Hey, slow down... I corrected my response later. I just thought you were talking about the normal foam you use.
I have no idea where you get the "blood sucking scoundral", that was out of no where.
Just a gut reaction to the ongoing claims that all I am interested in is lining my own pockets. Sorry if this does not apply to you.
Nope, I think you spend a lot of time posting because of your passion for audio science.
I just read your post wrong and I was assuming the foam is the common foam discussion I have wrt to waveguide builds.
The foam over the woofer is definitely interesting.
Sometimes different backgrounds might have some expressions that give different feelings to others of different backgrounds, it's good to try and look on the bright side like buddy buddy type talk.😉
I am not talking about any particular woofer, just all woofers in general. getting out to 90 degrees is the only way I know to be sure how complicated your crossover is going to have to be. Further might be better still. The break up of the woofer can change dramatically at these angles in unpredictable ways and interfere with your crossover even when it doesn't on axis or at 45 degrees. Without that information, you're just guessing. That's why I'm hoping to find a correlation to the CSD waterfall plot. Perhaps a better CSD may make for better off axis performance. That's what I'm hoping for, but not betting on.
I think that foam is the most interesting idea I've heard for a while. Wish I had thought of it. I'd bet it would need to be more dense, but how much more is for someone smarter than I to figure out. It would probably take some experimentation. How much more dense, how thick, and of what shape......? Interesting stuff. Another type of low pass filter. Maybe the good Dr. will work that out for us and give it away to us less fortunate.😀
Dr. Geddes, any chance of you posting some of that detailed information for those of us who would love to know? I'm particularly interested in the CSD and polar data. I'll show you all mine when I get it if anyone is interested.
Which brings me to my next point: What mic distance would be best for a waterfall plot?
Thanks guys! This continues to be an interesting thread.
Dan
I'm not a fan of CSD, doesn't tell you very much. But I will try and post the woofers that I have.
Hi Dan,
what exactly are you looking for in the off axis measurements? I assume you are looking at the individual drivers off axis behaviour. are you looking for (for want of a better phrase) non linear changes in the freq response off axis? ie maybe the curves track each other well being different by say 3db at all freq's past the point where the driver becomes directional, but then all of a sudden at a particular freq there is a change of different magnitude, for example 6db?
I've done some horizontal off axis measurements of my drivers at 15 degree intervals out to 45 degrees (I had trouble past that due to room reflections). I'm just not sure what it is I should be looking at, with respect to making decisions about my crossover 🙂 They were pretty non-directional out to about 4K (and I'm now crossing 2nd order acoustic at 3K at the moment) so wasn't sure if I needed to worry about it at all.
Also another thing I wasn't sure about when doing off axis was the relative shift of the drivers when rotating the cabinet. I simply rotated on the axis of the middle of the cabinet leaving the mic in the same place, but when the box goes past 45 degrees the mic is pointing at the side of the box rather than at the driver (due to the arc the front baffle follows), so I wasn't sure if I should move the whole stand back to compensate (probably need to re-read my d'appolito!!).
I don't have a graph with the individual drivers handy, but I do have one with the very first "crossover" which consisted of a single 4.3uf cap on the tweeter, and the midbass' running full range. This will show what I mean about being pretty non-directional to about 4K. My measurements have improved since then, and so has the crossover (I think!) 🙂
regards,
Tony.
what exactly are you looking for in the off axis measurements? I assume you are looking at the individual drivers off axis behaviour. are you looking for (for want of a better phrase) non linear changes in the freq response off axis? ie maybe the curves track each other well being different by say 3db at all freq's past the point where the driver becomes directional, but then all of a sudden at a particular freq there is a change of different magnitude, for example 6db?
I've done some horizontal off axis measurements of my drivers at 15 degree intervals out to 45 degrees (I had trouble past that due to room reflections). I'm just not sure what it is I should be looking at, with respect to making decisions about my crossover 🙂 They were pretty non-directional out to about 4K (and I'm now crossing 2nd order acoustic at 3K at the moment) so wasn't sure if I needed to worry about it at all.
Also another thing I wasn't sure about when doing off axis was the relative shift of the drivers when rotating the cabinet. I simply rotated on the axis of the middle of the cabinet leaving the mic in the same place, but when the box goes past 45 degrees the mic is pointing at the side of the box rather than at the driver (due to the arc the front baffle follows), so I wasn't sure if I should move the whole stand back to compensate (probably need to re-read my d'appolito!!).
I don't have a graph with the individual drivers handy, but I do have one with the very first "crossover" which consisted of a single 4.3uf cap on the tweeter, and the midbass' running full range. This will show what I mean about being pretty non-directional to about 4K. My measurements have improved since then, and so has the crossover (I think!) 🙂
regards,
Tony.
Attachments
I know what you mean, but I'm hoping to find something useful from it. It's just a hunch that may well be wrong. I truly enjoy learning when my hunches are wrong. It always seems to lead to better things. No one really seems to do what I've found to be the most useful measurements and/or posts them online, so I would appreciate your data immensely. If I can find something that correlates to something more useful it would be a big sigh of relief.
Thank you,
Dan
I'm looking for a low magnitude break up on all axis.
Here's what I'm talking about:
That's 11.25 degrees off axis out to 90 degrees in 11.25 degree steps. Nasty stuff.
Here's what the textbook LR4 filter would look like with that woofer:
Definitely not ideal.
This is how that improved after a notch filter was added:
Here's what that response would look like w/o the notch crossover to a waveguide:
A picture is worth a thousand words--in this last case, a thousand curse words.🙂
Now look at the response with that notch filter in place:
A step in the right direction.
Dan
CSDs do not show anything off axis unless a CSD is done from an off axis measurements.
I am curious to where you are getting all your info because I have never read any papers talkinga about or caring about 90degrees response of a woofer. 90degrees is not as meaningful as 40 degrees, you are better off designing for 40 degrees and less. Woofers have zero off axis different down low anyways and we are always crossing over before they beam so 90 degree measurements are pointless. Like you Im still learning so Im hoping there is some good links to read on this.
Last edited:
Thanks Dan, the graphs do show the story 🙂 I went and checked and I only did off axis measurements of the whole system not the individual drivers as I had thought, I also discovered that at 30 degrees off axis they are a lot more directional that that graph I posted I think I didn't do a very good job on those measurements 😉
Tony.
Tony.
Thats not the case for me at all. It's the beaming of the woofer that I am specifically looking for and that it beams while still maintaining a decent response on and off axis. Going to 90 degrees is useful I think (SL complains because I don't show to 180 degrees! Oh well.)
But you look for the -6dB @ 45deg, no?
and you look how it matches your waveguides, correct.
I guess I just think that we have a hard enough time with 45 degree matching and I can not see how 90deg below 1K is going to matter compared to all other things we need to worry about. Its almost impossible to be perfect and if the focus is on 90 so much then other important factors will be missed.
Last edited:
To match at 45 I need data well beyond that, right? But yes below 400-500 Hz its meaningless beyond 45, but I plot data all the way up to 20 kHz and its not meaningless there.
Doug, read post 228 and 229. Answers all your questions. You won't find it in any links I'm aware of, but there really aren't any links on designing a CD loudspeaker. Most audio geeks just don't realize how necessary CD is as they have never heard it. Maybe Pi Loudspeakers might have something, but after read 228 and 229 it will be obvious. There is no book telling you what is in my opinion the most important stuff. Loudspeakers have been holding back listening for too many years.😉
Dan
Yea, it's not below 1kHz that I'm worried about. I won't miss more important stuff just because I have more data. Neither will you.😎
dantheman:
gedlee (from Defo's horn loaded dome tweeter thread):
Thanks dantheman for pursuing the details of the impulse response. I doubt we're the only two who want to know more.
Continuing thanks Earl.
gedlee (from Defo's horn loaded dome tweeter thread):
Thanks dantheman for pursuing the details of the impulse response. I doubt we're the only two who want to know more.
Continuing thanks Earl.
Interesting note, Im taking apart some Tannoy V8s to look at building better XOs for them and I notice the impulse of the tweeter was really jagged but that should also show up in the CSDs or wavelets.
Geddes response was to me in that thread, I still wonder if that 'tail' effects of the impulse are also in the CSD.
Geddes response was to me in that thread, I still wonder if that 'tail' effects of the impulse are also in the CSD.
Ah, sorry...when you posted you wanted more data I thought it was a discussion on woofers.
You won't miss important stuff but you may focus on the wrong things for the wrong reasons. I doubt there will ever be a 90 deg issue that isnt handle by fixing other off axis issues first (the 0 to 45 deg issues).
Last edited:
- Status
- Not open for further replies.