Helge A Bentsen

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1753
  • Oslo, Norway.
Re: Combining drivers
« Reply #3 on: March 03, 2021, 04:42:50 AM »

Btw, it's a measurement indoors in my awful sounding loading bay with the speaker on top of a road case.
Logged

Russell Ault

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2484
  • Edmonton, AB
Re: Combining drivers
« Reply #4 on: March 03, 2021, 06:14:36 AM »

I've turned off coherence just to clear up the screenshot.
Also forgot to save the two captured traces showing overlap, this is with a Linear Phase BW crossover inserted.

Ah, okay; the reason I thought you had coherence blanking jacked way up is because your crossovers are super steep and pass-band overlap is tiny.

It's been a little while; let's see if I still remember how to do this. (If I don't, I know someone will be along shortly to correct me; my apologies in advance.)

It looks like your crossover frequency is ~1.26 kHz. In the area of overlap between the two pass-bands (i.e. where there's less than a 10 dB level difference between the two mag traces) it looks like the LF is leading by ~15 degrees, and the HF is lagging by ~165 degrees.

To time-align the two passbands you'll need to add delay to the LF; specifically, you'll need to delay the LF enough to produce ~210 degrees of phase shift at the crossover frequency (15 degrees to go from leading to in time, plus 180 degrees to produce a wrap, plus another 15 degrees to go from 180 to 165), which works out to ~0.46 ms.

Does that sound about right?

-Russ
Logged

Helge A Bentsen

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1753
  • Oslo, Norway.
Re: Combining drivers
« Reply #5 on: March 03, 2021, 08:34:09 AM »

Ah, okay; the reason I thought you had coherence blanking jacked way up is because your crossovers are super steep and pass-band overlap is tiny.

It's been a little while; let's see if I still remember how to do this. (If I don't, I know someone will be along shortly to correct me; my apologies in advance.)

It looks like your crossover frequency is ~1.26 kHz. In the area of overlap between the two pass-bands (i.e. where there's less than a 10 dB level difference between the two mag traces) it looks like the LF is leading by ~15 degrees, and the HF is lagging by ~165 degrees.

To time-align the two passbands you'll need to add delay to the LF; specifically, you'll need to delay the LF enough to produce ~210 degrees of phase shift at the crossover frequency (15 degrees to go from leading to in time, plus 180 degrees to produce a wrap, plus another 15 degrees to go from 180 to 165), which works out to ~0.46 ms.

Does that sound about right?

-Russ

Almost exactly what I did :) 0.44ms delay on LF if I remember correctly.

Sounds perfectly all right, both on and off-axis. However, I end up with one phase wrap through the crossover, is there a way of doing this without this wrap?

Logged

Mark Wilkinson

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1104
Re: Combining drivers
« Reply #6 on: March 03, 2021, 10:43:48 AM »

Almost exactly what I did :) 0.44ms delay on LF if I remember correctly.

Sounds perfectly all right, both on and off-axis. However, I end up with one phase wrap through the crossover, is there a way of doing this without this wrap?

It looks like you are using linear phase xovers, given how steep the mag response is, and that the phase curves are still relatively flat. Yes? No?

If yes, the remaining phase curvature is from the drivers' natural mag rolloff. 
So the way to get rid of that wrap, is to flatten the drivers' out of band response before appling the xovers.

Do this with minimum phase EQ's.
I like to try to get mag and phase flat out-of-band down to the frequency where they contribute -30dB to summation. But -20 dB is fine.

It's kinda cool how raw driver mag and phase mirror each other over a horizontal axis.
IOW, if mag is a frowny face (almost always); phase will be a smiley face.  Flatten one, the the other flattens too.

I's also cool how that steep linear phase xovers  take care of what would normally be dangerous excursion from out-of-band boost.

It's truly a have your cake and eat it too scenario, imho. 
As long as you're working with xover frequencies that don't require too much FIR latency for using the linear phase xovers.

If you're not using lin phase xovers with what you posted...please flush all this down the crapper!!!  And quick haha
Logged

Russell Ault

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2484
  • Edmonton, AB
Re: Combining drivers
« Reply #7 on: March 03, 2021, 02:28:29 PM »

{...} So the way to get rid of that wrap, is to flatten the drivers' out of band response before appling the xovers. {...}

I'm sure Mark's answer is better than mine; my inclination would be to do what you've already done and then apply processing to the pre-crossover signal to offset the phase wrap.

In FIR a single "maximum phase" all-pass filter should do the trick. In IIR (i.e. if you're John Meyer) you can achieve the same effect using a series of second-order minimum phase all-pass filters (and some patience).

-Russ

ETA: Of course it should be noted that, while there are many valid reasons for wanting speakers with a flat phase response, a phase wrap at the frequencies in question will be entirely inaudible to humans.
« Last Edit: March 03, 2021, 02:37:28 PM by Russell Ault »
Logged

Helge A Bentsen

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1753
  • Oslo, Norway.
Re: Combining drivers
« Reply #8 on: March 03, 2021, 03:07:38 PM »

Ok, I'll try to get outside next week and refine my pre-alignemt EQ and see what happens.
IIRC it's a linear phase brick wall filter, 1.2Khz/79db/oct.


Thanks :)
Logged

Russell Ault

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2484
  • Edmonton, AB
Re: Combining drivers
« Reply #9 on: March 03, 2021, 06:39:15 PM »

{...} If yes, the remaining phase curvature is from the drivers' natural mag rolloff.  {...}

So, I've just been having another think about this, and I feel like there must be more to this alignment than just the drivers' natural roll-off. If this speaker uses any kind of an HF waveguide (horn, etc.) there's almost certainly a physical offset between the HF and LF, and compensating for this requires some kind of time-domain solution, wouldn't it? Without knowing the speaker's design it's difficult to say for sure, but I still feel like the only way for this speaker to have a "pc75" phase response is going to involve delaying the LF and then dealing with the resulting wrap one way or another.

-Russ
Logged

ProSoundWeb Community

Re: Combining drivers
« Reply #9 on: March 03, 2021, 06:39:15 PM »


Pages: [1] 2 3  All   Go Up
 



Site Hosted By Ashdown Technologies, Inc.

Page created in 0.03 seconds with 21 queries.