We all know matching the exit angle of the compression driver to the throat of the horn is the best way to go. However if there will be a compromise which one would be a better choice. Using an exit angle that is lower than the throats angle or one that is higher.
For example, take Yuichi Arai's 290 horn. It is designed to match the 10degree exit angle of the Tad 4001 (or JBL 2450, or even BMS 4592nd which is also has 10 degrees exit)
What if we wanted to use a Beyma cp850nd with its 7degrees exit angle. Would it be better than using a Radian 950 with its 20degrees exit? As the shape opens up rather than closing down it makes me think having a lower exit angle may work better however I have no experience on this.
What kind of a mismatch is more detrimental and at what frequency do these detrimental effects arise. If it is higher frequency ripples correlating to 2" throat size, then they would they be around 6700hz. What if we cut below that, surely it will still have an effect but would it be audible enough to consider a different cd?
For example, take Yuichi Arai's 290 horn. It is designed to match the 10degree exit angle of the Tad 4001 (or JBL 2450, or even BMS 4592nd which is also has 10 degrees exit)
What if we wanted to use a Beyma cp850nd with its 7degrees exit angle. Would it be better than using a Radian 950 with its 20degrees exit? As the shape opens up rather than closing down it makes me think having a lower exit angle may work better however I have no experience on this.
What kind of a mismatch is more detrimental and at what frequency do these detrimental effects arise. If it is higher frequency ripples correlating to 2" throat size, then they would they be around 6700hz. What if we cut below that, surely it will still have an effect but would it be audible enough to consider a different cd?
I would concur that going from a smaller angle (CD) to a wider one (Horn) is likely to be less damaging than the other way around.
And yes, the aberrations will arise at high frequencies, related to the width of the cross-section where the discontinuity is. If you cross-over to another driver steeply enough below the frequency where the issue starts to manifest itself, its audibility will be greatly diminished.
Marco
I would say the distance (i.e. length) along the axis where the discontinuity takes place does also play a role in what frequencies are affected - not only the diameter.
Edit: And yes, my gut feeling is also that going from smaller to larger angle would be less detrimental than vice-versa.
Regards
Charles
Last edited:
Where did you get the Beyma exit angle ? I emailed Beyma about my CP380s and just got a reply pointing me to the spec sheet (that doesn't show an exit angle 🙁 )
A reverse taper is REALLY bad news.
IE, if you have a compression driver with an exit of 10° and a horn with an entrance of 0°, that's really bad.
If you have a compression driver with an exit of 6° and a horn with an entrance of 0°, that's not AS bad.
If you have a compression driver with an exit of 10° and a horn with an entrance angle of 22.5°, that's not great but it's not the end of the world.
You can simulate all of this with axidriver or ABEC.
The challenge of a reverse taper is that you get a dip in the high frequency response, generally somewhere around 15khz give or take a few khz. The thing that's so maddening about this dip is that:
1) there's no easy way to fix it
2) the frequency and the depth of the dip will depend on the wavefront produced by the compression driver and the geometry of it's phase plug
I've literally spent HOURS in front of ABEC working on phase plugs with a reverse taper and it's REALLY difficult to make it work. The Celestion CDX1-1425 has a reverse taper, and it works somehow. I'm guessing that Celestion likely tried a few dozen different geometries until they found one that works. Life is too short, and the best solution for the DIYer is simply to be sure to avoid reverse tapers at all costs. Unless, y'know, you can't hear above 10khz, in which case it's not an issue.
IE, if you have a compression driver with an exit of 10° and a horn with an entrance of 0°, that's really bad.
If you have a compression driver with an exit of 6° and a horn with an entrance of 0°, that's not AS bad.
If you have a compression driver with an exit of 10° and a horn with an entrance angle of 22.5°, that's not great but it's not the end of the world.
You can simulate all of this with axidriver or ABEC.
The challenge of a reverse taper is that you get a dip in the high frequency response, generally somewhere around 15khz give or take a few khz. The thing that's so maddening about this dip is that:
1) there's no easy way to fix it
2) the frequency and the depth of the dip will depend on the wavefront produced by the compression driver and the geometry of it's phase plug
I've literally spent HOURS in front of ABEC working on phase plugs with a reverse taper and it's REALLY difficult to make it work. The Celestion CDX1-1425 has a reverse taper, and it works somehow. I'm guessing that Celestion likely tried a few dozen different geometries until they found one that works. Life is too short, and the best solution for the DIYer is simply to be sure to avoid reverse tapers at all costs. Unless, y'know, you can't hear above 10khz, in which case it's not an issue.
The amount of energy reflected from the discontinuity depends on the area at which this occurs, but more importantly is the distance from the diaphragm to the discontinuity as that determines the frequencies of the peaks. Whether the angle gets larger or smaller is of little consequence - its the differential that counts.
I actually ended up here through google, this thread is appropriately titled and not ancient yet....
I didn't notice an explanation of how things go once one goes the route of using a throatless CD. The exit angle is what? 0? So now you'd want a horn with a 0 angle, as well, right?
I didn't notice an explanation of how things go once one goes the route of using a throatless CD. The exit angle is what? 0? So now you'd want a horn with a 0 angle, as well, right?
Last edited:
I've always been disappointed by how little my measurements change from careful matching of CD exit to throat, vs a pretty dang coarse matchup.
Good to know it's not just me Pano, thx
Hey, i'll go full out heretic and say i can't hear it either...
Hey, i'll go full out heretic and say i can't hear it either...
Brave man - but no, I have not been able to hear it either. I'll qualify that with "Yet"
On the other hand I had to quickly adapt a 2" BMS coaxial compression drivers to a 1.4" horn. An Eminence 1.4 to 2 inch adapter backward did the trick. It works, but kills the top octave. Similar to what Patrick talked about earlier with angle mismatches going backward.
On the other hand I had to quickly adapt a 2" BMS coaxial compression drivers to a 1.4" horn. An Eminence 1.4 to 2 inch adapter backward did the trick. It works, but kills the top octave. Similar to what Patrick talked about earlier with angle mismatches going backward.
I was certainly able to measure it, so your experience is not universal.
It's not huge and it is dominantly HF.
In no way did i mean to imply a universal experience.
Like said, i've been disappointed by how little my measurements change.
The small changes i've seen are always above 10k, and usually hard to call whether they help or not.
Do you feel your measurement experiences with matching exit to throat are universal?
Like said, i've been disappointed by how little my measurements change.
The small changes i've seen are always above 10k, and usually hard to call whether they help or not.
Do you feel your measurement experiences with matching exit to throat are universal?
I did not say that or imply that. Only that others have had a different experience. There are way too many variables involved for anything consistent to be found.
Somehow, in my mind, I recall measuring a difference, corrected the fit and it was objectively better. I don't have that data and don't remember much of the details as to why I was left with this impression. But this is what I remember. My fit might have been terrible and yours not so much, who knows. But from that point on I was always careful about the fit.
"Audibility?" that's anybody's guess. But I have always been of the opinion that if a change changes the results then one or the other must be better and I always stuck to the better one.