I get that. But there is huge price to pay, and it is not working seamlessly either. You dive into very deep power compression, danger of damaging the speaker or 10dB limiter kicking in, by which the excursion "leeway for EQ" is not very usable. Average will be milder, but around port tuning, we are talking 6+dB. That is very ugly. Also, if the design is not undertuned, and is aligned well, and frequency response is rather flat, why would you want to play woth EQ, adding unnecessary peak in frequency response?
At port tuning frequency, the driver is already driven at 150%, and only reason for its survival is the program content that is dynamic enough to not torture the speaker for too long in this spot to die. With synthetic music, it might not be the case, the speaker is tortured, and adding EQ will only result in overall lower gain to be had to not burn the speaker. And that is exactly the point by which this design is not winning so easily over mine. The conventional design is power limited, my design is excursion limited. Just a choice. Nothing is objectively better, I bet on my design because it is safer to watch long term overexcursion than smoking the speaker. Premature spending of the suspension as a consumable is cheaper than burning it down. With my use though, I am still in 6-10 years ballpark.
There is nothing inherrently wrong with both. There is no mistake, just chosen set of compromises, which is now being overturned by RCF and their 18mmXmax drivers, or some older beasts that can do that.
As shown, my design shows little compression. 100Watts at Xmax, little on the port, little around the impedance peak. Overall lower compression than with conventional design.
If I chose the driver with 18mm Xmax, perfectly able to do 21-22, and I need 15-16, it would be perfect fit with zero compromise whatsoever. LN19S400 and LN19S450 drivers can do exactly that even in long term use, which I am not really doing.
At port tuning frequency, the driver is already driven at 150%, and only reason for its survival is the program content that is dynamic enough to not torture the speaker for too long in this spot to die. With synthetic music, it might not be the case, the speaker is tortured, and adding EQ will only result in overall lower gain to be had to not burn the speaker. And that is exactly the point by which this design is not winning so easily over mine. The conventional design is power limited, my design is excursion limited. Just a choice. Nothing is objectively better, I bet on my design because it is safer to watch long term overexcursion than smoking the speaker. Premature spending of the suspension as a consumable is cheaper than burning it down. With my use though, I am still in 6-10 years ballpark.
There is nothing inherrently wrong with both. There is no mistake, just chosen set of compromises, which is now being overturned by RCF and their 18mmXmax drivers, or some older beasts that can do that.
As shown, my design shows little compression. 100Watts at Xmax, little on the port, little around the impedance peak. Overall lower compression than with conventional design.
If I chose the driver with 18mm Xmax, perfectly able to do 21-22, and I need 15-16, it would be perfect fit with zero compromise whatsoever. LN19S400 and LN19S450 drivers can do exactly that even in long term use, which I am not really doing.
Here depicted are issues of comparable setups simulated for the same output.
Top conventional, tuned at 39Hz, bottom 30Hz tune. Same box and port volume. Post thermal compression, post EQ, port compression mostly ignored, I am not sure what Hornresp does with it.
Additional EQ for resonating port of the low tuned design will need to be applied, but this thing can be banged on nearly without any limiter on such voltage.
The conventional design does not suffer from overexcursion, but banged between 35 and 45Hz full tilt will burn down within minutes. Needs 3-4dB limiter realistically.
The SPL outcomes are obvious. 0,6dB backpedalling, or 3-4db packpedalling. Choice for synthetic bassy music is obvious to me.
I still do not believe my design is exactly better. I believe it is equal, workable, and safer for me.
We know how the speaker behaves on 0,6dB overexcursion, and we know how it behaves with 3+dB overpowering. The choice is easy.
Another untold benefit of my design is, that with such cutoff, the amplifier will see something along the lines of 6Ohm load for two drivers in parallel. So two pieces can be used on the amp/channel compared to the conventional one. Heh. I have not tried this one though. That might be too cheap. The LF21N551s were running with amp for each. This would be the pinnacle of "kneedrilling for a penny" approach. 🙂
Top conventional, tuned at 39Hz, bottom 30Hz tune. Same box and port volume. Post thermal compression, post EQ, port compression mostly ignored, I am not sure what Hornresp does with it.
Additional EQ for resonating port of the low tuned design will need to be applied, but this thing can be banged on nearly without any limiter on such voltage.
The conventional design does not suffer from overexcursion, but banged between 35 and 45Hz full tilt will burn down within minutes. Needs 3-4dB limiter realistically.
The SPL outcomes are obvious. 0,6dB backpedalling, or 3-4db packpedalling. Choice for synthetic bassy music is obvious to me.
I still do not believe my design is exactly better. I believe it is equal, workable, and safer for me.
We know how the speaker behaves on 0,6dB overexcursion, and we know how it behaves with 3+dB overpowering. The choice is easy.
Another untold benefit of my design is, that with such cutoff, the amplifier will see something along the lines of 6Ohm load for two drivers in parallel. So two pieces can be used on the amp/channel compared to the conventional one. Heh. I have not tried this one though. That might be too cheap. The LF21N551s were running with amp for each. This would be the pinnacle of "kneedrilling for a penny" approach. 🙂
Last edited:
Looks like that's a situation where you can afford to use twice the excursion (above the port tuning) to allow certain benefits.
I don't think that will apply in every situation (some designs will be excursion-limited), but it's interesting to see. Thanks for posting a comprehensive set of graphs.
Chris
I don't think that will apply in every situation (some designs will be excursion-limited), but it's interesting to see. Thanks for posting a comprehensive set of graphs.
Chris
For many genres of music, that are not abnormal, which have 9dB+ dynamic range, this is possible indeed. The more conservative/conventional solution would withstand +3dB. The port would choke though in this case.
For heavy synthetic music though, it would either play less or it would be one minute show.
Application matters.
For heavy synthetic music though, it would either play less or it would be one minute show.
Application matters.
Mind you, if you use 19" with Xmax of 18-20mm, which LN19S450 will be comfortable to do, then you can up the input too. With 1700Watts RMS it will be heavily power limited again. 🙂
More measurements done with SECOND bottle of Port. My yearly alco consumption finished in two days. Maan, audio makes you drink.
At 40Hz, the SPL compression drastically reduced. Now the difference between 15mm excursion from 11mm (3dB) is just a fraction. No serious compression to speak of, at 320VA free air. Maaan I can't wait to make some sawdust.
At 40Hz, the SPL compression drastically reduced. Now the difference between 15mm excursion from 11mm (3dB) is just a fraction. No serious compression to speak of, at 320VA free air. Maaan I can't wait to make some sawdust.
I expect most software uses linear math models. So the results will only match the real world when the woofer is driven in its linear displacement region, below Xmax. I would expect the proof will be in the distortion measurements of the finished speaker. Xmax is often referred to as the excursion that produces 10% distortion. Unfortunately in my experience the people running the sound boards at local live music events drive the speakers to just below destruction and ignore any distortion. At a Mark Knopfler concert I attended a few years ago the bass was driven so hard that kick drum beats and bass guitar notes produced nearly identical burping sounds. Unfortunately many of these venues are more interested in shaking your guts than producing listenable music.
Last edited:
Know that sound. Unexplainably surreal. Witnessed it this year too. Just mix of resonances and noises, very specific, but very disturbing. Though further from the bins, quite acceptable bass was hearable. That must be a lot of masking.
Yes, that´s the issue too. You give em 1kW, they play it full tilt. You give em 10kW, they play it full tilt. You give em 50kW, they kill few people.
Thankfully, I am not doing regular gigs. It´s a personal hobby, and I rather attend the situations where someone else plays on my gear. Limiter ON, myself with an axe prepared to either to cut the wires or chop off the DJ´s hands off. Those atrocities will not happen around me.
My reality is 2% use most of the time, some normal gigs, and some "rave finish" at 101% last hour or so, couple times a year at most. More likely once or twice.
Two 21"s were enough. Heck even two 18NW100s were, where a village "militia" came at me for noise, even though the gig was fully approved.
Four 18"s are just for safety, overkill and good feel. Now, I can´t imagine going at or past 17mm one way with LF18X451s, because at that point more than half magnetic gap would be unoccupied by the coil, and that´s horsecrap. The 18,5mm test was awesome, it showed that the driver is absolutely safe at Xmax, and that was about it. Would the 18-20mm beast of LN19S450 be better? sure, but at that point it would need bigger box, bigger port, bigger amp. No free lunch.
More overload conditions to be tested though. I have a very interesting driver in my paws. It is a nice looking 12" chinese ~$50 subwoofer, with very soft suspension, with Xmax of around 13-14mm, that can be "railgunned" to the point of bottoming out at about 25-30mm one way, where the coil is completely off the top plate gap. Otherwise it looks decent, plays decent. So extreme and failure modes will be tested too. This is exciting knowledge, and this time not that expensive.
Yes, that´s the issue too. You give em 1kW, they play it full tilt. You give em 10kW, they play it full tilt. You give em 50kW, they kill few people.
Thankfully, I am not doing regular gigs. It´s a personal hobby, and I rather attend the situations where someone else plays on my gear. Limiter ON, myself with an axe prepared to either to cut the wires or chop off the DJ´s hands off. Those atrocities will not happen around me.
My reality is 2% use most of the time, some normal gigs, and some "rave finish" at 101% last hour or so, couple times a year at most. More likely once or twice.
Two 21"s were enough. Heck even two 18NW100s were, where a village "militia" came at me for noise, even though the gig was fully approved.
Four 18"s are just for safety, overkill and good feel. Now, I can´t imagine going at or past 17mm one way with LF18X451s, because at that point more than half magnetic gap would be unoccupied by the coil, and that´s horsecrap. The 18,5mm test was awesome, it showed that the driver is absolutely safe at Xmax, and that was about it. Would the 18-20mm beast of LN19S450 be better? sure, but at that point it would need bigger box, bigger port, bigger amp. No free lunch.
More overload conditions to be tested though. I have a very interesting driver in my paws. It is a nice looking 12" chinese ~$50 subwoofer, with very soft suspension, with Xmax of around 13-14mm, that can be "railgunned" to the point of bottoming out at about 25-30mm one way, where the coil is completely off the top plate gap. Otherwise it looks decent, plays decent. So extreme and failure modes will be tested too. This is exciting knowledge, and this time not that expensive.
If I had spare money I would just buy 20 of them "just in case". This is Europe, Czech Republic. These are About 300WRMS drivers.
Fs 27Hz, Qts 0.5.
Fs 27Hz, Qts 0.5.
Klippel can actually also do some long term testing as well.
So we get a graph with how a specific drivers behaves as function of time, power and thermals.
Also they can show the stability of the motor as well.
This is very rarely shown on large signal measurements you see on the web (eg Voice Coil's testbench).
Recently I had some drivers Klippel tested and one of them seems pretty decent at first sight, but long term behavior was really bad.