Speakerplans.com Homepage
Forum Home Forum Home > Plans > Punisher and X-tro
  New Posts New Posts RSS Feed - Xtro flare redevelopement thread
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

Xtro flare redevelopement thread
 Post Reply Post Reply Page  <1234>
Author
Message
  Topic Search Topic Search  Topic Options Topic Options
cravings View Drop Down
Old Croc
Old Croc
Avatar

Joined: 30 January 2007
Location: Ireland
Status: Offline
Points: 7204 		Post Options Post Options   Thanks (0) Thanks(0)   Quote cravings Quote  Post ReplyReply Direct Link To This Post Posted: 23 May 2017 at 9:30pm
to put a phase plug in the original straight horn.. you'd need to totally change the start of the horn to accomodate the plug, therefore the geometry.. it's a begin again project really isn't it?
Back to Top
toastyghost View Drop Down
The 10,000 Points Club
The 10,000 Points Club
Avatar

Joined: 09 January 2007
Location: Manchester
Status: Offline
Points: 10520 		Post Options Post Options   Thanks (0) Thanks(0)   Quote toastyghost Quote  Post ReplyReply Direct Link To This Post Posted: 23 May 2017 at 11:26pm
Originally posted by ceharden ceharden wrote:

Originally posted by gen0me gen0me wrote:

Do you have any directivity characteristics? Basicly any info really appreciated. Im fighting here with abec http://forum.speakerplans.com/abec-problems_topic98514.html
I started even cooling processor with ventilator. Got so hot
Im wondering xt8 was designed by try and makes methods or invented in some software? Or made partly as rough idea in hornresp and than perfected?

Wider openning at the begging may be due to extending top of bandwidth around 2.2k. As I was simulating with hornresp I could go higher with Sentry around 170cm2 than 100cm2


I suspect there was a fair amount of guesswork and no simulation.

I would concentrate on getting the dispersion right, frequency response is easier to fix with EQ...

The original straight Xtro flare, plus phase plug as Markk says would be a reasonable starting point.

When I was thinking about doing something similar, I wanted to try a Quatratic Throat Waveguide:





You can change the response with EQ but only in one position - if your directivity is way off at certain frequencies then you aren't going to fix that with filters, especially if you're not a serious whizz.
Back to Top
Teunos View Drop Down
Old Croc
Old Croc
Avatar

Joined: 23 November 2008
Location: The Netherlands
Status: Offline
Points: 1784 		Post Options Post Options   Thanks (0) Thanks(0)   Quote Teunos Quote  Post ReplyReply Direct Link To This Post Posted: 24 May 2017 at 7:14am
Originally posted by ceharden ceharden wrote:

Originally posted by gen0me gen0me wrote:

Do you have any directivity characteristics? Basicly any info really appreciated. Im fighting here with abec http://forum.speakerplans.com/abec-problems_topic98514.html
I started even cooling processor with ventilator. Got so hot
Im wondering xt8 was designed by try and makes methods or invented in some software? Or made partly as rough idea in hornresp and than perfected?

Wider openning at the begging may be due to extending top of bandwidth around 2.2k. As I was simulating with hornresp I could go higher with Sentry around 170cm2 than 100cm2


I suspect there was a fair amount of guesswork and no simulation.

I would concentrate on getting the dispersion right, frequency response is easier to fix with EQ...

The original straight Xtro flare, plus phase plug as Markk says would be a reasonable starting point.

When I was thinking about doing something similar, I wanted to try a Quatratic Throat Waveguide:





Its funny how many (fundamental) mathematical errors are in that paper.
(I did not click on the link but i assume this is the peavey document?)
Also, if you start drawing out the (correct) equations and throw them into a simulator, you will probably find out pretty quickly why it really isnt all that amazing.
Best regards,
Teun.
Back to Top
ceharden View Drop Down
The 10,000 Points Club
The 10,000 Points Club
Avatar

Joined: 05 June 2005
Location: Southampton
Status: Offline
Points: 11637 		Post Options Post Options   Thanks (0) Thanks(0)   Quote ceharden Quote  Post ReplyReply Direct Link To This Post Posted: 24 May 2017 at 9:00am
I would imagine a QTW can't be worse than just a straight conical waveguide without a smooth transition at the throat?

I did start doing some experiments a couple of years ago with the fibreglass versions of the Xtro flare.  I modified the throat from square to closer to rounded and trying to add Martin Wavefront style 'Phase Balls' although I was only making them from Polystyrene and probably weren't rigid enough.  Loaded with a an 18Sound 8M400 it did work fairly well.  I can't remember what the high extension was like.  I've still got the ones I started modifying if anyone wants to pick up the project because I can't see me doing it any time soon.
Back to Top
Mark James View Drop Down
Old Croc
Old Croc
Avatar

Joined: 15 January 2006
Location: rig side
Status: Offline
Points: 5264 		Post Options Post Options   Thanks (0) Thanks(0)   Quote Mark James Quote  Post ReplyReply Direct Link To This Post Posted: 24 May 2017 at 10:38am
how much?
If you want people to feel the bass then a reflex design is probably best as you get the most air movement per decibel from those.      errrrr yes wtf?
Back to Top
krestudon View Drop Down
Registered User
Registered User


Joined: 23 August 2011
Location: Madrid
Status: Offline
Points: 32 		Post Options Post Options   Thanks (0) Thanks(0)   Quote krestudon Quote  Post ReplyReply Direct Link To This Post Posted: 26 May 2017 at 11:42am
Do you still need measurements on a B&C 8pe21?

I'm available.
Back to Top
gen0me View Drop Down
Young Croc
Young Croc
Avatar

Joined: 20 February 2016
Location: UK
Status: Offline
Points: 999 		Post Options Post Options   Thanks (0) Thanks(0)   Quote gen0me Quote  Post ReplyReply Direct Link To This Post Posted: 01 June 2017 at 12:11am
Thanks krestudon, I have them already.


What inspires me now is this:
http://www.bluearan.co.uk/index.php?id=SONMR8XT&browsemode=manufacturer
Dont know how much of it is true(even italians dont have that nice dispersion chars). Ofc scale is high to blind eye but it should be doeable anyway!



Here are some of my tries:
https://drive.google.com/drive/folders/0B5y23T8xURtGNmZwVlRsM3pmeXM?usp=sharing
Mostly on the blind shoot. I was trying to do hornresp one(that would guarantee similar power on all the bandwidth) with very fast expanding after throat. Ofc hornresp sim has huge circumference value(from 43cm2 to 161cm2 in 3mm) but to do at lest something similar or go this way. Suprisingly this short phase plug that allows fast expansion after throat dont look so good on Abec sims.


Edited by gen0me - 01 June 2017 at 12:12am
Back to Top
Teunos View Drop Down
Old Croc
Old Croc
Avatar

Joined: 23 November 2008
Location: The Netherlands
Status: Offline
Points: 1784 		Post Options Post Options   Thanks (0) Thanks(0)   Quote Teunos Quote  Post ReplyReply Direct Link To This Post Posted: 01 June 2017 at 6:07am
Normalized directivity plots would be way more useful.
Best regards,
Teun.
Back to Top
gen0me View Drop Down
Young Croc
Young Croc
Avatar

Joined: 20 February 2016
Location: UK
Status: Offline
Points: 999 		Post Options Post Options   Thanks (0) Thanks(0)   Quote gen0me Quote  Post ReplyReply Direct Link To This Post Posted: 01 June 2017 at 7:13pm
Thats for 40V Distance almost 4m. Hmm smt doesnt calculate.
Back to Top
Teunos View Drop Down
Old Croc
Old Croc
Avatar

Joined: 23 November 2008
Location: The Netherlands
Status: Offline
Points: 1784 		Post Options Post Options   Thanks (0) Thanks(0)   Quote Teunos Quote  Post ReplyReply Direct Link To This Post Posted: 01 June 2017 at 8:38pm
Normalized means that everything off axis is eqed such that the on axis frequency response in the directivity plots is constant.
This way, you actually see what the horn geometry itself does to the directivity.
If you have a huge peak on axis, it will also show off axis.
Of course you would want to research the reason for this peak and battle it, but shallow trends still give twisted results.
Best regards,
Teun.
Back to Top
gen0me View Drop Down
Young Croc
Young Croc
Avatar

Joined: 20 February 2016
Location: UK
Status: Offline
Points: 999 		Post Options Post Options   Thanks (0) Thanks(0)   Quote gen0me Quote  Post ReplyReply Direct Link To This Post Posted: 13 June 2017 at 10:42pm
Need to figure out how to do it in abec.
I need some help in interpretation of the resaults. Is it possible that vertical and horizontal angles of horns influence dispersion from around 1.6, 1.8 kHz and lower much more important is mouth shape which works kind of against(verticaly wider, horizontaly narrower)? Top printscreens are horizontal, bottom vertical.

Next one.
Back to Top
Teunos View Drop Down
Old Croc
Old Croc
Avatar

Joined: 23 November 2008
Location: The Netherlands
Status: Offline
Points: 1784 		Post Options Post Options   Thanks (1) Thanks(1)   Quote Teunos Quote  Post ReplyReply Direct Link To This Post Posted: 14 June 2017 at 7:52am
Read up on Huygens theory and imagine how it would affect sound.
If the source is small compared to the wavelength, the source will become omnidirectional. If the source is much larger than the wavelength, it will automatically be directional.

Typically you have 3 regimes that control directivity.
  1. Low frequencies where the wavelength is much lower than the size of the horn in the measurement plane. In this area directivity is primarily controlled by horn size and efficiency/frequency response by expansion profile. Since your horn is wider than it is tall, it will be more directive down low in the horizontal plane than it will be in the vertical plane. This is contradictive to what you want if you for instance want a 60Hx40V horn. Below a certain frequency, the horn will become more directive in the plane with the bigger size, in this case the horizontal plane. This is a phenomenon commonly refered to as pattern flip.
  2. Midband frequencies where the wavelength of the sound is roughly equal to the size of the horn. In this area, both directivity and and frequency response are HEAVILY influenced by horn geometry more than it is influenced by the size. Since the wavelength is roughly equal to the horn dimensions, you can get all sorts of reflections and therefor standing waves inside your horn if the mouth does not perfectly terminate in a smooth transition towards the open space. This is because the sound undergoes a change in acoustic impedance and if this change is abrupt, it will cause the sound to be reflected. Usually in this area is where you see edge effects that cause comb filter like effects in the frequency response, and long, slow decaying trails in a waterfall plot. The effects you see in a directivity graph are what is commonly known as midrange waste-banding. 
  3. High frequencies where the wavelength is bigger than the sound. In this region, size does not matter anymore. The primary effects that affect the dispersion is simply the angle of the horn walls, and how well the sound can follow these walls. If you have a horn with a rapid expansion profile, the higher frequencies will ''break'' free from the walls very soon which means that for increasing frequency, the directivity increases and dispersion angles decrease. 
    To make sure that even the highest of frequencies are ''dispersed'' throughout the intented coverage angle, many horn designs use a diffraction slot. If the width of this slot is small enough compared to the wavelength, the slot will be omnidirectional. If the horn angle after that is for instance 90 degrees, the sound will be dispersed throughtout the full 90 degrees. (again, read up on Huygens). The diffraction slot however is a distortion generating beast since like i said in 2, it offers a HUGE abrupt change in acoustics impedance > reflected sound > standing waves > resonances> high distortion. (This is why i love my JABO KH53 horns on the BMS horns, no diffraction slot. These horns sound leagues better than the JBL 2385A i also sometimes use. (See the topic : http://forum.speakerplans.com/2-tractrix-horn-any-interest_topic97925_page1.html ) 
In your horn, which i guess is around 50cm wide, the point where the flip between 1 and 2 occurs is roughly around the 400~500Hz ballpark horizontal and higher for the vertical. This means that you will never, ever with these horn sizes, get a horn that is directive to below this frequency region. This shows in your simulations.
What your simulations also show, is that in this area (500Hz), the directivity decreases, typical midrange waste-banding behavior. I do need to add, that this might be exaggerated a bit since around 700Hz your on-axis sensitivity is higher. What your simulations also show is that this effect is much worse in the vertical plane than in the horizontal plane. This is logical, since your horizontal termination to the open air is WAYYYY more smooth than that for the vertical, which is basically an open ended pipe.

The funniest part of this all to me, is that most people have no idea that these effects exist. If you listen carefully to a sound system, you can spot them though. Especially when you stack multiple horns side by side, things like this can become problematic pretty fast.
It is THE main difference between a professionally engineered cabinet and a home built DIY system. Measuring frequency response never tells you the entire picture, and you have found that out for yourself, kudos!

Maybe this clarifies why i always recommend people, that if you can stick with one speaker per side with a wide coverage angle (they are way easier to design, and the horns are automatically more constant directivity since the transition from 1>2>3 is way more smooth (physics, yeah!!)), than it is something you should always do i.m.o. 

Hope you find some helpful point in this post.


Edited by Teunos - 14 June 2017 at 7:59am
Best regards,
Teun.
Back to Top
 Post Reply Post Reply Page  <1234>
  Share Topic   

Forum Jump Forum Permissions View Drop Down

Forum Software by Web Wiz Forums® version 12.03
Copyright ©2001-2019 Web Wiz Ltd.

This page was generated in 0.063 seconds.