Jump to content

Celestion Axi2050 2" compression driver on K-402 horn and Jubilee bass bin


Chris A

Recommended Posts

Yes, that's where the 8 dB boost at 18 kHz comes in.

 

Not to beat the subject to death, but I think a short story is warranted on this topic:

 

When I first dialed in the Axi2050 on the right and set it up against the TAD TD-4002 on the left (generally ±1.5-->2 dB using psychoacoustic smoothing), I played each loudspeaker in mono for a while, switching back and forth.  Then I realized that both the amplitude and phase response of the corrected outputs were probably within the driver-to-driver tolerances that I've witnessed over the years, I switched to stereo and listened carefully. I could move left and right from the centerline listening position with the relative channel gains within 0.2 to 0.3 dB using a calibrated microphone at the LP pointed at the ceiling.  (Generally DSP amplifiers have at least 0.25 dB gain resolution in their settings.) This is where I could easily hear the lack of top end or "air" from the Axi2050 side. 

 

I thought about this dilemma for a while (and went on a faster-than-usual dog walk 🦮🦮🦮 😰), then I realized that the Axi2050 has a clamped diaphragm on both the inner and outer diameters, and that, as the frequency rises, the effective moving mass of the diaphragm probably continues to shrink toward the position where the voice coil is attached (about midway between the ID and OD).  If this notion was true, then it means that the diaphragm would still respond to increasing frequency, but with less and less moving area.  So I tried the boost above the natural roll-off frequency (about 14.9 kHz), and found that it responded just as I had thought it might (unlike mass corner effects of other drivers).  Of course, there are phase issues that arise in this regime, but I've learned that "better is better" when it comes to this sort of thing.  The boost appears to work without added chatter, unlike other dome-type 2" compression drivers.

 

So, bottom line, if using the Axi2050, I strongly recommend the 18 kHz boosting PEQ, even though it might look a little extreme, it seems to just compensate for the decrease in on-axis SPL. 

 

D29178D3-8E4F-4270-B03C-85B841F95718.png

 

I still haven't run a polar plot on this driver and on the effect of the lower crossover frequency within the K-402-MEH.  This task is still waiting on me to move the Axi2050 into my center K-402-MEH and move it outside to measure its polars (in 10 degree increments--horizontally, then vertically).  I feel this is warranted from the conversation with Roy on this driver in the added throat-located acoustic lens to "spread out" the polars above 6 kHz.  I'm more than a bit leery of adding an additional object within the pathway of the horn throat entrance without it causing other issues--like higher order modes and other diffraction events--but I'd accept measurements--higher order mode measurements with a moving microphone across the throat entrance to measure for HOM generation with and without the added throat lens.

 

YMMV.

 

Chris

 

 

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

The spectrogram view of the loudspeaker can also be used to see the decay characteristics of the compression drivers more succinctly (IMO) than using waterfall views (which tend to be very sensitive to plot settings, in my experience).  First, the TAD TD-4002 on a K-402 horn, with Jubilee bass bin (KPT-KHJ-LF):

 

1438662211_TADTD-402onaK-402HornwithJubBassBIn.jpg.e3138fc3446957f5f66fe9b0b97e190a.jpg

 

and the Celestion Axi2050 on the K-402 over a Jubilee bass bin:

 

832923930_CelestionAxi2050onaK-402HornwithJubBassBIn.jpg.afbe8f94e8cdce7655244238b0e49fe8.jpg

 

The cleanness of the output above 10 kHz shows that the boost of the Axi2050 at 18 kHz seems to work without accompanying non-pistonic modes or phase plug issues. 

 

The small amount of acoustic energy shown around 0.7-1.1 ms above 10 kHz corresponds to the distance of the acoustic center of the driver to the depth of the K-402--the mouth distance (about 15-16 inches) with a horn mouth roll-out starting at about 2/3s that distance (a very slight acoustic impedance discontinuity). 

 

The following lower level echos shown above 4 kHz at 1.5 to 2.5 ms in the acoustic background could be an area for further investigation using a moving microphone across the horn mouth or throat in order to discover any higher order modes that are being generated by the compression driver or its interface with the horn throat.  These HOMs effectively have a longer path to travel within the horn and therefore will appear after the initial acoustic energy of the primary horn mode, so they will exit the K-402 horn mouth later.  This later acoustic energy could also be explained by the higher moving mass of the AXI2050 diaphragm than the TD-4002 diaphragm (and therefore longer delay times vs. frequency).

 

The dog leg in the Axi2050 peak energy time curve at 300 Hz is an artifact of the dialing in process to get a smooth crossover band between the K-402 and the Jubilee bass bin.  More iterations of the compression driver channel delay trades and resulting phase growth are required to smooth that dogleg in the curve, but the subjective listening performance is not materially affected by the current settings. 

 

rePhase (i.e., FIR filtering) can also be applied on top of these settings used to smooth out this crossover transition dogleg, but only for the microphone position being used.  This is one reason why I don't typically use FIR filters, since they only partially affect the underlying acoustics issues, and usually only for a very small location volume of the microphone--like "room correction software" packages suffer from because the microphone is picking up too many in-room reflections if placed at the listening position, and winds up trying to correct the non-minimum-phase reflections (which it can't correct). 

 

Chris

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

On 3/8/2021 at 12:55 PM, Bjorn said:

I tested the AXi2050 on K-402 in 2019 and did a direct comparison to Radian 950BePB. Back and forth for several days both in mono for quick comparisons and listening in stereo, though swapping for stereo takes time when you only have one pair of speakers. 

 

My experience was quite different from Chris'. The Radian with Be sounded considerably clearer and more open in both the upper midrange and tweeter. Distortion measurements confirmed this. Crossover lower was nice though and mye experience her was also different. While it didn't experience this as night and day, it was still more than a subtle improvement in a coherent sound stage and vocal presentation. But overall, the Axi2050 sounded dull in the upper frequencies making it boring to listen to compared to Radian. 

 

But I'm not surprised Klipsch is choosing such a driver for a commercial speaker and that solved issues with the bass bin they use.

 

Slightly off-topic, but you are one of the few I have seen to have tested the Radian 950BePB.  Have you also tested or heard TAD beryllium drivers?  Can you compare the two?

 

I have always been curious about the Radian, as they are available new for somewhat less than the TAD drivers are used.  They seem like they could be a good alternative if they perform similarly.  Very hard to find any information from those who have experience with both.  Thanks.

Link to comment
Share on other sites

14 minutes ago, Tarheel TJ said:

 

Slightly off-topic, but you are one of the few I have seen to have tested the Radian 950BePB.  Have you also tested or heard TAD beryllium drivers?  Can you compare the two?

 

I have always been curious about the Radian, as they are available new for somewhat less than the TAD drivers are used.  They seem like they could be a good alternative if they perform similarly.  Very hard to find any information from those who have experience with both.  Thanks.

No. I haven't tested TAD drivers. Today I'm using Radian 951BePB (1.4") exit with a horn that I got developed together with Don Keele. Picture below and the best horn I've heard by far. 

595254361_80x50HFhornandmidbasshorn(Liten).jpg.72cddf3aa4e52546f90a2d1797bded23.jpg

 

 

But I know some who have tested Radian Be (950 I think), different JBL drivers with Truextent Be (i.e. 2451) and different TAD drivers and have also measured them. If I remember correctly they were basically very close and perhaps not possible to distinguish audibly after applying EQ, apart from TAD 4003 which the person found to be slightly better than JBL 2451 Be (I don't think he tested Radian 951BepB). TAD 4003 is also the driver that measures overall best and seems to both go a bit lower in frequency and has more level/SPL in the highs. Unfortunately it's an exceptional expensive driver (not sold new anymore) and I assume we're talking about a minor improvement after equalization. The horn is certainly way more important, which the person who has tested these drivers also agrees to. 

 

  • Like 3
Link to comment
Share on other sites

On 3/8/2021 at 1:58 PM, Chris A said:

Yes, that's where the 8 dB boost at 18 kHz comes in.

One thing that I should point out: this is yet another instance where passive crossovers will not get you the performance that you desire.  You simply cannot correct the SPL response of the Axi2050 driver above 14.9 kHz without a DSP crossover (or some other PEQ capability upstream of the amplifier channel).  I actually do not recommend trying to use the Axi2050 without a DSP crossover.

 

When you look at the present economics and performance levels of using DSP crossovers, what you leave on the table in terms of performance by trying to use passives is just too wide a difference.

 

Chris

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

3 hours ago, Bjorn said:

No. I haven't tested TAD drivers. Today I'm using Radian 951BePB (1.4") exit with a horn that I got developed together with Don Keele. Picture below and the best horn I've heard by far. 

595254361_80x50HFhornandmidbasshorn(Liten).jpg.72cddf3aa4e52546f90a2d1797bded23.jpg

 

 

But I know some who have tested Radian Be (950 I think), different JBL drivers with Truextent Be (i.e. 2451) and different TAD drivers and have also measured them. If I remember correctly they were basically very close and perhaps not possible to distinguish audibly after applying EQ, apart from TAD 4003 which the person found to be slightly better than JBL 2451 Be (I don't think he tested Radian 951BepB). TAD 4003 is also the driver that measures overall best and seems to both go a bit lower in frequency and has more level/SPL in the highs. Unfortunately it's an exceptional expensive driver (not sold new anymore) and I assume we're talking about a minor improvement after equalization. The horn is certainly way more important, which the person who has tested these drivers also agrees to. 

 

Thanks for the response!  Sounds like any of the Be drivers are a good choice.

 

I have seen that Keele horn referenced a few times on the 'net.  Any more information you can share about it?  Are they commercially available?

 

 

Link to comment
Share on other sites

1 hour ago, Tarheel TJ said:

Thanks for the response!  Sounds like any of the Be drivers are a good choice.

 

I have seen that Keele horn referenced a few times on the 'net.  Any more information you can share about it?  Are they commercially available?

 

 

I'm not allowed to share any such information here. That would kick me out of here. 

Link to comment
Share on other sites

Interesting to see you have given this driver a try.

 

It was my intention to stay away but I will give it another try.

 

For the upper frequencies boost this is a good place to use the HI SHELF option on the xilica.

 

Use a very high Q setting around 7000 Hz.  Boost will be dependent upon your setup.  I would say exactly what it is but i am at work. You will gain a bit more phase change but also gain a smoother response.  I go with Geddes dictum of frequency response is the most important of all parameters.

 

I had noticed (using the RTA) that while listening to music there was very little output at the high frequencies so why not boost it up there?  I have boosted as high as the xilica will allow and at home listening levels I never heard any bad effects.  This was when i was trying to retain the sensitivity in the lower regions.  Now I am splitting the difference between the low and high to keep my filters in balance - too much boost or cut affects other parts of the response more than we want. I am not sure if amplifier quality has a bearing here asking for this boost at these frequencies.  I have not tried any amplifiers other than the ones I have.

 

I have found the MEH sounds best when you throw away most of the acoustic transformation.  Luckily even with low powered amplifiers the things get plenty loud.

 

I think it is a bit of a misnomer to consider the K401 MEH a horn loudspeaker in the classical sense - at its best we are using it as a waveguide.

 

I have come to conceive that most of what most people object to in a horn loudspeaker is the sound of the acoustic transformation,  The most notable exception is Romy Besnow who seems to thrive on it but then he is using his horns in very limited frequency ranges so the acoustic transformation is more equal between the various horns whereas with our systems we are blending the two modes and I suspect this discontinuity is what most people identify as "horn sound".

 

No question in my mind the CELESTION is the best compression driver I have ever heard.  I never got a chance to try the 4002 which is said to be greatly superior to the 4001s I did use for few years.  The 4001 is said to only work with its companion horn.  Using the Edgar round horn it sounded pretty bad - but the reputation of the thing kept me listening to it long after my ear/brain said REPLACE THEM. Of course, everyone knows this now.

 

My system has been completed for a few months now and I still keep tweaking the EQ - funny thing is - no matter what I do the intrinsic nature of the loudspeaker is always there and the intrinsic nature is very fine.  There is no doubt when one finds something that works better but this is not hyper critical. One could easily come up with a decent sounding setup without much trouble.  But of course, with a fascinating puzzle like the MEH and xilica the desire to try to get it a little better every day is impossible to resist,  getting really close though.

 

No question the speakers are dynamic beyond my expectations.  Drums sound more like drums than I have ever heard - even at reasonable volume, there is no need for them to be LOUD to hear something that sounds like a tom or a rimshot. - the center image is rock solid and the space between center to the sides are well defined - the loudspeaker disappears.  The tonality is simply realistic.  Massed strings sound very refined.  Vocals can be eerie in their realism.  Blah, blah, blah

 

I had hoped these speakers would be better than my previous setups - I knew they would be better and I knew they would take up much less space but I was and remain extremely happy with how much better they are.  I cannot imagine anything I could afford that could come close to what I am hearing.

 

I have found that placing a two inches sheet of 20 ppi reticulated foam with a cut out for the opening - but a bit smaller than the horn opening makes for a greatly improved sound.  This seems to be well documented as something that improves any horn..  In my case it was instantly apparent.  One would think there would be a loss somewhere in the response but in actuality dips in response in the 5 to 7 kHz region were mitigated.

 

Another thing to try is a very stretched out bit of long fiber wool right at the CD opening.  Visually transparent - I used two sided tape.to keep it in place.  Makes a minimal change in measured response at 18 to 20 kHz but the the sound becomes much more refined.  I think both of these things are affecting the horn far more than the driver.

 

No question the 401K MEH is a great loudspeaker, a great stereo loudspeaker and the Celestion driver is an important part of the whole in my experience.

  • Like 2
Link to comment
Share on other sites

Welcome back, Rick. BTW: I think you are referring to the K-402 horn--not the K-401:

 

Klipsch K-401 horns:

 

IMG_2131.JPG

 

Klipsch K-402 (with stand) taken in my garage in December 2007:

CRW_2205.gif

 

 

3 hours ago, rickmcinnis said:

I have found the MEH sounds best when you throw away most of the acoustic transformation. 

 

How do you do this, Rick?  Do you decouple the driver from the horn in order to break horn loading?

 

The K-402 has significant gain down to 170 Hz, and tapers down to boundary gain below that point rather smoothly (raw, un-EQed response, half space/quarter space):

 

K-402-MEH woofers only (No EQ).jpg

 

Chris

  • Like 1
Link to comment
Share on other sites

Interesting to see you have given this driver a try.
 
It was my intention to stay away but I will give it another try.
 
For the upper frequencies boost this is a good place to use the HI SHELF option on the xilica.
 
Use a very high Q setting around 7000 Hz.  Boost will be dependent upon your setup.  I would say exactly what it is but i am at work. You will gain a bit more phase change but also gain a smoother response.  I go with Geddes dictum of frequency response is the most important of all parameters.
 
I had noticed (using the RTA) that while listening to music there was very little output at the high frequencies so why not boost it up there?  I have boosted as high as the xilica will allow and at home listening levels I never heard any bad effects.  This was when i was trying to retain the sensitivity in the lower regions.  Now I am splitting the difference between the low and high to keep my filters in balance - too much boost or cut affects other parts of the response more than we want. I am not sure if amplifier quality has a bearing here asking for this boost at these frequencies.  I have not tried any amplifiers other than the ones I have.
 
I have found the MEH sounds best when you throw away most of the acoustic transformation.  Luckily even with low powered amplifiers the things get plenty loud.
 
I think it is a bit of a misnomer to consider the K401 MEH a horn loudspeaker in the classical sense - at its best we are using it as a waveguide.
 
I have come to conceive that most of what most people object to in a horn loudspeaker is the sound of the acoustic transformation,  The most notable exception is Romy Besnow who seems to thrive on it but then he is using his horns in very limited frequency ranges so the acoustic transformation is more equal between the various horns whereas with our systems we are blending the two modes and I suspect this discontinuity is what most people identify as "horn sound".
 
No question in my mind the CELESTION is the best compression driver I have ever heard.  I never got a chance to try the 4002 which is said to be greatly superior to the 4001s I did use for few years.  The 4001 is said to only work with its companion horn.  Using the Edgar round horn it sounded pretty bad - but the reputation of the thing kept me listening to it long after my ear/brain said REPLACE THEM. Of course, everyone knows this now.
 
My system has been completed for a few months now and I still keep tweaking the EQ - funny thing is - no matter what I do the intrinsic nature of the loudspeaker is always there and the intrinsic nature is very fine.  There is no doubt when one finds something that works better but this is not hyper critical. One could easily come up with a decent sounding setup without much trouble.  But of course, with a fascinating puzzle like the MEH and xilica the desire to try to get it a little better every day is impossible to resist,  getting really close though.
 
No question the speakers are dynamic beyond my expectations.  Drums sound more like drums than I have ever heard - even at reasonable volume, there is no need for them to be LOUD to hear something that sounds like a tom or a rimshot. - the center image is rock solid and the space between center to the sides are well defined - the loudspeaker disappears.  The tonality is simply realistic.  Massed strings sound very refined.  Vocals can be eerie in their realism.  Blah, blah, blah
 
I had hoped these speakers would be better than my previous setups - I knew they would be better and I knew they would take up much less space but I was and remain extremely happy with how much better they are.  I cannot imagine anything I could afford that could come close to what I am hearing.
 
I have found that placing a two inches sheet of 20 ppi reticulated foam with a cut out for the opening - but a bit smaller than the horn opening makes for a greatly improved sound.  This seems to be well documented as something that improves any horn..  In my case it was instantly apparent.  One would think there would be a loss somewhere in the response but in actuality dips in response in the 5 to 7 kHz region were mitigated.
 
Another thing to try is a very stretched out bit of long fiber wool right at the CD opening.  Visually transparent - I used two sided tape.to keep it in place.  Makes a minimal change in measured response at 18 to 20 kHz but the the sound becomes much more refined.  I think both of these things are affecting the horn far more than the driver.
 
No question the 401K MEH is a great loudspeaker, a great stereo loudspeaker and the Celestion driver is an important part of the whole in my experience.
Do you have any pics of the mods?

Sent from my SM-T830 using Tapatalk

Link to comment
Share on other sites

One of the interesting thoughts that came to mind when looking at the K-402/Axi2050 performance is that the Axi2050 driver basically makes full use of the K-402 horn, i.e., the K-402 begins to lose horizontal directivity below 170 Hz, and that's frequency at which the Axi2050 performance really drops off the deep end (relative to the sensitivity of the Axi2050 above 6 kHz requiring similar PEQ boost for flat SPL response, or conversely, PEQ attenuation between 170 Hz and 6 kHz for similarly flat SPL response):

 

Celestion Axi2050 on K-402 and Jubilee Bass Bin Raw SPL Response.jpg

 

What this means to Khorn owners is that the Axi2050/K-402 horn on top of a Khorn bass bin would be a very good combination.  The only remaining need would be subwoofer performance below ~25-30 Hz to achieve a full range performance at Jubilee-like sound quality levels.  The Khorn bass bin really doesn't like to go above ~230 Hz looking at its phase response, and noticing that little "dogleg" of phase response right around 230 Hz (red trace, below):

 

Comparison Klipsch Jubilee Bass bin  (blue) vs. Clone (red) Phase response.jpg

 

The other good thing about crossing so low with the Axi2050 and the Khorn bass bin (about 230 Hz, that is) is that it also avoids the dual mouth polar coverage issues that are exaggerated with the widely split dual-mouth Khorn bass bin (relative to the Jubilee bass bin--the KPT-KHJ-LF). 

 

So in balance, I don't believe that there would be much difference in sound quality if using the Khorn bass bin or the Jubilee bass bin under a K-402/Axi2050 high frequency horn crossing at 230 Hz.  The phase and SPL response--as well as the polar coverage angles--would be pretty close to one another.  What this means is that Jubilee-like performance is still available to Khorn owners that are willing to spring for the K-402/Axi2050 costs, and using a good DSP crossover, such as a Xilica 2040/4080 or even just a miniDSP 2x4 HD ($205).  

 

JMTC.

 

Chris

  • Thanks 1
Link to comment
Share on other sites

I read somewhere on diyAudio that some guy was using a pair of Axi2050s without horns--like they were full-range loudspeakers.  I assume he flattened the response a bit using some form of EQ somewhere upstream...but it is an interesting notion.

 

I bet they sound like really good table radios.

 

Chris

  • Like 2
  • Haha 2
Link to comment
Share on other sites

  • 2 weeks later...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...