The ripple tank is WYSIWYG and largely self explanatory. Let me know if you get stuck.
Mr. Arai originally published it as A-320FL in the May 1992 issue of the Japanese magazine MJ (which makes sense, since its Fc = 320Hz).
He later mis-labelled it as A290FL on his website.
Finally, he corrected his own mistake and reverted to calling it A-320FL in his 2015 book.
Anyone compared Yuichi A290 with fins with the finless A320FL for sound quality and clarity below say 8khz?
I once tried TAD TH4001 horns with the fins, very similar to Yuichi A290, and I felt they were a bit lacking in clarity and detail compared to my tractrix horns.
I suspect the fins may obscure som detail/clarity..
I once tried TAD TH4001 horns with the fins, very similar to Yuichi A290, and I felt they were a bit lacking in clarity and detail compared to my tractrix horns.
I suspect the fins may obscure som detail/clarity..
@Lament
You compare a highly beaming device like a tractrix horn with a high dispersion fin horn. The tractrix will concentrate most of the HF power along the central forward axis so it might sound "more direct". Although the fins generally cannot be made of an 100% reflecting material so I suspect there will be always some kind of minimal absorption which may depend on the material used for the horn.
Yamamoto is using a special kind of African black wood because of this issue for the fins of their F280 horn:
https://userweb.117.ne.jp/y-s/F280A-e.html
You compare a highly beaming device like a tractrix horn with a high dispersion fin horn. The tractrix will concentrate most of the HF power along the central forward axis so it might sound "more direct". Although the fins generally cannot be made of an 100% reflecting material so I suspect there will be always some kind of minimal absorption which may depend on the material used for the horn.
Yamamoto is using a special kind of African black wood because of this issue for the fins of their F280 horn:
https://userweb.117.ne.jp/y-s/F280A-e.html
@docali I forgot to mention that I lowpassed both horns at 5khz to compare only midrange.
This should be well below where beaming starts to be an issue for a 2" throat.
Very good info about the Yamamoto horn, I guess this is a good indication that the fins create issues.
My two cents following recent experimentation. More likely, on axis the Yuichi 290 plays less loud across its passband (not just at the top end) than a tractrix or JMLC. Therefore without adjusting the level relative to the mid-woofer, one will hear less direct sound from the horn mouth, and more sound coming from the room, as in "less clarity and detail." I measured the difference on axis to be about 2 dB between two other horns, the TH-4001 (with fins) and JMLC AH425. Once I padded the mid-woofer by 2 dB and cranked up the volume to compensate, the clarity did come through the TH-4001. I don't think I can hear the fins of this horn. IMO other differences that I am hearing trump the presence of the fins.
I should add that I am very happy to see this thread revived. Welcome back @Lynn Olson !
Thanks to you and many DIYer that I met here, for about a decade I have enjoyed a two way design using a 15 inch mid woofer on bottom and AH-425 on top. This simple two way has been hard to improve on!
Thanks to you and many DIYer that I met here, for about a decade I have enjoyed a two way design using a 15 inch mid woofer on bottom and AH-425 on top. This simple two way has been hard to improve on!
Thank you Pierre! The goal has always been a modern 2-way that's efficient, with generous headroom, flat response, and clean time decay, without a lot of crossover EQ. Lots of ways to do this, of course, and horn modeling software like we have now did not exist when the thread started. Bjorn Kolbrek's BEM software is what made the AH425 possible.
Now that the Raven Mark II preamp and Statement 300B (Karna Mark II) are in development by Don Sachs in BC, Canada, I have a target set of electronics to use with the big 2-way speaker. The damping factor of the power amp is 4, or Zout of 2 ohms, on the 8-ohm taps, with 27 watts of Class A power per channel. That should work satisfactorily with a 97 dB/meter/watt loudspeaker.
Now that the Raven Mark II preamp and Statement 300B (Karna Mark II) are in development by Don Sachs in BC, Canada, I have a target set of electronics to use with the big 2-way speaker. The damping factor of the power amp is 4, or Zout of 2 ohms, on the 8-ohm taps, with 27 watts of Class A power per channel. That should work satisfactorily with a 97 dB/meter/watt loudspeaker.
Last edited:
My 'speakers are 94dB/m/W efficient, and on dynamically recorded (and minimally compressed) music, I typically end up with peaks at ~50W on the Woofer amp (I bi-amp now). So ~25-27W with 97dB/m/W should be fine - provided you don't expect fully realistic concert levels. That's with classical and some jazz music; with more compressed recordings like most pop and electronica, the peak-to-rms ratio is much more squished, so the same power will allow much higher average SPLs.
What I'm far less convinced by is how any amp with such a high Zout (DF = 4 for a nominal 8-ohm load) can provide real high fidelity. The resulting frequency response will NOT be flat when operated into any real-world load (which is never anything like a constant 8-ohm resistor, even if you bi-amp and/or employ conjugate Z compensation networks). Also, the amp won't be able to shunt out effectively the back-EMF generated by the woofer cone.
I appreciate the attraction of Class A operation, I really do. But, IMHO there are objectively, measurably better options on the table.
For instance, I use an Accuphase A-45, which delivers 45W (*) of pure class A output into 8 ohms, via a push-pull array of MOS-FETs, with only moderate global feedback and a very low (<0.1 ohm) Zout.
(* this is now overkill, since I started bi-amping and only use it for the horn, but previously I was using it to mono-amplify the speakers via a passive crossover)
Marco, isn't this too much? What is your average listening level and at what distance? What would be the RMS vs Pk values of those recordings?
RMS to Pk approx. 20dB (for the best classical and jazz recordings)
Listening distance is approx. 2.5m. This is still fairly close, so in first approximation we can calculate the SPL assuming simple propagation in free space.
For 1 speaker: SPL [Pk,1 spkr] = SPL(1W,1m) + 10 · log (P/D^2) , where P = power (Watts), and D = distance (m)
So: SPL [Pk,1 spkr] = 94 + 10 · log (50/2.5^2) = 94 + 9 = 103 dB @ 2.5m
We have a stereo pair, so assuming uncorrelated L/R signals, we add 3 dB and get SPL [Pk,stereo] = 106 dB @2.5m
Thus, the average listening level is SPL[RMS,stereo] = 106 - 20 = 86 dB @2.5m
This is loud-ISH, but by no means "too much" for the enjoyment of most classical and jazz music.
Now, like I said, for typical overcompressed pop or electronic music (RMS to Pk ~ 6dB), the resulting SPL[RMS,stereo] would be approx 100dB @2.5m, and yes, THAT would be too loud for my comfort.
Bottom line is, the amount of power required "on tap" depends heavily on the music that you wish to listen to!
Also (corollary): most audiophiles listening to "audiophile" recordings of classical and jazz music with low-power amps and low-to-medium efficiency speakers are really listening to their amps CLIPPING on most peaks ;-)
That does seem a lot for a true 94dB/W system. Marco posted while I was writing, but I found similar numbers to his.
I don't like measuring amplifier in watts, I much prefer volts, but wattage is the common spec.
Many of the dynamic recordings that I own have an RMS to peak value of 20-24dB. So if Marco is hitting 20 volts peak (50 watts) then his average voltage is about 1.5 volts. Average power somewhere around 1/3 watt. For a high efficiency speaker in a domestic setting and listening only 2.5 meters away, that seems damn loud on average. I don't think I ever listed to classical music at an average level that high. That level seems a good bit higher than what you'd hear at a classical concert.
That said, I did like to have plenty of headroom on my woofer amps, even at 94dB/2,83V in a medium size room. So an amp rated at 24-50 watts into 8 ohms did not seem overkill.
I don't like measuring amplifier in watts, I much prefer volts, but wattage is the common spec.
Many of the dynamic recordings that I own have an RMS to peak value of 20-24dB. So if Marco is hitting 20 volts peak (50 watts) then his average voltage is about 1.5 volts. Average power somewhere around 1/3 watt. For a high efficiency speaker in a domestic setting and listening only 2.5 meters away, that seems damn loud on average. I don't think I ever listed to classical music at an average level that high. That level seems a good bit higher than what you'd hear at a classical concert.
That said, I did like to have plenty of headroom on my woofer amps, even at 94dB/2,83V in a medium size room. So an amp rated at 24-50 watts into 8 ohms did not seem overkill.
Last edited:
My TD-4001 drivers arrived yesterday, impedance/phase sweeps indicated they were functional. Inspection revealed the expected Be DP-4001 diaphragms. I installed them tonight on my Yuichi A-290 horns and am listening now.
Not too surprisingly they require even less EQ than the JBL, what little there is, is currently done by ear. I expect measurements will confirm it is OK.
No Dirac since with a change this big I need to do a calibration.
I'm pretty bowled over, contrary to what some may have heard these are voiced if anything towards very neutral, detail, imaging, and lack of distortion are in another league from anything I have heard here. (Or elsewhere really) I am very pleased.
The system is largely unequalized with the exception of that implemented in my heavily modified DCX.
Work has been very grueling the past week or so, so I apologize for not posting detailed measurements. It will be a few days.
Not too surprisingly they require even less EQ than the JBL, what little there is, is currently done by ear. I expect measurements will confirm it is OK.
No Dirac since with a change this big I need to do a calibration.
I'm pretty bowled over, contrary to what some may have heard these are voiced if anything towards very neutral, detail, imaging, and lack of distortion are in another league from anything I have heard here. (Or elsewhere really) I am very pleased.
The system is largely unequalized with the exception of that implemented in my heavily modified DCX.
Work has been very grueling the past week or so, so I apologize for not posting detailed measurements. It will be a few days.
Aren't these two really the same thing and wouldn't conjugating, or EQ which we might be using anyway, take care of this?
I now run the 2-way you can see in my avatar. It is a somewhat damped BR with a GPA 416 and a E-JMLC 300 + JBL 2451Be on top. In my room corner placement is working very well so a more dedicated corner implementation is the plan for the future. Among other changes is a new horn for the top to get two things, better low end loading and less beaming. Both the TH-4001 and the A290 have been an inspiration but i am looking at a slightly different implementation.
Thoughts
The plan is to start the throat at zero degrees/r=19mm, have constant height of 38mm and an OSWG profile horizontally for the first ~300mm. I plan to insert 5 fins into that section to get expansion correction to exponential/slight hyperbolic, the fins will not start as close to the throat entry as in the Tad and Yuichi. I will keep the 19mm "corner" radius throughout the horn, that will result in narrower outer "cells", a good thing looking in the Kolbrek/Dunker book. Total length of the horn will be around 600mm. I will use maybe the last 100mm for a gentle horizontal roundover so the horn will probably end up being close to 1m wide.
I also have another pair of 416 woofers so the plan is MTM.
Does any of this make sense?
//Anders
Thoughts
The plan is to start the throat at zero degrees/r=19mm, have constant height of 38mm and an OSWG profile horizontally for the first ~300mm. I plan to insert 5 fins into that section to get expansion correction to exponential/slight hyperbolic, the fins will not start as close to the throat entry as in the Tad and Yuichi. I will keep the 19mm "corner" radius throughout the horn, that will result in narrower outer "cells", a good thing looking in the Kolbrek/Dunker book. Total length of the horn will be around 600mm. I will use maybe the last 100mm for a gentle horizontal roundover so the horn will probably end up being close to 1m wide.
I also have another pair of 416 woofers so the plan is MTM.
Does any of this make sense?
//Anders
It does! Except this:
If you do that, will you not end up with a discontinuity in the expansion law, at the point where the fins begin. I.e., the expansion will NOT be hypex at first (i.e., from the throat until the fins begin, where you have constant height and OS horizontal expansion), and then it would abruptly transition to hypex (once the fins start doing their job of constraining the total horizontal expansion).
I think the reason why the fins start at the very beginning of the horn expansion in the TAD and Yuichi horns is to yield a single, continuous hypex expansion throughout the horn.
(I assume that then, after the first 300mm, you would continue the hypex expansion by combining straight side walls and curved vertical lips, like in the TAD horns).
Marco.
Isn't it crazy how a driver designed over 40 years ago still bests most (in not all) the latest ones developed since?
(provided that it is coupled to the appropriate horn, of course)
Goes to prove that "progress" is often aimed at reducing cost rather than improving ultimate performance. Also, priorities have changed in the pro/PA world, and most (virtually all) recent compression drivers (even large format ones) have been designed to be robust first and foremost, and to be coupled with rapid flare rate waveguides, with little to no regard for mid-range response below 800-1000Hz.
Exactly. And that's a VERY DIFFERENT use case vs. the design goals for this "Beyond the Ariel" speaker (and also for the projects mentioned in many other similar threads on this forum).
So one shouldn't be surprised when - for the latter applications - virtually none of the more "modern" compression drivers manage to perform as well as those "old dinosaurs" like the TADs, or even the JBLs from the same era (the latter ideally "upgraded" with Be diaphragms to get the best of both worlds).
So one shouldn't be surprised when - for the latter applications - virtually none of the more "modern" compression drivers manage to perform as well as those "old dinosaurs" like the TADs, or even the JBLs from the same era (the latter ideally "upgraded" with Be diaphragms to get the best of both worlds).