Phase ring??
Sorry, what is this "phase ring"??
A spacer at the end of the screw on section where it meets the horn, or something you place somewhere inside the driver's throat area??
Maybe a jpeg... curious.
And maybe you could post a "before and after" FR or other measurement??
Thanks.
_-_-bear
<snip>
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.
Sorry, what is this "phase ring"??
A spacer at the end of the screw on section where it meets the horn, or something you place somewhere inside the driver's throat area??
Maybe a jpeg... curious.
And maybe you could post a "before and after" FR or other measurement??
Thanks.
_-_-bear
Possibly true, but it isn't because of the waveguide.
The driver itself is primarily to blame because of it's high "q" resonance, NOT the waveguide.
Yeah but thee driver measured as well as any other low cost driver. Its after all only $60.
I have the BMS4550 too in an identical speaker so I will get the wave file posted tomorrow.
We can figure out if its the CD or not.
......
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....
I missed this post, could you give us the link?
also a picture of the "phase plug" would be awesome!
I think I mentioned this in another thread, but you just have to get rid of those horns!!
Take a look at the fins in the front first.
Then look at the shape of the throat area.
Then of course the material the horn is made from - damped by you or not.
Imho, you can easily build a horn that will outperform that one just doing it by hand and eye... imho, of course!
_-_-bear
Actually I think it measured better than most drivers irrespective of cost. The one glaring exception however is it's resonance. It's entirely possible that the B&C DE12 (mylar) is better in this respect (..and about the same cost).
It's also possible that the Celestion could be modified to achieve better performance at resonance (..though likely at the cost of some eff. and power handling.)
Hello,
I agree that this CSD is not very good.
You surely noticed that this the only horn of the study for which an Altec 288 was used in place of the B&C DE250.
I have several mesurements of the AH425 with several ALTEC288 and there is a noticeable variability. This is due to the 288.
Then, you can have a better view of what is due to the horn and is dure to the driver looking at the off axis evolution of the frequency response of the very same driver + horn you gave the CSD (same website).
http://sites.google.com/site/driver...ncy-response/Polarresponse.PNG?attredirects=0
What you see here is that from angle to angle, the details on the response curve are the same even at frequency and angle for which there is a large attenuation (e.g. -60dB) this is a clear indication that those details (peaks, holes, ...) are due to the driver not to the horn.
Additionally all 288 for which I have measurements on the AH425 are also leading to specific resonances (more or less important from driver to driver) at 2.5k-3k ; 4.5k ; 7k ; 8.5k ; 12kHz.
Best regards from Paris, France
Jean-Michel Le Cléac'h
I agree that this CSD is not very good.
You surely noticed that this the only horn of the study for which an Altec 288 was used in place of the B&C DE250.
I have several mesurements of the AH425 with several ALTEC288 and there is a noticeable variability. This is due to the 288.
Then, you can have a better view of what is due to the horn and is dure to the driver looking at the off axis evolution of the frequency response of the very same driver + horn you gave the CSD (same website).
http://sites.google.com/site/driver...ncy-response/Polarresponse.PNG?attredirects=0
What you see here is that from angle to angle, the details on the response curve are the same even at frequency and angle for which there is a large attenuation (e.g. -60dB) this is a clear indication that those details (peaks, holes, ...) are due to the driver not to the horn.
Additionally all 288 for which I have measurements on the AH425 are also leading to specific resonances (more or less important from driver to driver) at 2.5k-3k ; 4.5k ; 7k ; 8.5k ; 12kHz.
Best regards from Paris, France
Jean-Michel Le Cléac'h
Hello Michael,
ARTA (Impulse Response module) is a very powerful piece of software, but exception made of the frequency response and the distortion curves for H2, H3 and H4 which are easy to use, the other features as CSD and GD (by example) need some experience to use. Also for those advanced features the interface is not very user friendly (a pity ...)
That's why here in France I'll plan to give in the near future some lessons to an audiophile association (Melaudia) in order to show how to use ARTA.
Example taken on your file HP_LR48_1kHz.pir
To obtain a good Group Delay curve:
Open the pir file. In the impulse responce window, put your cursor by a left click (yellow line on 4.40ms, as an example) then place your cursor around 8.80ms and with a right click set a marker (red line).
Then on the right margin of the IR window, below "zoom" action [Max].
Then put your cursor on the top of the main peak (positive), right click.
On top of the IR window near "delay for phase estimation" action [Get]. You should obtain something like 2.404 (ms). Then on top right near "Marker" action [Del].
Then open the "smoothed frequency response" window.
Action [ Low Fr > ] in order to enlarge the useful part of the frequency response curve (as an example try to have 500Hz as fmin). Eventually do a screen copy of the response curve...
Then at the bottom of that window action [ Gd ]. On the right of the GD window action [Set] this open the "Graph Set Up" window. On right you put Gdtop 1.7ms and on Gd range 1.7ms. Then action [OK].Eventually do a screen copy of the group delay curve...
Close the "group delay curve window" (which still have the name "smoothed frequency response".
To obtain a good CSD.
Remember you have preciously set the cursor (yellow line) at 4.40ms and you have deleted the marker (red line). In the case of mesurement done with a mircophone in a room you may have to put the marker (red line) just befor the pulse corresponding to the first reflection.
action CSD button (on top of the Impulse response window). The Cumulative Spectrum Setup window opens.
- check Log frequency axis
- check remove antialising range
- dB range : 25 (but this value may be change latter)
- maximum frequency (Hz) : 20000
- uncheck Smoothing (better to use the rough results than a smoothing even if the result is better looking it doesn't have more information inside)
- uncheck Use FR compensation
Time -ferquency Foirier analysis:
- Analyze as CSD using apodizing window
- FFT length (in sample) : 128
- FFT block shift (in samples) : 5
- Max. number of FFT blocks : 30
- Apodizing window rise time : 1ms
The last 4 parameters can eventually be retouched but as such they are quite well adpated to the example.
Then action [OK]
On the Cumulative Spectral Decay window then:
- Mode : Sonogram (this is my preferred presentation type)
- palette : Jet
- check : "Colored Waterfall"
- check : "Contoured"
- Uncheck : "Grid"
- Range (dB) : 25 (this can now be retouched at you convenience)
Eventually if you cannot see the top of the CSD, then close the CSD window, go back to the impulse response window and move a bit the cursor on left (4.286ms as an example). Then action [CSD] and [OK].
Best regards from Paris, France
Jean-Michel Le Cléac'h
ARTA (Impulse Response module) is a very powerful piece of software, but exception made of the frequency response and the distortion curves for H2, H3 and H4 which are easy to use, the other features as CSD and GD (by example) need some experience to use. Also for those advanced features the interface is not very user friendly (a pity ...)
That's why here in France I'll plan to give in the near future some lessons to an audiophile association (Melaudia) in order to show how to use ARTA.
Example taken on your file HP_LR48_1kHz.pir
To obtain a good Group Delay curve:
Open the pir file. In the impulse responce window, put your cursor by a left click (yellow line on 4.40ms, as an example) then place your cursor around 8.80ms and with a right click set a marker (red line).
Then on the right margin of the IR window, below "zoom" action [Max].
Then put your cursor on the top of the main peak (positive), right click.
On top of the IR window near "delay for phase estimation" action [Get]. You should obtain something like 2.404 (ms). Then on top right near "Marker" action [Del].
Then open the "smoothed frequency response" window.
Action [ Low Fr > ] in order to enlarge the useful part of the frequency response curve (as an example try to have 500Hz as fmin). Eventually do a screen copy of the response curve...
Then at the bottom of that window action [ Gd ]. On the right of the GD window action [Set] this open the "Graph Set Up" window. On right you put Gdtop 1.7ms and on Gd range 1.7ms. Then action [OK].Eventually do a screen copy of the group delay curve...
Close the "group delay curve window" (which still have the name "smoothed frequency response".
To obtain a good CSD.
Remember you have preciously set the cursor (yellow line) at 4.40ms and you have deleted the marker (red line). In the case of mesurement done with a mircophone in a room you may have to put the marker (red line) just befor the pulse corresponding to the first reflection.
action CSD button (on top of the Impulse response window). The Cumulative Spectrum Setup window opens.
- check Log frequency axis
- check remove antialising range
- dB range : 25 (but this value may be change latter)
- maximum frequency (Hz) : 20000
- uncheck Smoothing (better to use the rough results than a smoothing even if the result is better looking it doesn't have more information inside)
- uncheck Use FR compensation
Time -ferquency Foirier analysis:
- Analyze as CSD using apodizing window
- FFT length (in sample) : 128
- FFT block shift (in samples) : 5
- Max. number of FFT blocks : 30
- Apodizing window rise time : 1ms
The last 4 parameters can eventually be retouched but as such they are quite well adpated to the example.
Then action [OK]
On the Cumulative Spectral Decay window then:
- Mode : Sonogram (this is my preferred presentation type)
- palette : Jet
- check : "Colored Waterfall"
- check : "Contoured"
- Uncheck : "Grid"
- Range (dB) : 25 (this can now be retouched at you convenience)
Eventually if you cannot see the top of the CSD, then close the CSD window, go back to the impulse response window and move a bit the cursor on left (4.286ms as an example). Then action [CSD] and [OK].
Best regards from Paris, France
Jean-Michel Le Cléac'h
Last edited:
Hello,
Here is for comparison a CSD obtained by Jzajaga from one my measured impulse response ( 2 measurements one with the TAD TD2001 current driven and the other one voltage driven).
http://www.diyaudio.com/forums/attachment.php?s=&postid=1531292&stamp=1212610725
Best regards from Paris, France
Jean-Michel Le Cléac'h
Here is for comparison a CSD obtained by Jzajaga from one my measured impulse response ( 2 measurements one with the TAD TD2001 current driven and the other one voltage driven).
http://www.diyaudio.com/forums/attachment.php?s=&postid=1531292&stamp=1212610725
Best regards from Paris, France
Jean-Michel Le Cléac'h
Hello Jean-Michel,
Would it be possible for you to post an impulse response file of one of your Le Cléac'h horns? I will calculate the wavelet transform and post the result here.
Your horn is presumably having the least horn honk in the world of horns because of the genious mouth termination profile. That would give us an ultimate reference.
- Elias
Would it be possible for you to post an impulse response file of one of your Le Cléac'h horns? I will calculate the wavelet transform and post the result here.
Your horn is presumably having the least horn honk in the world of horns because of the genious mouth termination profile. That would give us an ultimate reference.
- Elias
This looks like an interesting sample to build.
But it's pretty big. 60cm diameter and 30 cm long.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
But it's pretty big. 60cm diameter and 30 cm long.
Last edited:
How about the bifurcation at 4.1?The pulses at 6.3 and 7.5 look interesting.
I don't suspect that's my beer bottle.
[You didn't expect I'd clean up, surely....
]
Last edited:
Hello Elias,
You'll find in attached file (in text format) as .wav (48kHz, 16 bits) the impulse reponse I measured on one of my horns (Fc = 320Hz, T = 0.707). The driver is the TAD TD2001.
Best regards from Paris, France
Jean-Michel Le Cléac'h
You'll find in attached file (in text format) as .wav (48kHz, 16 bits) the impulse reponse I measured on one of my horns (Fc = 320Hz, T = 0.707). The driver is the TAD TD2001.
Best regards from Paris, France
Jean-Michel Le Cléac'h
Hello Robh,
The GD problem of the HP should be separated of the problem of the GD of the horn.
Why?: because the GD of the couple LP + HP has to be considered not only the GD of the HP. Choosing a quasioptimal crossover (as the Le Cléac'h crossover) allows the couple LP + HP to have a very constant GD inside the "common interval of frequency" between the bass loudspeaker and the medium loudspeaker (2 ways example taken as an example).
Best regards from Paris, France
Jean-Michel Le Cléac'h
The GD problem of the HP should be separated of the problem of the GD of the horn.
Why?: because the GD of the couple LP + HP has to be considered not only the GD of the HP. Choosing a quasioptimal crossover (as the Le Cléac'h crossover) allows the couple LP + HP to have a very constant GD inside the "common interval of frequency" between the bass loudspeaker and the medium loudspeaker (2 ways example taken as an example).
Best regards from Paris, France
Jean-Michel Le Cléac'h
Almost all of the delay is from the crossovers. So it's the crossovers or the horns?? That delay would be there for cone drivers as well so again I don't get the significance as it relates to horn honk.
To avoid the "worst" GD from the horn all you have to do is go up an octave with your crossovers but the crossver is going to add more Group delay than the horn has. The GD is certainly not an issue with the horn in that case.
Rob
Thank You, I have had ARTA for over a year and I love the analyzer for in room tweaking but you just posted the best information about it I have seen. The manual is horrible, the FAQ, User forums really don't exist. Just want to say thanks!!
btw, will your ARTA lessons be in French?
Member
Joined 2003
This looks like an interesting sample to build.
But it's pretty big. 60cm diameter and 30 cm long.
Yes it does look interesting. Go for it!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.