Author Topic: Time alignment of drivers - is it worthwhile?  (Read 335 times)

Offline fuz940510

  • Trade Count: (+3)
  • AVForums Veteran
  • ***
  • Posts: 581
Time alignment of drivers - is it worthwhile?
« on: March 19, 2017, 07:45:37 PM »
Hi everyone

Not sure if this is a quick question or not, but here goes:

Is it worthwhile to time align drivers in a 3-driver, 3 way speaker? As far as i understand it is to mitigate the vertical lobing that happens with multiple drivers, as well as get the phase tracking to line up. I can follow the vertical lobing part, but phase tracking is still unknown to me.

I've found a design by Troels called the Ellam Flex 3W that uses a very similar host of drivers to what i currently have hidden in a cupboard. I have the Seas W21EX001, SS 15M/4531K00 and SS D3004/66000, whereas his design uses the 22W/4851T00 and 15W/8530-K00. Tweeters are the same model.

Seeing as the 15M looks to have a similar cone geometry, if not identical, to the 15W, i thought it would be useful to emulate his design. I was contemplating using a similar crossover order and frequency

Only problem is that the stepped baffle leads to a damn ugly looking speaker. Doesn't appeal to my design sense at all, and they need to look half decent if they're going to be occupying the same living space as I am. I prefer them to look a touch more organic - more curves, less hard-edge geometry.

Thoughts/experiences/opinions?

Kean
If you want to be taken seriously, always check your fly.

Offline bbe22

  • Trade Count: (+1)
  • AVForums Veteran
  • ***
  • Posts: 824
Re: Time alignment of drivers - is it worthwhile?
« Reply #1 on: March 20, 2017, 03:06:57 PM »
So................plenty views but no feedback
Flex 3W   ugly speaker    I have to agree
If you do not have the exact compliment of drivers and exact cab you cannot use the published X over.
You are better off to design the cab to your liking  and then have someone redo the X over to suit your drivers. It the drivers allow you can still aim  for LR2 as a target ?
Time align is useful sometimes  but not a prerequisite for a good design
example of another style of cab for 3 way ....

Also you may want to check T/S on the mid  / bass drivers  just to make sure they measure well, before you start the build , (because these have been in storage for some time)

Offline fuz940510

  • Trade Count: (+3)
  • AVForums Veteran
  • ***
  • Posts: 581
Re: Time alignment of drivers - is it worthwhile?
« Reply #2 on: March 20, 2017, 03:22:49 PM »
Thanks bbe22  :thumbs:

The mid and tweeter have never been played, and sourced from Gearslave about 1.5 years ago - bass driver has seen some years though.

Good to know that time alignment isn't one of those miracle, fix-all solutions that can successfully polish a turd.

I wasn't planning on using the published crossover, more of using it as a possible starting point for my own experimentation.

I haven't had the chance for trying any filters so far, so from a purely academic and theoretic point of view, i like the idea of an LR2 filter. Seems like an elegant way of solving the component cost/power handling/musicality issue. I am definitely looking to challenge that pre-conception though
If you want to be taken seriously, always check your fly.

Offline bbe22

  • Trade Count: (+1)
  • AVForums Veteran
  • ***
  • Posts: 824
Re: Time alignment of drivers - is it worthwhile?
« Reply #3 on: March 20, 2017, 07:45:31 PM »
My comment re drivers is because I also have a pair of NewOldStock Seas Excel 8  inchers (not your model though) and their rubber surround have deteriorated (harder / brittle) despite never been used, and that changes the T/S parameters quite a bit. I am still contemplating if I should use them as is or replace surround first , if available...   Other drivers seem to age very well :)

Offline fuz940510

  • Trade Count: (+3)
  • AVForums Veteran
  • ***
  • Posts: 581
Re: Time alignment of drivers - is it worthwhile?
« Reply #4 on: March 20, 2017, 09:50:37 PM »
I'll definitely get them checked out.

When I got the bass drivers i took them over to Ian to check for scraping voice coils etc, and the only problem he could find was that the frames near the mounting holes were bent, like one of the previous owners had overtightened the bolts  - otherwise perfectly fine. The bends should be easy enough to fix, with a delicate but firm hand
If you want to be taken seriously, always check your fly.

Offline Air

  • Commercial Member
  • Trade Count: (0)
  • ***
  • Posts: 474
  • Music lover and involved in industry
    • Airmusic
Re: Time alignment of drivers - is it worthwhile?
« Reply #5 on: March 21, 2017, 10:11:59 AM »
Just a bit of general input from my side. I tend to find that every little bit helps to get closer to that elusive magic SQ but nothing on its own break the camel's back. (if this makes sense). Time alignment, phase tracking etc are all worthwhile goals to pursue but on their own don't guarantee a good design. Often these factors are intricately linked to each other and with other aspects. A slanted baffle might bring a bit of time alignment at a certain frequency but the drivers will be listened to off axis, if this is not compensated for, it could introduce some unintended consequences.

However, I found the better the phase tracking with a particular typology is the more focus and almost "neatness" in total SQ I perceive. It is almost as if everything just snaps into place, like imaging, harmonics etc. 

Also, consider placing the midbass above the tweeter to get some form of time alignment or even consider to offset the tweeter horizontally to get the same effect but less so than with the tweeter below the midbass(designing to listen off-axis can help then even a bit more).


Offline fuz940510

  • Trade Count: (+3)
  • AVForums Veteran
  • ***
  • Posts: 581
Re: Time alignment of drivers - is it worthwhile?
« Reply #6 on: March 21, 2017, 10:31:33 AM »
Just a bit of general input from my side. I tend to find that every little bit helps to get closer to that elusive magic SQ but nothing on its own break the camel's back. (if this makes sense). Time alignment, phase tracking etc are all worthwhile goals to pursue but on their own don't guarantee a good design. Often these factors are intricately linked to each other and with other aspects. A slanted baffle might bring a bit of time alignment at a certain frequency but the drivers will be listened to off axis, if this is not compensated for, it could introduce some unintended consequences.

However, I found the better the phase tracking with a particular typology is the more focus and almost "neatness" in total SQ I perceive. It is almost as if everything just snaps into place, like imaging, harmonics etc. 

Also, consider placing the midbass above the tweeter to get some form of time alignment or even consider to offset the tweeter horizontally to get the same effect but less so than with the tweeter below the midbass(designing to listen off-axis can help then even a bit more).


Thanks for the input  :2thumbs: The Ellam Flex 3W design does place the midrange above the tweeter, and also moves the mid 19mm further forward.

I'm not too sensitive to imaging, however the "neatness" you describe is definitely something i'm interested in.

Sum of parts being greater than the whole is kind of what i'm going for. I've been faffing around with a design for these drivers for almost 3 years now. I get a design that i think i'm happy with, then i read something else, and the whole cycle starts again...
If you want to be taken seriously, always check your fly.

Offline Timber_MG

  • Trade Count: (+2)
  • AVForums Veteran
  • *****
  • Posts: 6,313
Re: Time alignment of drivers - is it worthwhile?
« Reply #7 on: March 21, 2017, 10:34:55 AM »
Time alignment/phase tracking needs to be balanced over a range of angles, it's not simply a matter on axis. With the offset between a baffle mounted tweeter and mid-woofer the group delay is still well under the audable shift so phase tracking/coherency is as much of an issue with integration on as it is off axis.

When you start adding more significant time offsets (read group delay) the matter does become a very real concern. One might make it phase coherent with say integer multiples of 180 degrees phase shift, but the precedence effect starts to become distracting and even fatiguing, especially where the tweeter leads the rest.