Sure – but we can't do it all in a single step - no?
To look deeper (in a quantifiable way) into that topic I'd like to translate horn HP into an electric filter first – then we can process that filter to our liking in SPICE and import the results into acoustic analysis software like ARTA.
Unfortunately there are some hurdles to take :
- first, lack of GD plots from horns of interest (to simplify estimation of filter order and type of those horns)
- second, ARTA refuses to correctly import IR files created by LT SPICE in a first run
After all above is overcome – then I think we also should look into "loading" – but rather in a more intuitively form of excursion over frequency for a given SPL – or in other words horn gain over frequency.
This is something that is traced down in my doppler IM spreadsheet for imported files of HornResp for example
Quite a task – possibly one has already looked at horn honk that way?
Michael
To look deeper (in a quantifiable way) into that topic I'd like to translate horn HP into an electric filter first – then we can process that filter to our liking in SPICE and import the results into acoustic analysis software like ARTA.
Unfortunately there are some hurdles to take :
- first, lack of GD plots from horns of interest (to simplify estimation of filter order and type of those horns)
- second, ARTA refuses to correctly import IR files created by LT SPICE in a first run
After all above is overcome – then I think we also should look into "loading" – but rather in a more intuitively form of excursion over frequency for a given SPL – or in other words horn gain over frequency.
This is something that is traced down in my doppler IM spreadsheet for imported files of HornResp for example
Quite a task – possibly one has already looked at horn honk that way?
Michael
Hello Soongsc,
The horn acts as a transformer so the way it is loaded is directly in relation to the power it can delivers on a reference input signal.
(Think impedance adaptation...)
Best regards from Paris, France
Jean-Michel Le Cléac'h
The horn acts as a transformer so the way it is loaded is directly in relation to the power it can delivers on a reference input signal.
(Think impedance adaptation...)
Best regards from Paris, France
Jean-Michel Le Cléac'h
Hello Michael
The compensation I illustrated uses one more channel of the DCX2496.
The first channel receives the input signal and you set the mentionned delay( plus the High Pass crossover you normally use).
The second channel receives also the input channel and is programmed with a low-pass Linkwitz-Riley filter 48dB/octave with Fc a bit less than 2 times the acoustical cut-of of the horn. Its polarity is inversed.
A 2 (or 3) resistors summator is used to sum the outputs of the 2 channels.
As an example, for the graph I attached in a previous message (for a Le Cléac'h horn, T = 0.8, Fc = 320Hz):
first channel:
delay 550 millimeters
second channel :
low-pass LR48dB/octave Fc = 738Hz
delay = 0 millimeter
polarity: inverted
Both channels fed with the same input signal
Best regards from Paris, France
Jean-Michel Le Cléac'h
The compensation I illustrated uses one more channel of the DCX2496.
The first channel receives the input signal and you set the mentionned delay( plus the High Pass crossover you normally use).
The second channel receives also the input channel and is programmed with a low-pass Linkwitz-Riley filter 48dB/octave with Fc a bit less than 2 times the acoustical cut-of of the horn. Its polarity is inversed.
A 2 (or 3) resistors summator is used to sum the outputs of the 2 channels.
As an example, for the graph I attached in a previous message (for a Le Cléac'h horn, T = 0.8, Fc = 320Hz):
first channel:
delay 550 millimeters
second channel :
low-pass LR48dB/octave Fc = 738Hz
delay = 0 millimeter
polarity: inverted
Both channels fed with the same input signal
Best regards from Paris, France
Jean-Michel Le Cléac'h
Last edited:
Hello Michael,
I know quite well Arta but not the other piece of software you mentionned.
When I import an impulse response obtained throught the use of another software (e.g. Adobe Audition + Angelo Farins's plugins) some useful features of Arta disappear.
In fact I succeeded in recovering all features in Arta after dimensioning the Wav file to be imported to 262144 samples 16bits .wav format. (That's easy to perfom under CoolEdit or Adobe Audition)
Best regards from Paris, France
Jean-Michel Le Cléac'h
I know quite well Arta but not the other piece of software you mentionned.
When I import an impulse response obtained throught the use of another software (e.g. Adobe Audition + Angelo Farins's plugins) some useful features of Arta disappear.
In fact I succeeded in recovering all features in Arta after dimensioning the Wav file to be imported to 262144 samples 16bits .wav format. (That's easy to perfom under CoolEdit or Adobe Audition)
Best regards from Paris, France
Jean-Michel Le Cléac'h
Last edited:
Thanks for your input, Jean-Michel
#####
Hello, back after some time it took me to manage data transfer between ARTA and LT SPICE.
For a moment I'd really would like to stay at the topic of GD of horns and it's impact or relevance to the "guttural" sonics I found so annoying if XO comes too close to cut off.
Though "first hurdle" is basically overcome, I'd would need your help to look at a sample HP filter's performance validity.
Best visualization for the effect seems to be a CSD plot in its sonogram variant:
Below we see a text book 3rd order Chebyshev high pass filter at 1kHz :
I calculated with SPICE – converted the export file – and imported into ARTA.
It mimics a 88kHz sampling rate and the pulse is set after 0.5ms in a window of 100ms
IR and FR looks as it should IMO
GD has a strange notch below 1kHz – any thoughts on that ?
CSDs looks "not completely" unreasonable - though - I would have expected more uniformity below 1kHz – any thoughts on that ?
Above is just a sample to warm up in the topic – if reasonable – I'd like to apply it to more real world estimations of horns / XO combinations.
I remember John Kreskovsky having once shown a CSD of a pure HP filter – but digging it up in Lynn's thread seems the pix is gone...
Anybody seen CSD of "text book" filters elsewhere for comparison / validation ?
Michael
#####
Hello, back after some time it took me to manage data transfer between ARTA and LT SPICE.
For a moment I'd really would like to stay at the topic of GD of horns and it's impact or relevance to the "guttural" sonics I found so annoying if XO comes too close to cut off.
Though "first hurdle" is basically overcome, I'd would need your help to look at a sample HP filter's performance validity.
Best visualization for the effect seems to be a CSD plot in its sonogram variant:
Below we see a text book 3rd order Chebyshev high pass filter at 1kHz :
I calculated with SPICE – converted the export file – and imported into ARTA.
It mimics a 88kHz sampling rate and the pulse is set after 0.5ms in a window of 100ms
IR and FR looks as it should IMO
GD has a strange notch below 1kHz – any thoughts on that ?
CSDs looks "not completely" unreasonable - though - I would have expected more uniformity below 1kHz – any thoughts on that ?
Above is just a sample to warm up in the topic – if reasonable – I'd like to apply it to more real world estimations of horns / XO combinations.
I remember John Kreskovsky having once shown a CSD of a pure HP filter – but digging it up in Lynn's thread seems the pix is gone...
Anybody seen CSD of "text book" filters elsewhere for comparison / validation ?
Michael
Last edited:
Can someone summarize this discussion in a "dummy's version" for me?
Im just curious on what I can do to eliminate the honk in my DIY build. The other thread on this topic "HOM" stuff says to make sure the baffles have large round over (Im using 1/2" right now but Im going to do 1 1/4" next build) and have foam (ala Geddes).
Im just curious on what I can do to eliminate the honk in my DIY build. The other thread on this topic "HOM" stuff says to make sure the baffles have large round over (Im using 1/2" right now but Im going to do 1 1/4" next build) and have foam (ala Geddes).
Hy doug20
Myself being a noob / dummy / twit on the topic, I try to first get together all ingredient (horn gain, GD, IM, mouth reflection etc. ) that make up for the "melange" horn honk seems to be IMO. Havent found that elsewhere in a "compendium" - too many schools out there...
So it may take a while until you get some "cook book" suggestions here.
LOL - if at all !
Michael
Myself being a noob / dummy / twit on the topic, I try to first get together all ingredient (horn gain, GD, IM, mouth reflection etc. ) that make up for the "melange" horn honk seems to be IMO. Havent found that elsewhere in a "compendium" - too many schools out there...
So it may take a while until you get some "cook book" suggestions here.
LOL - if at all !
Michael
Last edited:
I remember this thread being pretty straight forward on the HOM issue which seems to include the "Honk"
http://www.diyaudio.com/forums/multi-way/151376-homster-how-i-learned-how-fix-horn.html
Is this topic on the same path? Or are we talking strictly about XO stuff ?
http://www.diyaudio.com/forums/multi-way/151376-homster-how-i-learned-how-fix-horn.html
Is this topic on the same path? Or are we talking strictly about XO stuff ?
About HOM and its definition and relevance there already was a exhausting discussion I participated to some extend.
I do not want to warm that up - this book is closed for me.
On the other hand - mouth reflection, what Jean-Michel refers to as "tuned pipe effect" some postings back, certainly is "in the pot" .
You may have a look to the simus I have shown in Earl's thread, where the impact of different mouth round over can be estimated pretty well...
For now it may go into acoustic respectively electric filter / XO stuff - simply because the cut off might be seen equivalent to a high pass filter and hence "with the right preparation" (impulse shaping) a lot might be gained already in terms of improving on horn honk (as far as concerning the guttural tone coloration).
Michael
I do not want to warm that up - this book is closed for me.
On the other hand - mouth reflection, what Jean-Michel refers to as "tuned pipe effect" some postings back, certainly is "in the pot" .
You may have a look to the simus I have shown in Earl's thread, where the impact of different mouth round over can be estimated pretty well...
For now it may go into acoustic respectively electric filter / XO stuff - simply because the cut off might be seen equivalent to a high pass filter and hence "with the right preparation" (impulse shaping) a lot might be gained already in terms of improving on horn honk (as far as concerning the guttural tone coloration).
Michael
Last edited:
If you look at the acoutic impedance of a tuned pipe, you might find some interesting issues. From close to 100 simulations, it seems every part of the horn/guide effects the acoustic impedance. I suspect that Earl's foam plays an important roll of smoothing acoustic impedance. The Newell and Holland book on LOUDSPEAKERS talks about the effects of this. The real difficult issue is selecting a balance between directivity control and loading. Currently it seems the throat section sets the main trend of the acoustic impedance, and the outer horn/guide sort of fine tunes it.
Regarding the flat response you posted somewhere, it's possible the acoustic impedance becomes higher in the upper frequency range. My sims show this trend.
Regarding the flat response you posted somewhere, it's possible the acoustic impedance becomes higher in the upper frequency range. My sims show this trend.
Horn honks if these criteria are not met
Posted this on another thread but it SO applies here just had to post it again. Hope no one minds...
Somewhere in this thread someone said a horn is a directivity control device. Actually a horn is an acoustic transformer which transforms the low amplitude high pressure at the diaphragm to the higher amplitude lower pressure suitable for radiation into free air. An impedance match throat to mouth. Modern horns have change the shape of the matching element to control directivity and in the process change the impedance away from that of the straight elliptical horn. Even though the modern horn with constant directivity does not offer constant impedance at the throat the impedance seen by the driver is within the operational limits of the diaphragm design (usually) so they work fine. Reflection back down the throat of the horn (this is the honk part) occur when the match at the mouth in not equal to the impedance of the free air at the mouth. By far the biggest offender of this is the dual rate horn with always bad sound and always lots of reflections. To minimize reflections horn mouth really needs to be at least 1/2 wavelength at the lowest frequency used and really should be the wavelength of the lowest frequency used.(this is how to greatly reduce the honk) So you may see the dual rate horn has a mouth the size of where the rate changes and is therefore limited to narrow band operation. And please note that is the mouth minimum dimension if the mouth is not round. (example a 4" by 10" horn the number of concern is 4") This means (using the 1/2 wave rule) a horn mouth of 4 inches is only good down to 1700Hz absolute minimum including the crossover region. This means there is no such thing as a good small horn.
Posted this on another thread but it SO applies here just had to post it again. Hope no one minds...
Somewhere in this thread someone said a horn is a directivity control device. Actually a horn is an acoustic transformer which transforms the low amplitude high pressure at the diaphragm to the higher amplitude lower pressure suitable for radiation into free air. An impedance match throat to mouth. Modern horns have change the shape of the matching element to control directivity and in the process change the impedance away from that of the straight elliptical horn. Even though the modern horn with constant directivity does not offer constant impedance at the throat the impedance seen by the driver is within the operational limits of the diaphragm design (usually) so they work fine. Reflection back down the throat of the horn (this is the honk part) occur when the match at the mouth in not equal to the impedance of the free air at the mouth. By far the biggest offender of this is the dual rate horn with always bad sound and always lots of reflections. To minimize reflections horn mouth really needs to be at least 1/2 wavelength at the lowest frequency used and really should be the wavelength of the lowest frequency used.(this is how to greatly reduce the honk) So you may see the dual rate horn has a mouth the size of where the rate changes and is therefore limited to narrow band operation. And please note that is the mouth minimum dimension if the mouth is not round. (example a 4" by 10" horn the number of concern is 4") This means (using the 1/2 wave rule) a horn mouth of 4 inches is only good down to 1700Hz absolute minimum including the crossover region. This means there is no such thing as a good small horn.
Sure, in other words you can break down a conical to be sort of transmission line.
You just have to narrow the included angle by half - and again by half – and if needed, again by half.
It then becomes immediately intuitively that the sonic pattern of a transmission line - its strong reso at certain frequencies – MUST be part of conicals to some degree.
A round over at the mouth may help some, but a TM with a round over still is a TM –no?
Seen from this perspective, the foam trick is nothing else than the art of dampening a transmission line (with a loooot of woodoo dancing around).
To be fair – the TM issue of horns applies to any abrupt change of boundary contour that are not optimized in diffraction alignment – with the effects clearly seen in sound field defects.
The best contour I find in this respect is the LeCleach where Jean-Michel has investigated the propagation of a wave front and accounted his expansions laws for.
Yes that's the "trick" I had in mind when answering to
###########
Can you share some work on horn gain you possibly have done and summarize here in the light of horn honk (with a lot of text of explanation
!)?Michael
Last edited:
me certainly not...
I think – we should look at a topic from as many points of view as possible.
"Directivity control device", "diffraction alignment device", "acoustic transformer" etc - all emphasis on strong points regarding this horn animal.
The discussion of horn dimensions with respect to band width is just another point of view at that horn animal – no ?
summarizing above, I'd rather put it the way that "there is no such thing as a good *low frequency* small horn "
The interesting part IMO in what you brought up is what also is outlined in detail form Jack Bouska (Pano set the links there).
Towards the low cut of a horn, control over directivity and mouth reflection sadly get lost more or less - well there are limits to each and everything...
Michael
I don't think I'm quite there yet , but still working along the concepts that I've mentioned. The difficulties are at the two extreme ends. Even if they do get to acceptable levels in the analysis and sims, they would have to be listened to before I can say how successful they are.Can you share some work on horn gain you possibly have done and summarize here in the light of horn honk (with a lot of text of explanation
Michael
Some more examples of acoustic impedance variation: Highlighted some. Have no explanation for what's going on.
An externally hosted image should be here but it was not working when we last tested it.
Ahh - uumpf - grrr - huuu - GOT IT !
These are simus !
Ok.
- and the light blue for sure is a OSWG for reference - where did I have my thoughts...
- the dark blue and is one of your earlier works with the wiggle in the 15kHz range
How many points for the candidate ?
##########
Seriously - can you explain a little bit ?
Michael
These are simus !
Ok.
- and the light blue for sure is a OSWG for reference - where did I have my thoughts...
- the dark blue and is one of your earlier works with the wiggle in the 15kHz range
How many points for the candidate ?
##########
Seriously - can you explain a little bit ?
Michael
These are beautiful graphs!
When the Z goes down and suddenly up that is phase flip of the signal. The gigantic peak is probably a standing wave or other strong resonance such as Fs. The very narrow peak in blue is likely a diaphragm resonance or other resonance within the driver. With such a high Q that is about the only choice.
So may I suggest you take any horn and do this kind of curve and then stand out in front of your horn about 1 meter with a 30cm square piece of cardboard and position the cardboard perpendicular to the axis of the horn and look at the Z curve. Now tilt the cardboard at some angle and see the changes in the curve again. The region the horn does not change Z much with this test is the region where the horn is making a good match. This test is for a tweeter horn used above 1kHz. For lower frequency horn use bigger cardboard and stand further away. Anyone who does this will learn more than they want to know about horns.
It may interest some of you I have found this Dayton 1 1/8 inch screw on horn from Parts Express with a 11 inch plus square mouth. This is the best horn I have ever tested with a decade of bandwidth where the horn stays in phase. I added a little felt phase ring between the driver and the horn (1 1/8"OD and 5/8 inch ID 1/8 inch thick) and this amazing horn worked (I mean stayed in phase and did not honk!!) from 1300Hz to 13,000Hz...WOW! It is so good I am trying out a PA speaker for the first time.
Wish I could find a better compression driver with wider bandwidth....
Always liked the looks of the Altec 802-8G but those things are now many hundreds of dollars. Known as the "tangerine" driver. Oh well.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.