Kolbrek said:You mean the cutoff concept here, right? I would say the importance of cutoff related issues depends on how close the cutoff is to the pass band of the driver/horn combination. In bass horns (which I believe you are not much involved in designing), it gets more important. But given too small mouth, wich is common, there is virtually not difference on horn shapes.
True, but high resistive loading also affects diaphragm displacement, which can be important sometimes. The picture is complex here, and the relative importance of the factors depends on the frequency range of the horn/driver combination.
I do, BTW, want to thank you for the all the interesting and useful work you have done on horns/waveguides, I believe I have read all your JAES papers on that topic
But we (I) are maybe throwing this thread far off topic - sorry for that.
Last first, I don't think that we are far off topic, or I wouldn't be here. The topic was Horns and that is my main interest. I should point out my other thread on waveguides which are horns to most people.
My book has more info on waveguides, some of which is only published there.
As far as loading goes, there is a very minimal effect on cone excursion from loading beyond the compression ratio. Air loading, even with a horn, is not that great that it has a pronounced effect on the excursion. Not until one reachs efficiencies of 30% or more does this happen. But at these efficiencies the devices are not very usable because the compression ratios are so high that the amplification of the parasitic effects makes things difficult.
One wants a horn with no reactive part, which is doable at higher frequencies, but never at low frequencies. Cutoff is such an overblown concept - the HP function is gradual in any waveguide, and especially ones that are not of infinite length. The theoretical concept of cutoff comes from an infinite horn in Websters Eq. Infinite horns are impossible and Websters Eq. is incorrect, so the entire concept is obsolete.
Kolbrek said:
The sound of an acoustic instruments can be seen as a base frequency plus alot of harmonics plus some noise, shaped by the resonances of the body. Distortion is mainly harmonics. The only difference is, that the harmonic distortion is generated by the speaker and is not in the original signal. As our ears dont know the original signal, they cant distinguish between the "intended" and "not intended" harmonics. Thus, i doubt that distortion leads to fatique, since that would mean, that listening to an acoustic instrument (and human voice and nearly everything we can hear around us) would also lead to fatique.
As we get into the realm of electric instruments as guitars + amplifiers, most people speak of the "warm" sound of the distortion. Actually, distortion sounds good in an instrument, in the right dosage. There is a huge industry living from producing audio plugins and hardware, that produce distortion of all kind to make records sound sweet.
MaVo said:
What you say is true for a single instument, but when two instuments are played together, the same nonlinearity will cause the two instuments to interact and modulate each other. This has no analog in nature and is highly audible - the real reason why nonlinearities tend to sound bad. The more instuments the worse the sound. A low order of nonlinearity will actually give that "warm" sound because these orders don't tend to be offensive.
gedlee said:
I was recently conversing with PaulW on a very similar topic..
There we had drivers that displayed low non-linear distortion at high output for *most* of the passband.
However, at and below (what presumably was) the driver's resonance - non-linear distortion increased dramatically (from a low of less than .1% to well beyond 5% and rising as freq.s decreased).
Anyway..
My point was that while excursion may be marginally effected with a lower freq. "flair" or "cut-off", (please pardon the terminology), as opposed to one with a higher freq. flair, that with such driver's you are better of with the higher freq. flair to effectively act as an additional highpass filter. Now I'm not certain of this, rather it was just my reasoning at the time. (..and of course we both agreed that a steeper highpass filter above the driver's resonance was most important.) In my "mind" - ONLY if you were to actively boost lower freq.s would their be a significant increase in driver excursion.
Does this seem correct to you?
Additionally, gain represented by a waveguide/horn seems to be little more than a trade-off in directivity. (i.e. decrease directivity with a "supportive" waveguide ='s more gain all else equal.)
From my simplistic reasoning its rather like a pie (..from a horizontal perspective, or a ball from a spherical perspective). (..or if you consider an infinite baffle - half a pie, or half a ball.)
(Keeping with just the horizontal for simplicity..) With half a pie we have our *3*60 degree slices - each "slice" bearing a measure of non-linear distortion with the fundamental.
When we measure at a certain axis (say 0 degree), are we in fact measuring the full amount of non-linear distortion, or is some portion not fully represented because its not on the same axis as the microphone? (..said differently - of the *3* 60 degree "slices", are the 2 side slices, to the sides of the front slice having the 0 degree axis, having their share of non-linear distortion fully measured?)
In a similar but perhaps different respect - perhaps because of the brain's ability to track very minute changes in horizontal radiation, do we hear the full amount of non-liner distortion the driver is producing, or again is some portion lower in apparent level than what should be the total?
In any event what I'm trying to get at here is that IF non-linear levels are somewhat lower than the total THEN decreasing directivity and thereby increasing gain might well increase either/or both measured or heard levels of non-linear distortion. Not the percentage mind you (which should decrease), but the total output.
OK, yeah.. that was a bit nutty.
So am I sniffing to much glue, or is there something to my madness?Hi Earl,
I came across a couple of minor errors myself as I worked through the text, but nothing to the extent that you are indicating. I would be very interested in the specific details of any data errors that you may be aware of - many thanks.
Hornresp can model a conventional oblate spheroidal waveguide - if that is close enough for your needs, then check the Help file for details. If you have any problems running the program, please let me know. I would be very interested in your findings.
Kind regards,
David
gedlee said:
I came across a couple of minor errors myself as I worked through the text, but nothing to the extent that you are indicating. I would be very interested in the specific details of any data errors that you may be aware of - many thanks.
gedlee said:
Hornresp can model a conventional oblate spheroidal waveguide - if that is close enough for your needs, then check the Help file for details. If you have any problems running the program, please let me know. I would be very interested in your findings.
Kind regards,
David
Efficiencies that high is not very hard to achieve in bass horns, even with moderate compression ratios. Efficiency in the order of 70% was achieved in the bass horn of the so-called Fletcher-system developed by the Bell labs, with a compression ratio of 4.5.
This assumes efficiency defined as (Acoustical power output / Electrical power input) *100%.
I would rather say that Websters eq. only gives a limited amount of information. It's an approximation, and should be treated as such. It is not usable for designing horns intended primarily for directivity control.
To make it clear for everyone, cutoff is the frequency where the acoustic throat resistance of an infinite exponential or hypex horn is zero. This point does not exist in a finite horn, where throat resistance will never be zero. The reason is that the radiation impedance at the horn mouth is not zero at any frequency.
But, taking for example an exponential horn, at some point the throat resistance will drop quickly, and this is always above the cutoff frequency of the corresponding infinite horn, higher in frequency the less the difference between throat and mouth areas, especially with undersize mouths.
So strictly speaking, the concept should not be applied to finite horns.
Best regards,
B
The highpass filter effect of the horn does not affect what is coming into the driver. Placing a highpass filter at the output of a device does not protect the device from low frequencies.
Your reasoning would be true if the horn acted as a series capacitor at low frequencies, but it is more like a shunt inductor.
Using a smaller rear chamber is equivalent to using a smaller series capacitor, and will reduce diaphragm displacement.
Best regards,
B
If the maximum power transfer occurs when the source impedance equals the load impedance, then 50% efficiency is the maximum that could ever be achieved. How could one ever exceed this number? The driver itself would have to have NO electrical impedance at all.
I would guess that bass horns have about 10% efficiency max. And even that would be very hard to do. Most LF drivers are about 1% efficient.
Thats my point exactly.
From Keele, "Maximum efficiency of Compression Drivers", AES Convention Paper 6193, October 2004:
"Small-signal calculations show that the maximum nominal efficiency of a horn loudspeaker compression driver is 50% and the maximum true efficiency is 100%. Maximum efficiency occurs at the driver’s resonance frequency. In the absence of driver mechanical losses, the maximum nominal efficiency occurs when the reflected acoustic load resistance equals the driver’s voice-coil resistance and the maximum true efficiency occurs when the reflected acoustic load resistance is much higher that the driver’s voice-coil resistance."
Bell Labs used the "true efficiency" value, not the "nominal efficiency" value. IMHO, true efficiency makes most sense to use, as it relates directly to how much power is converted to sound.
The efficiency of bass horns depend on many things, but by using an efficient driver it can easily reach 30% and more. A quick simulation in Hornresp using a Beyma 102Nd driver gives efficiency in the order of 50-70% efficieny in the pass band, depending on frequency. (4Pi space, 4:1 compression ratio).
Regards,
B
Attachments
Efficiency - in numbers - is not something that I deal with regularly, but your numbers seem highly suspect to me. It would be very very difficult for the acoustic impedance to be greater than the electrical impedance. Your numbers seem to be based on HornResp calculations and I'm not sure where those come from. The only definition of efficiency that I know of is the nominal, which cannot exceed 50% as I said. I don't know what "true" efficiency is.
If what you say were true then compression drivers would not even get warm, but the fact is that they melt down regularly.
If what you say were true then compression drivers would not even get warm, but the fact is that they melt down regularly.
gedlee said:
Horn Response.. Google. Top of the list. Follow the links.
http://www.google.com/search?hl=en&q=horn+response&btnG=Google+Search
gedlee said:
Hi Earl,
Just to clarify, Hornresp calculates loudspeaker efficiency based on the net power available to the loudspeaker, not on the total power generated by the constant voltage source (some of which, as you say, is absorbed by the amplifier output resistance).
Kind regards,
David
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.