REW’s Auto EQ only uses peaking filters in it’s algorithm so it will never suggest anything else 😀
You can input a shelving filter yourself and as long as that eq is set to manual and not auto it won’t be overwritten when you run the auto algorithm again.
Where the target level is set has a big influence on what the algorithm will do experimenting with that is helpful to understand how it works.
Yes I added the shelving filter manually in this case. This what I ended up with after some listening. Dip at 1.4kHz, -3.1dB, Q =1 and the shelving filter.
Yes, I use a DCX for mains cross over, time delays for driver offset correction, and limited EQ. My DCX2496 has outboard linear power supplies, improved master clock, Amb Labs Analog I/O board, and countless other modifications. As a recent convert from analog line level signal processing the fact that the DCX could be made to sound good seemed worth the gamble.(Digital in only)
Yes, I use a DCX for mains cross over, time delays for driver offset correction, and limited EQ. My DCX2496 has outboard linear power supplies, improved master clock, Amb Labs Analog I/O board, and countless other modifications. As a recent convert from analog line level signal processing the fact that the DCX could be made to sound good seemed worth the gamble.(Digital in only)
Any likelihood of a repost somewhere outside that walled garden? If not, I know someone with a key...
Radian 950PB. I do not understand a thing about CD's but, the step in the exit, looks awful. I am not sure all the S/N have this termination.
Attachments
Could it be that the R950PB is the R951PB with a conical exit section? From these pictures it should be clear that the 2" driver is a mismatch to TH4001 as you should not use a high flare rate section into a low flare rate section of the horn. What would better work to use the R951 and design an appropriate adapter.
Btw, it would be still nice to see the impedance of the JBL attached to the fin horn. Thanks.
Btw, it would be still nice to see the impedance of the JBL attached to the fin horn. Thanks.
It's really the flare rate (i.e., rate of area change over distance, divided by area), not the length per se.
Expansion calculations, adapter length and driver angle
The angles for expansion (to manage axial reflection) are not necessarily the same as the waveguiding angles in either dimension (to manage HOM). In regards to expansion...
1. The Existing Throat Angle
Assuming a radial wavefront (cylindrical), the wavefront area expansion over a sample of the first 5.08cm(2") of travel (as shown, being from 28.7cm2-43.9cm2), is the equivalent of a conical expansion with a total included angle of no more than 16 degrees.
2. The Adapter
Matching that(*), going back, and assuming the adapter does the double duty of wavefront shaping. The area of the 5.08cm diameter (assumed plane) wavefront at the driver exit, to the radial area of the horn throat (being from 20.3cm2-28.7cm2), means that for the equivalent of 16 degrees it should be 3.4cm long (the wavefront varies, but the 25mm shown in the plan is about right in the middle).
* A more narrow angle might be considered in the interest of following the original intended expansion, which I haven't covered here, otherwise...
3. The Driver
Then only needs to have a total angle of about 16 degrees (or, again, less). If it doesn't, the adapter length might be adjusted to compensate.
The angles for expansion (to manage axial reflection) are not necessarily the same as the waveguiding angles in either dimension (to manage HOM). In regards to expansion...
1. The Existing Throat Angle
Assuming a radial wavefront (cylindrical), the wavefront area expansion over a sample of the first 5.08cm(2") of travel (as shown, being from 28.7cm2-43.9cm2), is the equivalent of a conical expansion with a total included angle of no more than 16 degrees.
2. The Adapter
Matching that(*), going back, and assuming the adapter does the double duty of wavefront shaping. The area of the 5.08cm diameter (assumed plane) wavefront at the driver exit, to the radial area of the horn throat (being from 20.3cm2-28.7cm2), means that for the equivalent of 16 degrees it should be 3.4cm long (the wavefront varies, but the 25mm shown in the plan is about right in the middle).
* A more narrow angle might be considered in the interest of following the original intended expansion, which I haven't covered here, otherwise...
3. The Driver
Then only needs to have a total angle of about 16 degrees (or, again, less). If it doesn't, the adapter length might be adjusted to compensate.
Mine look exactly like that. I thought briefly that it might be possible to make an adaptor that actually fits into the driver mouth, but machining something like that around here could cost more than a pair of TD-4001. @docali I will do an impedance/phase measurement on one of the horns today.
These days you could 3D print an adaptor like that, even when using an online print service it should be more affordable.
All you'd need is the right 3D model, there are people here that could help with that. Or make your own with an App like Fusion 360 from Autodesk.
@docali
Here are the phase and impedance plots for both JBL drivers on horn and in free air for comparison.
The device driver 1 performs a little worse in terms of overall THD in sweeps and despite being NIB from JBL many years ago this diaphragm has a slight blemish which seems to have made it past JBL QC.
Driver 1 JBL2440 with 2441 diaphragm (left channel) on horn.
Driver 1 (left channel) JBL2440 with 2441 diaphragm in free air
Driver 2 (right channel) JBL2440 with 2441 diaphragm on horn
Driver 2 (right channel) JBL2440 with 2441 diaphragm
Here are the phase and impedance plots for both JBL drivers on horn and in free air for comparison.
The device driver 1 performs a little worse in terms of overall THD in sweeps and despite being NIB from JBL many years ago this diaphragm has a slight blemish which seems to have made it past JBL QC.
Driver 1 JBL2440 with 2441 diaphragm (left channel) on horn.
Driver 1 (left channel) JBL2440 with 2441 diaphragm in free air
Driver 2 (right channel) JBL2440 with 2441 diaphragm on horn
Driver 2 (right channel) JBL2440 with 2441 diaphragm
Last edited:
Mayne thanks.
In my previous post I mixed up your curves - my bad. The yellow like curve is impedance - right?
Here the Radian and it is very clear that it is a serious mismatch. This is near to a no-go. That impedance peak is responsible for your fr hole.
Compared to this the JBL has a very even impedance attached to the horn indicating a good match:
From these figure I would clearly recommend to use the JBL when you would not have said that it distorts to some degree. Sad to here this but is it still possible to get new diaphragms? It would be worth imo.
In my previous post I mixed up your curves - my bad. The yellow like curve is impedance - right?
Here the Radian and it is very clear that it is a serious mismatch. This is near to a no-go. That impedance peak is responsible for your fr hole.
Compared to this the JBL has a very even impedance attached to the horn indicating a good match:
From these figure I would clearly recommend to use the JBL when you would not have said that it distorts to some degree. Sad to here this but is it still possible to get new diaphragms? It would be worth imo.
The phase impedance plots of the JBL "2441" look pretty benign to me, easy amplifier load, and no phase shift between 1 - 3kHz, above and below very gentle phase shift with frequency. It sounds fairly good with Dirac3 calibration and a slightly tweaked HK target curve. Detail is quite good, still some concerns over distortion, might be a bias given the history of these drivers in my system previously.
Hi Docali,
Yes the yellow is impedance, and like you I interpreted that impedance peak as basically a no go indication for the 950 on this horn - that in fact was what drove the switch.
Somewhat ironic that I have all of this lovely test equipment and I could have known this immediately upon installing these drivers on the horns had I tested them. It would likely have been obvious to me in a couple of days that no amount of EQ was going to fix this problem. Part of this is learning curve (not sure I would have fully understood the implications), but I really ought to have done this test, it takes me literally less than 5 minutes to set up and do this test.
The most interesting thing was that the response measurements with appropriate EQ looked pretty good, it just never sounded quite right. Above 2kHz I would say it outperformed the JBL, below 2kHz the reverse is very much the case. It is disappointing because the 950 is a better driver than the 2440/2441, but incompatible with this horn, and I was stubbornly naive to think I could make it work. I hope if nothing else that others will benefit from my experience.
Last edited: