Another thing that I would point out is that absorbent materials such as plastic foams have non linear with frequency absorption that is usually rated as for instance so much attenuation per meter.
If as has been stated in these pages that we are looking to absorb waves with three times the path length, i.e. for an o.s. device around 0.6 metres, then if the forward path attenuation is 3db. the total attenuation of the unwanted part is 6db. higher than this.
The delay of a 0.6metre path length is 5.8 milliseconds, in a shallow device the, such as the one I have described the path length is 0.066 metres giving a delay of 0.2 milliseconds.
Rcw.
If as has been stated in these pages that we are looking to absorb waves with three times the path length, i.e. for an o.s. device around 0.6 metres, then if the forward path attenuation is 3db. the total attenuation of the unwanted part is 6db. higher than this.
The delay of a 0.6metre path length is 5.8 milliseconds, in a shallow device the, such as the one I have described the path length is 0.066 metres giving a delay of 0.2 milliseconds.
Rcw.
pooge said:
Yes - your answer was correct - but only as a statement of over-generalism.
In the very context of this thread its all about Earls claims and explanations to do transition of wave front most smoothly ever possible by the OS contour - and in *this* context - your statment was not helpfull.
Michael
Finally got to do some measurements. Here is a comparison between BEM sim in infinit baffle and measurement free standing.
BEM shows up to 45 deg, while the measurements show 60deg. Both in 7.5 deg intervals.
I have just started a round of BEM sim to see if free standing simulation is possible.
An externally hosted image should be here but it was not working when we last tested it.
BEM shows up to 45 deg, while the measurements show 60deg. Both in 7.5 deg intervals.
I have just started a round of BEM sim to see if free standing simulation is possible.
The source here is a compression driver? At any rate those are quite impressive results.
I would ask to see the data on a more reasonable scale since 90 dB is just not usable. It appears like once equalized there will be a 4-6 dB hole at about 4 kHz. Quite consistant with what I always find.
I would ask to see the data on a more reasonable scale since 90 dB is just not usable. It appears like once equalized there will be a 4-6 dB hole at about 4 kHz. Quite consistant with what I always find.
Many thanks for sharing the measurenemts of your contour too, soongsc. Very much appreciated!
Could you do a "normalised to 0deg" plot in SE as well?
I would be very interested to compare measurements of yours with mine (dipole configuration)
hope to gain some further insight about possible differencies of my cylinder wave horn to a standard horn.
Michael
Could you do a "normalised to 0deg" plot in SE as well?
I would be very interested to compare measurements of yours with mine (dipole configuration)
hope to gain some further insight about possible differencies of my cylinder wave horn to a standard horn.
Michael
On a lighter note:
I was passing by some of the new style "basin sinks" in a home improvement store today. You know, they type that look like a bowl sitting on the counter top.
I thought "hey!, those might make a cool waveguide." Some even had a nice flange. But then I took a look at the prices. Yikes!
They make horns and waveguides look like a bargain. Glass, marble, ceramic, metal - they are all $$$. Just crazy for a mass produced item.
I was passing by some of the new style "basin sinks" in a home improvement store today. You know, they type that look like a bowl sitting on the counter top.
I thought "hey!, those might make a cool waveguide." Some even had a nice flange. But then I took a look at the prices. Yikes!
They make horns and waveguides look like a bargain. Glass, marble, ceramic, metal - they are all $$$. Just crazy for a mass produced item.
Im renovating my house, and I have followed what there is of kitchen and bathroom sinks etc
Every time a new design enters the scene it can be exstremely expencive
And the the next year its available much cheaper
Different finish, but same design
At one time they they showed how much of expencive fancy bathroom stuff are made in china from scrap metals containing lead, nickel and all sorts of nasty stuff, and cost next to nothing
Until they arrive at our doorstep costing a small fortune
Every time a new design enters the scene it can be exstremely expencive
And the the next year its available much cheaper
Different finish, but same design
At one time they they showed how much of expencive fancy bathroom stuff are made in china from scrap metals containing lead, nickel and all sorts of nasty stuff, and cost next to nothing
Until they arrive at our doorstep costing a small fortune
panomaniac said:So true, so true. I used to think audio was an expensive hobby - until I saw the prices of home decor and prices women pay for "designer" clothes! Makes Hi-Fi seem resonable.
Quite true. Lidia happens to be a frugle buyer who never wears the fashion. She also doen't go in for the "decorating" mode of home furnishings. Makes my hobby harder to justify. Thats why I call it my job!
The source is the same as with all the other measurements, a concaved dome direct radiating tweeter. Here is a simple equalized version of measurement(cap in series with some zobel components), but turning the waveguide in a opposite direction.gedlee said:
An externally hosted image should be here but it was not working when we last tested it.
This is the scale I normally look at as a standard unless there is anything in particular detail I'm looking for. Is there anything you suspect that might be hidden by the scale settings?
No, there is no way to set referrence to 0db in SE that I know of. I actually find that much more confusing because there is no way to equally set two measurements to 0db reference in a comparible way. But I think if you change your scale to 5db per tic as I use, it's quite comparible. Note the two notches in the highest frequency? Seems to be a similar trend.mige0 said:Many thanks for sharing the measurenemts of your contour too, soongsc. Very much appreciated!
Could you do a "normalised to 0deg" plot in SE as well?
I would be very interested to compare measurements of yours with mine (dipole configuration)
hope to gain some further insight about possible differencies of my cylinder wave horn to a standard horn.
Michael
Here is a summary of measurements I've posted. Bottom graph is the earliest. All 0~45 deg in 7.5 deg increments. Latter three are about the same size.
An externally hosted image should be here but it was not working when we last tested it.
In some of my recent measurements, I tend to find a funny reflection that equates to about 75cm distance from the main signal as shown below.
Guess what! It equates almost exactly to double distance from mic tip to mike clamp. I forgot to put on an acoustic shield over the clamp. Lynn's way of setting up the mic is really a better way.
An externally hosted image should be here but it was not working when we last tested it.
Guess what! It equates almost exactly to double distance from mic tip to mike clamp. I forgot to put on an acoustic shield over the clamp. Lynn's way of setting up the mic is really a better way.
soongsc said:
Excellent structured work IMO, soongsc.
There are some observations that strike me even if we don't know the very details of your contours and about the driver used.
I have found them in my measurements and simus so far as well.
There are obviousely three reginons that have to be destinguished from each other
1.) the part below roughly 2-3kHz
2.) the part at and above HF peaking (around 20kHz)
3.) the part in between
ad 1 all of your designs clearly excel over your second (second up from bottom) design
ad 2 this differs on how good we can equalize - for advanced equalising your first (first plot up from bottom) and last (last plot up from bottom) design are best as the SPL *difference* of the curves is almost kept constant throughout the peak
ad 3 your third (third up from bottom) desingn is a clear winner
All above statments are meant in the light of optimisation of directivity "smoothing" - and *not* in optimising (smoothing) FR.
The "issues" in the three regions can - and have to be - adressed quite differently when developing a horn (form my noob experiance).
Thanks again - very revealing, I'm looking forward to your next contour and hopefully can contribute some own development soon as well (I already ordered a quad core system to cut down high res / high bandwidth simulation time)!
Michael
The bottom graph (first waveguide), as jokingly described by the shop, is a nice looking baby bathtub Note good to start out with a real speaker that big. But it was helpfull as a first cut at wave guides. Tradeoff between directivity, low frequency loading, size, FR shaping, high frequency filtering, etc. is really a big task and quite complicated. Since I've pretty much set the size around 9" diameter at the lip @ baffle angle, it leaves one less parameter to worry about even though it seems a larger wave guide will be better.
Note good to start out with a real speaker that big. But it was helpfull as a first cut at wave guides. Tradeoff between directivity, low frequency loading, size, FR shaping, high frequency filtering, etc. is really a big task and quite complicated. Since I've pretty much set the size around 9" diameter at the lip @ baffle angle, it leaves one less parameter to worry about even though it seems a larger wave guide will be better.