Hyperbolic bass horns |
Post Reply | Page 123 4> |
Author | |
citizensc
Young Croc Joined: 16 October 2015 Location: Perth,Australia Status: Offline Points: 532 |
Post Options
Thanks(1)
Posted: 01 June 2020 at 11:57am |
Hi Guys
Got a rather simple question that is defeating me a little bit. After doing a bit of reading, I have come to understand that one of the main advantages of a hyperbolic horn is increased sensitivity near the low cutoff frequency. This is because the throat of the horn has a slower flare rate and this increases loading on the driver at lower frequencies. < Please correct me is this is wrong. Page 4 of this discusses it https://www.grc.com/acoustics/an-introduction-to-horn-theory.pdf The bit that is confusing me is hornresp is showing the exact opposite. A T value of 1 is an exponential horn, <1 is hyperbolic and >1 tends towards conical therefore as T decreases, sensitivity near cutoff should increase. Here is a sim of T=0.8 in black and T=0.5 in grey. T=0.5 clearly has less sensitivity near cutoff. Could someone please fill in the gap in my understanding here? Thanks Hugh |
|
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5482 |
Post Options
Thanks(0)
|
If I remember the GRC paper it is discussing "complete" horns, in that the mouth area is large enough to support the horn path length with constant expansion. The first thing I would do is double check that is the case in your simulations.
As I remember the real part of acoustic load of a hyperbolic can rise faster from cutoff to its steady state level vs an exponential. But this is not the only factor when calculating total horn output. It is a multiplying factor I would have thought. Think in electrical analog terms, if you want to know the power dissipated in a load, you have to multiple it by current squared. The analog to current in the driver side of a cabinet model is volume of air moved per second, or average(driver excursion*Sd)^2/time Just a thought, might be off base but it is the first thing that occurs. |
|
snowflake
Old Croc Joined: 29 December 2004 Location: Bristol Status: Offline Points: 3122 |
Post Options
Thanks(0)
|
optimal T is dependent the horn cutoff frequency and the bandwidth of the system. and those affect Vb and St. assume you want to keep Sm, F0 and horn length constant - so try increasing Vb and St as you decrase M. |
|
citizensc
Young Croc Joined: 16 October 2015 Location: Perth,Australia Status: Offline Points: 532 |
Post Options
Thanks(0)
|
After some more reading and experimenting, I think I am starting to understand it better now. As you said T is linked to a number of other elements in the system. You cant just change T and watch what happens, it will throw the reactance annulling off. Vb, St and throat area need to be changed compensate. There are limits to this, where reactance annulling cant be achieved without impractical chamber volumes or compression ratio. If you adjust it all as one system, lower T values = a little bump near cutoff. T is linked to the rest if this system via this formulae (M=T):
Please let me know if anything I am saying is wrong! There are still two things i'm not sure about. 1. The f0 figure that formulae solves for, should that be your exact cutoff frequency or does it just need to be at or below your cutoff frequency? 2. What precisely is cutoff frequency? I understand its where the horn stops loading the driver but how do I identify where exactly this is happening. Where in the graph in my above post is it? The red line at -3db? the point where the horn starts to roll off at ~41hz? Should I be looking at another graph like acoustic impedance - pictured below? Sorry if this is a dumb question, just trying to get my fundamentals right. |
|
smoore
Young Croc Joined: 30 March 2011 Location: N.Devon Status: Offline Points: 937 |
Post Options
Thanks(0)
|
Cut-off is the point where pressure in the horn drops exponentially as it moves towards the front of the horn and pressure moves as 'evanescent waves'. Like you said the horn is no longer 'loading' the speaker.
Personally I wouldn't get too caught up in the precise frequency at which this is occurring as it's dependant on many complex factors you can't fully account for in anything other than a perfect horn and environment. Its just something you should know will happen at roughly the calculated frequency and should base your design on that knowledge before actually building and testing it and adjusting accordingly.
|
|
smoore
Young Croc Joined: 30 March 2011 Location: N.Devon Status: Offline Points: 937 |
Post Options
Thanks(0)
|
This is good reading if you're looking for stuff! - https://www.grc.com/acoustics/an-introduction-to-horn-theory.pdf
******* Just realised you've posted the same link haha **** Edited by smoore - 02 June 2020 at 10:14am |
|
snowflake
Old Croc Joined: 29 December 2004 Location: Bristol Status: Offline Points: 3122 |
Post Options
Thanks(0)
|
EDIT: see Kolbrek paper further down thread
to get an optimally flat response with highest possible efficiency you would set F0=FL. setting F0>FL is just throwing away efficiency. But maybe setting F0<FL and M<<0.7 gives you a bumpy response with a lower F3. This must also interact with the effect from 1/4wave resonance of a non-infinite horn to give the overall bass response. it's hard to come up with a design procedure that makes sense. maybe try reducng M and increase horn length at the same time. you might then need to reduce the mouth size to keep the same FL and system volume.
Edited by snowflake - 12 September 2020 at 6:35pm |
|
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5482 |
Post Options
Thanks(0)
|
The horn stops loading the driver completely when Ra = 0, or the real part of the acoustic impedance is nothing. That means there is no in-phase pressure and air volume movement and so no power transfer occurs.
The red line is the horn reactance, when this is non-zero it means there is air movement in the horn but it is not in phas with driver pressure so a proportion of electrical power delivered to the driver is not converted into acoustical output. If your ignore the peaks due to horn resonances and trace a line through the black data outside the resonance and the black data betweeen the resonances, you get an idea of when it is going to cross through zero. Zooming in the y-axis would help see this. My money would be on somewhere between first and second resonance. You can see under the first resonance frequency that the impedance is pretty much all reactive, so that is a definite cut off there. |
|
citizensc
Young Croc Joined: 16 October 2015 Location: Perth,Australia Status: Offline Points: 532 |
Post Options
Thanks(0)
|
So you are saying the cutoff is actually around 50-60 hz? I have been assuming 40hz when I did the reactance annulling and chamber volumes based on the formulae on this paper. http://leachlegacy.ece.gatech.edu/papers/HornPaper/HornPaper.pdf My target -3db point for the response is 35-36hz, should I be using a cutoff half an octave higher than that when doing my calculations? |
|
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5482 |
Post Options
Thanks(0)
|
It is hard to tell from the graph, but that is what the Ra value is suggesting. You have to zoom into see it properly. Looks like the real part of the first resonance goes down to about 40Hz but you're talking one pixel on a screen at that point.
I suspect there is more to total output than that just value, there will be scaling factors with frequency etc involved. I have never done the detailed analysis for deeper understanding myself, it is on the one day list (along with everything else). |
|
snowflake
Old Croc Joined: 29 December 2004 Location: Bristol Status: Offline Points: 3122 |
Post Options
Thanks(0)
|
model the system in Hornresp with Ang set to infinite horn and it will you show you very definitely where the cutoff frequency is.
|
|
snowflake
Old Croc Joined: 29 December 2004 Location: Bristol Status: Offline Points: 3122 |
Post Options
Thanks(0)
|
has anyone worked out the maximum efficiency design procedure from Leach's paper? He say it 'invloves slightly more complicated expressions'. I can't see how you find Qec or Vat without specifying an additional driver paramter.
|
|
Post Reply | Page 123 4> |
Tweet |
Forum Jump | Forum Permissions You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |