Announcement

Collapse
No announcement yet.

How important is a roundover on the top and bottom?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Zilch
    replied
    Re: How important is a roundover on the top and bottom?

    DXT has elements of a constant-directivity waveguide.

    Perhaps Mark would post his polars here?

    [I don't have access to Praxis, alas.... :( ]

    Leave a comment:


  • jkim
    replied
    Re: How important is a roundover on the top and bottom?

    Originally posted by Zilch View Post
    I think it's worth a direct quote:

    Originally posted by Zaph
    Related to the directivity benefit is the fact that a waveguide is not affected by the baffle it's on. Baffle diffraction is a non-issue compared to a standard dome mounted in an enclosure. In a waveguide system, you will not see the lower treble "ripple response" present in typical box systems.
    I don't think this is any surprise considering what a waveguide/horn does. Zaph had empirical evidence as he measured the waveguided 27TDFC on both infinite baffle and a cabinet baffle (Waveguide TMM).

    Jay

    Leave a comment:


  • dlr
    replied
    Re: How important is a roundover on the top and bottom?

    Originally posted by Dirk View Post
    I don't know if I dare to tread where giants walk, but one thing I haven't seen mentioned is how the baffle diffraction effects change as one moves off-axis. When working with the BDS (and in my own measurements), I've noted that the majority of the on-axis hash reduces or goes away when the measurement is taken even just 10 degrees off axis.

    It seems to me that a speaker utilising a tweeter with good dispersion would allow you to ameliorate the worst of these issues just by fiddling with the amount of toe in. I'm not the horney-waveguidey type, but that dispersion is probably another advantage to them.
    The advantage to using diffraction control (chamfers/roundovers/felt) is that as one moves off-axis, the variability may be greatly reduced. If you design for some specific off-axis with no diffraction control, then you may be locked into that axis for optimal response. Flexibility of room arrangement may be more limited.

    Another problem with designing on any specific axis, whether on-axis or at 10 deg. or some other, is that there is the power response that should be considered. When that is considered, the response on some single axis may be somewhat at odds with a smoother power response.

    Use of diffraction control helps in optimizing for a wider listening window (more flexibility as to axis) and may help the power response as well. But if for nothing else, it makes design easier due to that reduction in off-axis variability.

    dlr

    Leave a comment:


  • Dirk
    replied
    Re: How important is a roundover on the top and bottom?

    I don't know if I dare to tread where giants walk, but one thing I haven't seen mentioned is how the baffle diffraction effects change as one moves off-axis. When working with the BDS (and in my own measurements), I've noted that the majority of the on-axis hash reduces or goes away when the measurement is taken even just 10 degrees off axis.

    It seems to me that a speaker utilising a tweeter with good dispersion would allow you to ameliorate the worst of these issues just by fiddling with the amount of toe in. I'm not the horney-waveguidey type, but that dispersion is probably another advantage to them.

    Leave a comment:


  • Zilch
    replied
    Re: How important is a roundover on the top and bottom?

    Looks like Zaph agrees also:

    http://techtalk.parts-express.com/sh...50#post1615850

    Leave a comment:


  • Zilch
    replied
    Re: How important is a roundover on the top and bottom?

    http://sound.westhost.com/articles/waveguides1.htm

    Leave a comment:


  • charlielaub
    replied
    Re: How important is a roundover on the top and bottom?

    Originally posted by Zilch View Post
    First rule of thumb is that the distances should be unequal.
    I'll second that. I don;t know of any "golden ratio/rule" kind of thing. Depending on the baffle size (width and height) and position there will be some ripple. The amplitude, number and positions of the ripple(s) changes as you move the tweeter around. Sometimes, it is better to live with one ripple/hump that you can easily clean up with a notch filter than to try an minimize everything.

    -Charlie

    Leave a comment:


  • Zilch
    replied
    Re: How important is a roundover on the top and bottom?

    First rule of thumb is that the distances should be unequal.

    With waveguides, the edges are less of an issue than their own mouth contours, and perfect axisymmetric round ones generate an on-axis notch, hence the Geddes "I don't listen on axis" syndrome.... ;)

    Leave a comment:


  • rogoll
    replied
    Re: How important is a roundover on the top and bottom?

    Originally posted by charlielaub View Post
    That's right, your component count would quickly climb. But there are other ways to better the situation, namely via some diffraction modeling that is done before you cut holes for the drivers. By moving the drivers around on the baffle (as space permits) you can sometimes find a location that has significantly smoother response, e.g. the higher frequency squiggles are minimized. You are then left to deal with one main hump, which can be notched out or perhaps integrated in to the crossover overlap region, depending on its frequency and breadth.

    -Charlie
    Are there any guidelines or rules of thumb that you can give for positioning the tweeter, other than just moving it around somewhat randomly? In his book, Ray Alden suggests making the distance to one edge an imaginary number, but of course he was a math teacher by profession.

    Leave a comment:


  • charlielaub
    replied
    Re: How important is a roundover on the top and bottom?

    ...What I was referring to, was all the little squiggly stuff in the treble region. You'll have a real tough time crossovering all of that out.
    That's right, your component count would quickly climb. But there are other ways to better the situation, namely via some diffraction modeling that is done before you cut holes for the drivers. By moving the drivers around on the baffle (as space permits) you can sometimes find a location that has significantly smoother response, e.g. the higher frequency squiggles are minimized. You are then left to deal with one main hump, which can be notched out or perhaps integrated in to the crossover overlap region, depending on its frequency and breadth.

    -Charlie

    Leave a comment:


  • replied
    Re: How important is a roundover on the top and bottom?

    I wasn't talking about one big noticeable midrange hump. What I was referring to, was all the little squiggly stuff in the treble region. You'll have a real tough time crossovering all of that out.


    Originally posted by charlielaub View Post
    Sure, why obsess over it?? Who cares about that exaggerated upper midrange, anyway? BTW, point sources would be the worst in terms of diffraction effect, because the smaller the driver the worse the diffraction, 2-inch drivers and smaller, especially.

    Actually, this is one of the main reasons why you are supposed to take actual measurements of your drivers in the cabinet before designing the crossover - because you will then be able to compensate for the diffraction peaks in the crossover network. This is what all good designers do.

    Diffraction peaks are definitely audible. Most every cabinet has them except a perfect "sphere".

    -Charlie

    Leave a comment:


  • Zilch
    replied
    Re: How important is a roundover on the top and bottom?

    Originally posted by spasticteapot View Post
    Do I understand this correctly?
    I believe so, yes.

    I build with waveguides almost exclusively, and consider baffle step a separate issue, as Linkwitz suggests....

    Leave a comment:


  • charlielaub
    replied
    Re: How important is a roundover on the top and bottom?

    RE roundovers and tweeter response - I think that someone posted earlier in this thread that the roundover has to be pretty large to be effective, so even a 3/4" roundover might not have much effect at 2kHz. Make sure that you flush mount the tweeter, as the reflection from the edge of the faceplate will cause ripples in the response.

    RE your design and Edge: looks good. You are correct, it is pretty smooth up to 1.5kHz so that should work well. You might want to model the tweeter location as well, and try to minimize the ripple amplitude above 1.5 kHz by adjusting the location until you get the best looking response. Also, Edge does not take in to account the fact that as some frequency the off-axis response starts to become less intense than the on axis response (e.g. the driver starts to "beam"). You can look at the tweeter's off axis response plot (if available) and see if you can discount the Edge's prediction of response irregularities, e.g. if the response is down 6dB at 45degrees off axis then any predicted irregularities at that frequency are probably not really going to happen because the cabinet edge is not sufficiently "illuminated" compared to the on-axis response.

    -Charlie

    Leave a comment:


  • spasticteapot
    replied
    Re: How important is a roundover on the top and bottom?

    Originally posted by charlielaub View Post
    From my experience (with direct radiator drivers and not horns) there are two main effects of diffraction:
    1. the baffle step
    2. a hump near where the baffle step starts to drop off, e.g. around 1 kHz
    I'm not arguing that these are an issue - far from it. However, I thought that roundovers were more of an issue for tweeter response - wouldn't a roundover be too small to make much of a difference below 2khz? (I'm a bit confused.)

    Incidentally, while I'm new to speaker design, I have taken a look at the EDGE, and it gives me pretty promising results. This is what I got when I put a 180mm driver on a baffle 8" x 12":



    Considering that the net result will be used with an active crossover, it seems like it's a pretty usable result - everything is nice and smooth well past the planned ~1.5khz crossover point. Or am I looking at it wrong?

    Leave a comment:


  • charlielaub
    replied
    Re: How important is a roundover on the top and bottom?

    Originally posted by spasticteapot View Post
    I'm aware that it's a well-documented phenomenon, but nobody seems to agree on what to do about it. Depending on who you ask, the best solution is a spherical baffle, a heavily padded baffle, a baffle with big roundovers, or just ignoring the issue because there's nothing you can do about it.

    I guess the question is: How do I apply this data in my projects?



    I've wondered if this was true. Perhaps Zaph has some data on the subject?
    From my experience (with direct radiator drivers and not horns) there are two main effects of diffraction:
    1. the baffle step
    2. a hump near where the baffle step starts to drop off, e.g. around 1 kHz

    There is lots of material out there about the baffle step. What I see is a lack of much mention of the hump that remains after the baffle step transition has been compensated for. Let me illustrate:
    response of driver in box, no baffle step compensation:


    response of driver in box, with baffle step compensation (blue line):


    As you can see from the lower plot, even after you use baffle step compensation, you are left with a 2dB peak that is more than an octave wide. You can bet that this will be audible. This is typical behavior for direct radiator drivers that are not large compared to the baffle width.


    Here is another example. This time the driver diameter is small (2 inches, e.g. RS52 mid dome):
    response of driver in box, no baffle step compensation:


    response of driver in box, with baffle step compensation (blue line):


    The smaller diameter driver has accentuated the diffraction "oscillations" that were more damped in the first example that used a larger diameter driver. If this 2-inch diameter driver has strong off axis energy (and the RS52 does) then this effect WILL occur and the response will have similar peaks and valleys. This will be audible.

    So, I hope I have illustrated my point. I am not saying that this occurs for every combination of baffle width, shape, driver diameter and so on, but for the majority of cases involving a driver mounted in the center of the baffle of a rectangular box, you really need to take these effects in to account if you want to end up with a good sounding speaker. IMHO. So I really must disagree with the statement from Dickason posted by Zilch.

    You may feel differently. That's fine.

    -Charlie

    Leave a comment:

Working...
X