Josh Ricci's comment on the Klipsch forum:
"These drivers are rugged and do seem to give more useful low end extension than most other 2" format cd's. It may be a bit over hyped depending on what you read and the application. For domestic use, where it'll never see more than a couple volts on peaks, it can run very low without any risk. For large format / HO use I'd not cross it lower than 500Hz. For my situation and type of use, which is somewhere smack in between light domestic use and heavy duty reinforcement, I wouldn't run it lower than 350-400Hz on the 402 to keep diaphragm excursion and THD in check."
"These drivers are rugged and do seem to give more useful low end extension than most other 2" format cd's. It may be a bit over hyped depending on what you read and the application. For domestic use, where it'll never see more than a couple volts on peaks, it can run very low without any risk. For large format / HO use I'd not cross it lower than 500Hz. For my situation and type of use, which is somewhere smack in between light domestic use and heavy duty reinforcement, I wouldn't run it lower than 350-400Hz on the 402 to keep diaphragm excursion and THD in check."
First off, congrats on landing what is looking like a very nice horn!
And nice to see you out making measurements, and very nice to see guys helping with great specific advice...all of which I say ditto to
I'll make a few suggestions, hopefully they help.
Strongly agree with everyone saying to high-pass at no lower than 350-400Hz.
Just because a CD shows loading down to a certain frequency on a large horn, doesn't mean it's going to sound good doing so.
I keep finding i like what i hear, when i high pass a half-octave or better from the CD/horn's low corner.
Maybe it's because the low loading looks good at small signal, and falls apart driven harder...i dunno...kyleneuron gave a good way to check that.
Maybe IMD increases substantially the lower the high-pass....dunno that either....although that's my current belief (that hopefully can be proven/disproven with multi-tone testing.)
Anyway, for me, i'd be far more concerned with what xover freq/type/order mates the axi/horn with your lower driver(s)/box,
than trying to take the CD/horn as low it looks like it will go, and still give adequate headroom at desired SPL...
Far more......
Oh, for measurements.....strongly suggest you get a 2-ch or greater USB sound card and XLR mic...
UMIK is ok for starters, but dual channel loopback much better....
If at all possible, do try to get to far-field.
And don't sweat much over measurements above 15kHz...too hard to get exactly right anywhere, and not worth the trouble when you do, imo/ime
Besides, every good system should have easy to use high and low shelf EQs available, at playback time.
So far there is no insight into the impedance of Camplo's horn + Axi2050.
However, based on the response, in combination with scientific research, educated guesses can be made about this.
As indicated by, among others, DonVK, the farfield response is important, because "there is excellent correlation between the frequency peaks in the far-field pressure and the throat impedance".
Lord Rayleigh stated "when the section of a pipe is variable, the problem of the vibrations of air within it cannot be generally solved."
An important phenomenon in classical horn types (Expo, Hypex, Tractrix) is the group velocity effect near cut-off that could cause waveform distortion.
This 25-year-old paper, using BEM analysis of a Klipsch K400 horn, provides important insights.
However, based on the response, in combination with scientific research, educated guesses can be made about this.
As indicated by, among others, DonVK, the farfield response is important, because "there is excellent correlation between the frequency peaks in the far-field pressure and the throat impedance".
Lord Rayleigh stated "when the section of a pipe is variable, the problem of the vibrations of air within it cannot be generally solved."
An important phenomenon in classical horn types (Expo, Hypex, Tractrix) is the group velocity effect near cut-off that could cause waveform distortion.
This 25-year-old paper, using BEM analysis of a Klipsch K400 horn, provides important insights.
Last edited:
There would be a high correlation between the far field power response and the impedance. However, along any single axis one could have diffraction effects from the mouth, which would not correlate to the electrical impedance. That paper talks about the throat impedance, not the electrical impedance.
Regarding the approximate 3volt measurements... removing limiting is familiar from sound design so the concept is understood and in practice.... In other words this
is 1m and ~listening distance. Though 52" was nice when demoing the mid and bass together. Adding the horn will provide a complete view of drivers involved
my voltmeter needs batteries... I have the usp4 dsp card...
Crown CTS
I think they are clean enough...I do hear the noise floor with the gain turned way up, but when will I ever run 115db average from my tweeter 🤷🏽♂️...theres some sensitivity settings that might fix this little issue....
I just plugged the Umik into my laptop running windows `10.... no special windows processing, as that has been for me, status quo for some time.
An unbalanced stereo output from the laptop to the amps inputs.
I'd like to think that the simulation is at least in the ballpark for displacement.... anything to make the sim better, I would like to do....
Impedance on Horn
is 1m and ~listening distance. Though 52" was nice when demoing the mid and bass together. Adding the horn will provide a complete view of drivers involved
my voltmeter needs batteries... I have the usp4 dsp card...
Crown CTS
I think they are clean enough...I do hear the noise floor with the gain turned way up, but when will I ever run 115db average from my tweeter 🤷🏽♂️...theres some sensitivity settings that might fix this little issue....
I just plugged the Umik into my laptop running windows `10.... no special windows processing, as that has been for me, status quo for some time.
An unbalanced stereo output from the laptop to the amps inputs.
I'd like to think that the simulation is at least in the ballpark for displacement.... anything to make the sim better, I would like to do....
Impedance on Horn
Last edited:
That's true, of course, but since Camplo's horn is what it is, there's not much he can do about it.
Regarding mouth termination (or the lack thereof), Bjørn Kolbrek's horn seems suitable for comparison.
Thanks!
The conclusions indeed remain virtually the same:
- Fc: around 150 Hz
- Proper loading > 200 Hz
- Usable XO > 300 Hz
- Optimal XO > 500 Hz.
Last edited:
My guess is that a mouth termination that would eliminate the resonances on the lower end would have to be huge. The longer the horn, the bigger it must be to have an effect - with such a termination it could easily grow twice in size. It's a lot easier to properly terminate a shorter horn, which would be usable from 500 Hz as well, without any inherent ringing. Just my two cents.
This particular measurement has a high noise floor and visible ripple in the raw PIR time response, so I wouldn’t trust it personally. It could be that you were hitting the limiters in the amp at some frequencies, depending on how the limiter is configured. Peak limiting on these amps is typically ‘look ahead’.
The processing is on by default for new devices in Windows 10 & 11, so there’s a good chance it is affecting your data. As the REW manual states, the UMIK-1 presents as a stereo device with identical data on both channels. This is likely to trigger any echo cancellation software.
I’ve fallen afoul of this myself when I recently got an SSL2+ audio interface. Despite using ASIO for measurement software, Windows had enabled it as the default input device via WDM/WASAPI, with ‘enhanced processing’ set to on. It took me two hours of wondering why my data looked so weird before I checked.
I now manually disable my measurement interfaces from being accessed by anything other than ASIO, just to be safe. You can’t do that with the UMIK.
As well as potential noise leakage from the laptop power supply or cross talk from the amp mains cable, most laptops have a Realtek chip for the built in audio device. Those can exhibit timing errors and minor distortion when Wi-Fi is enabled, due to hidden battery optimising power management settings in Windows.
I presume you have a good external soundcard for your studio? Use that as the output device if you can, with balanced cable to the amp.
If you can’t, then I strongly recommend disconnecting your laptop from the mains and turning off wifi before you initiate a measurement sweep.
A good electrical impedance trace is the best tool to improve your model. When the measurement and simulated peaks in Ze match, then you have a good model. If they’re off, you need to tweak the model parameters accordingly or check your measurement setup.
Again there appears to be a lot more noise in this data than I would expect. I have a DATS box too, and while it’s far from perfect my traces rarely show so much ‘fur’.
All of those closely spaced peaks suggest noise in the measurement setup. This could be from timing errors, noise from cables running alongside mains power, or a physical reflection such as having the horn mouth very close to a large & dense boundary surface when making the sweep.
I would pause and check everything before making any major conclusions.
It could be an attempt to keep the group delay low at the acoustic crossover, and maximise efficiency by keeping the velocity maxima coupled at the 'big end' and 'small end'. It took me a while to find the post but to quot Tom Danley:
It's worth noting that these comments are regarding a theoretical 'perfect horn' where the mouth radius is at least 1 wavelength for all frequencies above cutoff.
Since a horn is a phase inverting device, it might also be desirable to enable the use of passive crossover networks with even order electrical filter slopes.
It's difficult to read the dimensions in the diagrams Ro808 posted, but the drawing appears to show two different horns. Eyeballing the relative sizes, neither appear to have a mouth size in either the horizontal or vertical to support a cutoff below 500Hz. It would need a mouth radius of >0.7m to do that.
Last edited:
Kyle, do you think the results will be fundamentally different with better measurement conditions/tools/methods?
I completely agree with your observations about noise etc., by the way.
As I mentioned before, impedance+phase (at the low end) should broadly correspond to the plot of the eTrak 300 + BMS 4590 (mid), only shifted to the left.
I completely agree with your observations about noise etc., by the way.
As I mentioned before, impedance+phase (at the low end) should broadly correspond to the plot of the eTrak 300 + BMS 4590 (mid), only shifted to the left.
Tom knows what matters
The mouth of that long horn is 'substantial' compared to the Arai A480FL-sized radial horn next to it, though still relatively small compared to the length.
Last edited:
"Transient perfect" was never shown to be a factor in subjective sound quality. The only relevant point I see would be the maximization of efficiency, which would make sense for PA but certainly not for a home/studio use.
In other words, I see a long horn with quite low "cut-off" but which is also pretty nasty. So what they do is they don't use the horn that low to not sound bad. It just makes no sense to me. I only asked to be sure I didn't overlook something.
In other words, I see a long horn with quite low "cut-off" but which is also pretty nasty. So what they do is they don't use the horn that low to not sound bad. It just makes no sense to me. I only asked to be sure I didn't overlook something.
Last edited:
Subjective sound quality is exactly what it is: subjective (in spite of double blind ABX testing and what have you).
Tom Danley is an knowledgeable and inspirational audio engineer, but that doesn't mean I value all of his products equally.
The big horn system could indeed be an exercise in horn loading in order to boost efficiency.
The majority of the 2-way systems of this brand are more conventional.
One reason for the relatively high crossover frequency might be the vulnerability of the Be diaphragm.
Tom Danley is an knowledgeable and inspirational audio engineer, but that doesn't mean I value all of his products equally.
The big horn system could indeed be an exercise in horn loading in order to boost efficiency.
The majority of the 2-way systems of this brand are more conventional.
One reason for the relatively high crossover frequency might be the vulnerability of the Be diaphragm.
