| Consequences of non-optimal throat transition in mid and high frequency horns [message #19175] |
Mon, 19 March 2007 05:55  |
Peter Krojgaard
Messages: 30
Registered: May 2009
|
Baron |
|
|
|
Hi, If the transition from a compression driver to a rectangular horn is less than optimal, where in the frequency band will this be most audible - and how? I ask because I use a pair of large rectangular 200 Hz mid/high horns for a pair of BMS 4592 ND drivers (I use this combo from 266 Hz and up). The throat transition is not a disaster, but it is not super smooth either, and I have some harshness in the treble (6-9kHz, I think) and wonder to what extent this harchness may be due to the non-optimal transition. Thanks! All the best
Peter
|
|
|
|
|
|
| Re: Consequences of non-optimal throat transition in mid and high frequency horns [message #19177 is a reply to message #19176] |
Mon, 19 March 2007 11:53  |
Peter Krojgaard
Messages: 30
Registered: May 2009
|
Baron |
|
|
|
Hi Wayne, Thanks a lot for your reply! To be more specific: In my horns the first 3/4 of an inch (the "adapter")is not flared at all, but resembles a "tube" (I hope it is possible to visualize my description!). After that, the circular entry is slowly (over a distance of app. 2 inches) converted to the rectangular form of the exponential horn. This transition piece is fairly smooth, but not perfect. Given your expertise, do you find it likely that a "transition" as described above would (or could) course harshness in the 6-9 kHz range (in my case, especially too pronounced s-sounds in female voices)? Thanks again Wayne! All the best
Peter
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| LC Peaking [message #19184 is a reply to message #19179] |
Tue, 20 March 2007 23:25 |
 |
Wayne Parham
Messages: 18680
Registered: January 2001
|
Illuminati (33rd Degree) |
|
|
Yes, I think that might be caused by LC peaking in the crossover. When capacitance and inductance are in the circuit, a resonant condition is formed. Without the right amount of damping, it can cause peaking. The electrical properties of the driver (including the transformations of acoustic resistance/reactance into mechanical resistance/reactance and eventually into electrical impedance) come into play as well.
Some people, including even some well respected speaker designers, consider a loudspeaker to be resistive, and estimate it as such when doing calculations. It is particularly inappropriate for horns because they are 1/4 wave devices that have several impedance peaks near cutoff. Since this is where they are likely to be crossed-over, it is important to consider the impedance of the horn in the crossover region. Even though a horn approaches pure resistance, it only does this well into its passband and only if adequately sized. Many times, horns are undersized to meet specific design criteria and the crossover point is almost always low in the pasband, so the horn's complex impedance near the low-frequency crossover point must always be considered for best results.
Each year, I give a seminar called "Crossover Electronics 101" at the Lone Star Audiofest. The main emphasis of this seminar is to familiarize people with reactive circuits, to show exactly how peaking can (inadvertently) creep into a crossover design and to show how to damp the circuit properly to prevent it. I present a series of slides with schematics and response charts that show what each circuit does. Then we connect each circuit to a horn tweeter and play sound through it, listening to each one to hear how they sound. I think it's interesting for people to actually hear what various levels of peaking sounds like, and to compare different circuit types.
|
|
|
|
Posting Rules
Members
Search
Help
Register
Login
Get a T-Shirt!
