Hello,
I build my loudspeakers but I have some problem.
When I test my loudspeakers in near field measurement (tweeter and woofer separately in the real cabinet) and use with phase and combine, the frequency range is beautiful and very flat.
But when I measure at Far field method, the graph look with problem at 1500Hz to 2800Hz. maybe is my room. maybe not.. I'm dont know.
I use with very high resolution - 24bit/96Khz.
I can to rely on near field measurement or is not say nothing?
Thank you
I build my loudspeakers but I have some problem.
When I test my loudspeakers in near field measurement (tweeter and woofer separately in the real cabinet) and use with phase and combine, the frequency range is beautiful and very flat.
But when I measure at Far field method, the graph look with problem at 1500Hz to 2800Hz. maybe is my room. maybe not.. I'm dont know.
I use with very high resolution - 24bit/96Khz.
I can to rely on near field measurement or is not say nothing?
Thank you
... here you will find some information regarding your question
http://www.fesb.hr/~mateljan/arta/AppNotes/AP4_FreeField-Rev03eng.pdf
Heinrich
http://www.fesb.hr/~mateljan/arta/AppNotes/AP4_FreeField-Rev03eng.pdf
Heinrich
Audist said:
I think that what you see in the far field measurements is the baffle step. The baffle step is supressed in close field measurements due to that the distance to the baffle edges is larger than the distance to the driver.
Below is an example on how the baffle step varies when the microphone is moved between 3 metres (green) and 5 cm (red).
An externally hosted image should be here but it was not working when we last tested it.
Audist said:
Hi Audist,
the NF measurement is made with the mic very close to the woofer dustcap, and it's valid till a frequency=10950/D where D is the driver diameter in centimeter. Said this, measuring a tweeter in NF is useless.
The NF technique is used for the low frequency spectrum.
The far field measurement you performed is gated or not?
Hi Audist,
let's assume your woofer has a diameter D=15 cm so Fmax = 730 Hz that means that we can use the near-field response till this frequency. However this is valid for a driver mounted on infinite baffle, while lowers for a driver mounted on a panel, cause of the baffle effects.
You say your woofer problem are in the 10000 Hz region: what kind of woofer are you using? Sounds a little too high frequency for a woofer. Isn't that you are measuring with both drivers (woofer+tweeter) working?
For a tweeter let's say it has a diameter of 2 cm: Fmax=5475 Hz, but you are already in the FAR FIELD range, so it's not necessary to make a NF for a tweeter.
Far Field: you didn't say if it is a gated or not measurement, the one you performed.
Looks you did room response, that is an ungated Far Field, with both drivers playing.
let's assume your woofer has a diameter D=15 cm so Fmax = 730 Hz that means that we can use the near-field response till this frequency. However this is valid for a driver mounted on infinite baffle, while lowers for a driver mounted on a panel, cause of the baffle effects.
You say your woofer problem are in the 10000 Hz region: what kind of woofer are you using? Sounds a little too high frequency for a woofer. Isn't that you are measuring with both drivers (woofer+tweeter) working?
For a tweeter let's say it has a diameter of 2 cm: Fmax=5475 Hz, but you are already in the FAR FIELD range, so it's not necessary to make a NF for a tweeter.
Far Field: you didn't say if it is a gated or not measurement, the one you performed.
Looks you did room response, that is an ungated Far Field, with both drivers playing.
Re: Re: Near field vs Far field
You have to weigh in the baffle size as well into that equation too. I am convinced that the fluctuations in the far field comes from the baffle step. Also, 10950 sounds terribly accurate, of course there is a transition zone where the measurements gets gradually better/worse.
claudio said:
You have to weigh in the baffle size as well into that equation too. I am convinced that the fluctuations in the far field comes from the baffle step. Also, 10950 sounds terribly accurate, of course there is a transition zone where the measurements gets gradually better/worse.
Hi friend's.
My problem is not with the woofer. the problem is with the tweeter in FF.
My tweeter is morel supreme 110 and I did 2 measurements (FF and NF).
The problem as you can see is on FF measurements where I have problem between 2300Hz to 3400Hz. I want cross it around 2000 to 2500 and is very difficult...
The baffle dimensions is w=20cm, h=40cm.
My problem is not with the woofer. the problem is with the tweeter in FF.
My tweeter is morel supreme 110 and I did 2 measurements (FF and NF).
The problem as you can see is on FF measurements where I have problem between 2300Hz to 3400Hz. I want cross it around 2000 to 2500 and is very difficult...
The baffle dimensions is w=20cm, h=40cm.
Attachments
Ok, so assuming that you measured on 5 cm and 50 cm, the baffle steps would look like this. The difference is terribly close to the difference between your curves.
The bottom line here is that the dip at 2800 Hz originates in the placement of the driver on the baffle. Placing it at the same distance from 3 edges focuses the frequency response effects on a narrow frequency range (around 2800 Hz obviously). The solution is, sadly, to make a new baffle with another driver placement. Download The Edge in my signature first and make some simulations of placements. It is rather easy, if I may say so.
The bottom line here is that the dip at 2800 Hz originates in the placement of the driver on the baffle. Placing it at the same distance from 3 edges focuses the frequency response effects on a narrow frequency range (around 2800 Hz obviously). The solution is, sadly, to make a new baffle with another driver placement. Download The Edge in my signature first and make some simulations of placements. It is rather easy, if I may say so.
An externally hosted image should be here but it was not working when we last tested it.
Problematic situation.
Most of loudspeakers place the tweeter there.
I try to change the baffle and see that the most baffle be similar +/- if you place the tweeter on center line.
Only in side og the baffle is work good.
You have some idea?
N.B: "speakers source density" select 1 or more for dome tweeter? what about ribbon?
Thank you
Most of loudspeakers place the tweeter there.
I try to change the baffle and see that the most baffle be similar +/- if you place the tweeter on center line.
Only in side og the baffle is work good.
You have some idea?
N.B: "speakers source density" select 1 or more for dome tweeter? what about ribbon?
Thank you
Audist said:
Yes it is true that most loudspeaker manufacturers place use symmetrical placement. I think that the main reason for that is to avoid two products instead of one (left and right version), but there are also arguments regarding directivity.
The simulation gets more accurate the higher source density you use. There is a tradeoff there between accuracy and computaional speed. Ribbon tweeters can be simulated by using a rectangular shape of the driver.
If you increase the source density, you can get a feel for how large the errors are. Mostly, more than 10 is not needed, and 6 mostly simulates the low frequency behaviour good enough.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.