This is my current driver's excursion from when I first got them. Reckon this is much more than 4mm?
More Break-in footage - YouTube
More Break-in footage - YouTube
This enclosure has recently seen some resurgence in the German speaking space, since a user at HiFi-Forum.de converted the plan to the metric system, for 15mm MPX wood, and suggested the ULB (55€) chassis the box 15LB075-UW4 for it.
I recently built one, purely out of curiosity:
I never encountered more than rudimentary frequency response measurements in this thread; mostly taken with single sines and SPL meters, with questionable environment variables or no info on measurement setup at all. I believe it is helpful if I share more extensive measurements taken with dedicated hardware and software, calibrated in frequency as well as SPL to 0,5 dB accuracy, taken in true free field / outdoor:
FR and sensitivity, 2V (note: 4 ohm chassis), 1m GPM; overlaid with impedance:
Burst Decay:
More measurements, details on hardware used, listening impressions, video/audio sample etc. in my build thread: JBELL SS15 + the box 15LB075-UW4 build @ HiFi-Forum.de
Use Google Translate if necessary.
A lot of fun for very little money. Props to Jbell for the original plan and lonelybabe69 (yes, that IS his username
) for the plan with cm values and suggestion of the chassis.
I recently built one, purely out of curiosity:
An externally hosted image should be here but it was not working when we last tested it.
I never encountered more than rudimentary frequency response measurements in this thread; mostly taken with single sines and SPL meters, with questionable environment variables or no info on measurement setup at all. I believe it is helpful if I share more extensive measurements taken with dedicated hardware and software, calibrated in frequency as well as SPL to 0,5 dB accuracy, taken in true free field / outdoor:
FR and sensitivity, 2V (note: 4 ohm chassis), 1m GPM; overlaid with impedance:
An externally hosted image should be here but it was not working when we last tested it.
Burst Decay:
An externally hosted image should be here but it was not working when we last tested it.
More measurements, details on hardware used, listening impressions, video/audio sample etc. in my build thread: JBELL SS15 + the box 15LB075-UW4 build @ HiFi-Forum.de
Use Google Translate if necessary.
A lot of fun for very little money. Props to Jbell for the original plan and lonelybabe69 (yes, that IS his username
) for the plan with cm values and suggestion of the chassis.Seems very well behaved up to about 110Hz, then you get into a bit of resonance overlays showing up as smearing in the time domain, very hard if not impossible to get away from in this design principle, and in it's intended range it looks really good, sensitivity wise as well, impressive!
Yup, a TH is basically a series-tuned 6th order BP alignment with pretty poor out-of-band resonances
Interesting though that Fb is down at around 42 Hz it seems. That does suggest that a little EQ can be done near the lower end of the passband to flatten it out to that frequency, without causing too much excursion issues.
Interesting though that Fb is down at around 42 Hz it seems. That does suggest that a little EQ can be done near the lower end of the passband to flatten it out to that frequency, without causing too much excursion issues.
And, BTW, my BOXPLAN-SS workbook will allow you to design a TH with this type of layout, using any panel thickness that you want to use. I've used it for example to create a 32 Hz version. Of course that build did not use just one sheet of ply...
The Subwoofer DIY Page - Horn Folding
The Subwoofer DIY Page - Horn Folding
Proper interpretation. I have little to add. Two points / responses:
1. sensitivity is indeed quite absurd. I've recently measured an Electrovoice Xsub, and at the same 2 volts into 4 ohm it had about the same sensitivity as the SS15 with the 15LB075. So yes, this has as much output at 1 watt as a double 18" bassreflex double the size / volume
2. tuning is 43 Hz determined by the impedance minimum, and 43,5 Hz determined by the 0° value of the electrical phase (a chart I can add if requested). If you view the measured Max. SPL values in the linked thread at HiFi-Forum, you can see maximum output is very good until about down to 40 Hz, after which it drops off very sharply.
So boosting the lower end would be possible, but has to be done very carefully. Also mind possible electrical / thermal difficulties because of the excursion minimum at / around the tuning frequency.
While listening (upper XO 100 Hz LR24 both sides), frankly, anything down to 45 Hz didn't seem underemphasized at all. A track with a heavy bassline at 40 Hz though did sound quite thin already. So if you want 40 Hz instead of 45 Hz, you'll have to put some effort in to get them; 45 Hz upwards it sounds pretty much perfect already as is, without any EQing.
1. sensitivity is indeed quite absurd. I've recently measured an Electrovoice Xsub, and at the same 2 volts into 4 ohm it had about the same sensitivity as the SS15 with the 15LB075. So yes, this has as much output at 1 watt as a double 18" bassreflex double the size / volume
2. tuning is 43 Hz determined by the impedance minimum, and 43,5 Hz determined by the 0° value of the electrical phase (a chart I can add if requested). If you view the measured Max. SPL values in the linked thread at HiFi-Forum, you can see maximum output is very good until about down to 40 Hz, after which it drops off very sharply.
So boosting the lower end would be possible, but has to be done very carefully. Also mind possible electrical / thermal difficulties because of the excursion minimum at / around the tuning frequency.
While listening (upper XO 100 Hz LR24 both sides), frankly, anything down to 45 Hz didn't seem underemphasized at all. A track with a heavy bassline at 40 Hz though did sound quite thin already. So if you want 40 Hz instead of 45 Hz, you'll have to put some effort in to get them; 45 Hz upwards it sounds pretty much perfect already as is, without any EQing.
Last edited:
Nice to see calibrated measurements in a correct environment, thank you, but the link to the German hififorum did not work for me, are there any other diy designs of this type measured in the same way on there?
This shows the strengths of the th principle, and what to look out for, the strengths are a very well defined pass and with well behaved response and high sensitivity and output capabilities in a fairly economical package, to me this is hard to beat for the application.
It also shows why putting focus solely on the spl response in simulations and measurements can be deceiving when it comes to this and other higher order design principles, even with a well behaved spl response above the actual passband seen from fb it may prove tricky to make use of it since you will most probably encounter time domain issues making integration to a topsystem a challenging task.
Still, it is a hard to beat design principle, wery well executed in this case, for it's intended range when economics (number of amp and drivers) becomes more important, which is usually the case for DIY'ers, but even so the performance is very good and can for sure rival or even beat commercial products that are more expensive.
This shows the strengths of the th principle, and what to look out for, the strengths are a very well defined pass and with well behaved response and high sensitivity and output capabilities in a fairly economical package, to me this is hard to beat for the application.
It also shows why putting focus solely on the spl response in simulations and measurements can be deceiving when it comes to this and other higher order design principles, even with a well behaved spl response above the actual passband seen from fb it may prove tricky to make use of it since you will most probably encounter time domain issues making integration to a topsystem a challenging task.
Still, it is a hard to beat design principle, wery well executed in this case, for it's intended range when economics (number of amp and drivers) becomes more important, which is usually the case for DIY'ers, but even so the performance is very good and can for sure rival or even beat commercial products that are more expensive.
Here's the link again - hope it works this time: JBELL SS15 + the box 15LB075-UW4, Erfahrungs- und Bauberichte - HIFI-FORUM
If not, just Google the thread title, "JBELL SS15 + the box 15LB075-UW4", and you shall find it.
I'm afraid I'm a bit of a lone ranger on the measurement front there, especially to the detail and with the accuracy provided. Many users might lack the funds required to purchase high quality hardware, the environment (I am lucky to own an outdoor property where there's no obstacle for hundreds of metres), or the knowledge - which is of course perfectly understandable, especially among hobby DIYers. Those who do might have already gone into professional loudspeaker development and either keep their results to themselves or charge for them.
I haven't measured any other THs. In fact this is the first I've built. I have only ever listened to one other too, the Lambda Labs DH18. Data-Bass.com does have a measurement on the Othorn, and perhaps others.
I too value decay charts highly. You can often see resonances in the frequency chart as well, which would show as narrow dips and elevations. As a rule of thumb, where the frequency response is flat, there's no resonance. But the decay charts show this much more clearly.
Indeed very well executed plan. Not only is efficiency extremely high, max. SPL in the useable frequency band is very solid as well. Reproduction accuracy is very good, almost ideal, within the useable band too.
Absolutely this can and will equal and outperform many commercial products at the same price, and quite a bit above it. Actually I already did one comparison in my linked thread.
I can hardly emphasize enough how much value this design with this speaker (15LB075) offers. ~150€ all in all with wood, paint, terminal etc., for really, really solid performance. With a 2.2 setup like the one in my video you could already throw a really nice "little" party where the very fewest would walk away disappointed due to a lack of bass. Outdoor the sensitivity also really helps - large stacks on small generators very much possible.
I suppose that's why I'm posting here - if I was like "meh" after measuring and listening to the box, I wouldn't have bothered. But what I've measured and heard did leave me quite excited
Btw., I have to mention you and Brian. I instantly recognized you from the first pages of this thread. Somehow I find that way cool that you're the first replies to my post, 10 years and 220 pages later
If not, just Google the thread title, "JBELL SS15 + the box 15LB075-UW4", and you shall find it.
I'm afraid I'm a bit of a lone ranger on the measurement front there, especially to the detail and with the accuracy provided. Many users might lack the funds required to purchase high quality hardware, the environment (I am lucky to own an outdoor property where there's no obstacle for hundreds of metres), or the knowledge - which is of course perfectly understandable, especially among hobby DIYers. Those who do might have already gone into professional loudspeaker development and either keep their results to themselves or charge for them.
I haven't measured any other THs. In fact this is the first I've built. I have only ever listened to one other too, the Lambda Labs DH18. Data-Bass.com does have a measurement on the Othorn, and perhaps others.
I too value decay charts highly. You can often see resonances in the frequency chart as well, which would show as narrow dips and elevations. As a rule of thumb, where the frequency response is flat, there's no resonance. But the decay charts show this much more clearly.
Indeed very well executed plan. Not only is efficiency extremely high, max. SPL in the useable frequency band is very solid as well. Reproduction accuracy is very good, almost ideal, within the useable band too.
Absolutely this can and will equal and outperform many commercial products at the same price, and quite a bit above it. Actually I already did one comparison in my linked thread.
I can hardly emphasize enough how much value this design with this speaker (15LB075) offers. ~150€ all in all with wood, paint, terminal etc., for really, really solid performance. With a 2.2 setup like the one in my video you could already throw a really nice "little" party where the very fewest would walk away disappointed due to a lack of bass. Outdoor the sensitivity also really helps - large stacks on small generators very much possible.
I suppose that's why I'm posting here - if I was like "meh" after measuring and listening to the box, I wouldn't have bothered. But what I've measured and heard did leave me quite excited
Btw., I have to mention you and Brian. I instantly recognized you from the first pages of this thread. Somehow I find that way cool that you're the first replies to my post, 10 years and 220 pages later
We diy designers might never get any adoption in to the truly professional realm but this I strongly believe are for reasons other than the performance aspects of our designs, this shows exactly that.
For the same reasons we can remain free to explore whatever our hearts...no sorry, that should be stone cold engineering minds desire all without having an entire organisation setting financial restrictions or marketing boundaries for us, which in turn leads to a very diverse and interesting development environment without excluding the engineering competence.
I'm convinced that if there ever was a comparrison shootout the results would potentially surprise alot of the established professionals, but in this industry the product is but a small yet important part of a huge production machinery.
For the same reasons we can remain free to explore whatever our hearts...no sorry, that should be stone cold engineering minds desire
all without having an entire organisation setting financial restrictions or marketing boundaries for us, which in turn leads to a very diverse and interesting development environment without excluding the engineering competence.I'm convinced that if there ever was a comparrison shootout the results would potentially surprise alot of the established professionals, but in this industry the product is but a small yet important part of a huge production machinery.
1. sensitivity is indeed quite absurd. I've recently measured an Electrovoice Xsub, and at the same 2 volts into 4 ohm it had about the same sensitivity as the SS15 with the 15LB075. So yes, this has as much output at 1 watt as a double 18" bassreflex double the size / volume
...boosting the lower end would be possible, but has to be done very carefully.
Stoneeh,
As far as boosting the low end of the SS15 using a light cone like the 15LB075-UW4, some may find its distortion to be unpleasant even at a level well below Xmax.
Your SS15 response curve in post #2227 indicates 105dB 1w1m sensitivity at 100 Hz, dropping to about 98 dB at 40Hz.
The Xsub uses a pair of EVX180B. EV’s EVX180B spec sheet indicates 100-800Hz average sensitivity at 98dB, this includes it’s +5dB peak between 500-800Hz. At 100Hz and below, sensitivity for a single EVX180B is below 95dB.
Considering the Xsub’s relatively small (for a dual 18”) net volume, and Fb around 40Hz, would be surprised if it’s sensitivity at any frequency below 100Hz exceeds 98dB, which would indicate your sensitivity measurements may be high, considering you wrote the Xsub has about the same sensitivity as your SS15.
Could you also post the Xsub measurement for comparison?
Thanks,
Art
Attachments
The SS15 may roll off below 40, but if compared to an equivalent 15" reflex that would normally be tuned for a bit of an efficiency boosting bass hump with the reflex port, rather than an flat response. You will find that the TH can use a lower high pass filter.
Once the highpass is addded I find that the TH is louder over the entire pass band.
Adding bass boost to the TH might work at lower levels but you will soon run out of excursion. I find it preferable to run it flat and with minimal peak compression.
With a small set of 15 reflex cabinets trying to keep up with a pair of mid tops you soon run into bass compression at rock & roll levels.
The SS15 is so efficient that it can keep up with the mid tops with minimal bass processing preserving the dynamic range - so you can have your bass line and your kick riding on top. You have the added bonus of running the driver at lower average power levels so the dynamic peaks can be even bigger with less thermal compression. The TH also cools the driver better so the overall effect is much better dynamic range.
Well worh the extra cabinet volume in my opinion!
Once the highpass is addded I find that the TH is louder over the entire pass band.
Adding bass boost to the TH might work at lower levels but you will soon run out of excursion. I find it preferable to run it flat and with minimal peak compression.
With a small set of 15 reflex cabinets trying to keep up with a pair of mid tops you soon run into bass compression at rock & roll levels.
The SS15 is so efficient that it can keep up with the mid tops with minimal bass processing preserving the dynamic range - so you can have your bass line and your kick riding on top. You have the added bonus of running the driver at lower average power levels so the dynamic peaks can be even bigger with less thermal compression. The TH also cools the driver better so the overall effect is much better dynamic range.
Well worh the extra cabinet volume in my opinion!
Nice to see a measurement.
To do a ground plane measurement that will be the equal of a 2Pi measurement you need to place your mic 2 metres away. Not sure if that is the case in the posted measurement. Otherwise it will have 3db greater efficiency than in real life.
Nice enclosure for the money. Pretty much unbeatable.
To do a ground plane measurement that will be the equal of a 2Pi measurement you need to place your mic 2 metres away. Not sure if that is the case in the posted measurement. Otherwise it will have 3db greater efficiency than in real life.
Nice enclosure for the money. Pretty much unbeatable.
Ground plane IS half space / 2pi. And of course this does represent "real life", since subwoofers are in the vast majority of cases ground stacked.
You might be referencing full space / 4pi, which would represent flown subwoofers. This is done only in large venues and large commercial systems and is very unlikely to ever happen to a SS15
Jfyi, a point source with omnidirectional directivity, which a single subwoofer at the bottom end of its frequency range is, loses 6 dB SPL (sound pressure level) per doubling of space - not 3 dB.
4pi (flown) = 0 dB gain
2pi (ground without wall, or flown with wall) = 6 dB gain
1pi (ground & wall) = 12 dB gain
0,5 pi (corner) = 18 dB gain
You might be confusing sound power level with sound pressure level. The gain from halfing of space in sound power level would indeed be 3 dB.
Sound power is the cause of sound pressure, and an intangible quantity that cannot be measured.
What a microphone measures is sound pressure level. Acoustic measurements done in half space yield 6 dB higher values than done in full space (again, only valid for an omnidirectional source).
You might be referencing full space / 4pi, which would represent flown subwoofers. This is done only in large venues and large commercial systems and is very unlikely to ever happen to a SS15
Jfyi, a point source with omnidirectional directivity, which a single subwoofer at the bottom end of its frequency range is, loses 6 dB SPL (sound pressure level) per doubling of space - not 3 dB.
4pi (flown) = 0 dB gain
2pi (ground without wall, or flown with wall) = 6 dB gain
1pi (ground & wall) = 12 dB gain
0,5 pi (corner) = 18 dB gain
You might be confusing sound power level with sound pressure level. The gain from halfing of space in sound power level would indeed be 3 dB.
Sound power is the cause of sound pressure, and an intangible quantity that cannot be measured.
What a microphone measures is sound pressure level. Acoustic measurements done in half space yield 6 dB higher values than done in full space (again, only valid for an omnidirectional source).
weltersys: there's some issues with your assessment of the EV sub.
1. the sensitivity figure of 98 dB SPL 1w / 1m from 100 to 800 Hz provided in the EVX 180B datasheet you referenced is given without data on space or enclosure.
2. the closest enclosure to the Xsub in the EVX 180B datasheet (though somewhat larger net volume and lower tuning) is figure 2D, which we might assume (not specified) is the 1w / 1m / 2pi value.
The SPL values given in that chart are abysmally low and I know of no double 18 BR that would yield such a poor sensitivity - even a L'Acoustics KS28 with a similar tuning (31 Hz), but much smaller volume and inefficient drivers with large moving masses does on average 98 dB 1w / 1m / 2pi from 30 to 100 Hz.
So basically figure 2D cannot be the 1w / 1m / 2pi response for a double 18 with relatively efficient chassis in such a large enclosure.
3. you ignored the sensitivity value EV itself cited for the Xsub in the Xsub datasheet, which is 103 dB 1w / 1m / 2pi, down to a -3 dB point of 40 Hz - meaning 40 Hz = 100 dB. This is actually exactly what I measured. It however does stand in stark contrast to the values provided in the datasheet of the chassis, which surely does not help to clarify things.
I hope this helps to clarify things. If I just posted my measurement without commentary, it wouldn't be of benefit if the underlying principles are not understood. I will gladly post it when I feel it is appropriate.
Ad boosting vs. distortion: the assessment / recommendation was based on measured distortion figures, as well as the impedance plot.
From the impedance plot we can see that loading of the resonator extends to about 40 Hz. "Loading" means the chassis exciting the resonator, which reduces membrane excursion, which leads to lower distortion.
The distortion plot (see my linked report) confirms this - down to about 40 Hz, max. SPL is high with little distortion.
However, frequency response / sensitivity drops off somewhere above that, and audibly there's not much there around 40 Hz.
So as already stated, there's a narrow window between about 40 and 45 Hz where the response can be equalized, without restraints on max. SPL.
One has to be careful, because a) boosting means putting more power into the chassis, which could run it into its electrical / thermal limit b) below 40 Hz, resonator loading drops off quite sharply.
So equalizing this sub is not a trivial undertaking, and should only be done if you know exactly what you're doing - preferrably using measurement equipment. If you're not sure, just leave it alone - a lot of possible harm for little gain.
I actually experimented with that equalization just yesterday. I used a narrow PEQ with about 2 dB peak gain on 40 Hz. The results were promising - 40 Hz basslines didn't sound as shallow anymore as on the unfiltered box. But it would require a LOT of testing for me to give a clear specific recommendation.
Also mind my measurements are for the the box 15LB075-UW4. Other chassis might and will respond differently.
1. the sensitivity figure of 98 dB SPL 1w / 1m from 100 to 800 Hz provided in the EVX 180B datasheet you referenced is given without data on space or enclosure.
2. the closest enclosure to the Xsub in the EVX 180B datasheet (though somewhat larger net volume and lower tuning) is figure 2D, which we might assume (not specified) is the 1w / 1m / 2pi value.
The SPL values given in that chart are abysmally low and I know of no double 18 BR that would yield such a poor sensitivity - even a L'Acoustics KS28 with a similar tuning (31 Hz), but much smaller volume and inefficient drivers with large moving masses does on average 98 dB 1w / 1m / 2pi from 30 to 100 Hz.
So basically figure 2D cannot be the 1w / 1m / 2pi response for a double 18 with relatively efficient chassis in such a large enclosure.
3. you ignored the sensitivity value EV itself cited for the Xsub in the Xsub datasheet, which is 103 dB 1w / 1m / 2pi, down to a -3 dB point of 40 Hz - meaning 40 Hz = 100 dB. This is actually exactly what I measured. It however does stand in stark contrast to the values provided in the datasheet of the chassis, which surely does not help to clarify things.
I hope this helps to clarify things. If I just posted my measurement without commentary, it wouldn't be of benefit if the underlying principles are not understood. I will gladly post it when I feel it is appropriate.
Ad boosting vs. distortion: the assessment / recommendation was based on measured distortion figures, as well as the impedance plot.
From the impedance plot we can see that loading of the resonator extends to about 40 Hz. "Loading" means the chassis exciting the resonator, which reduces membrane excursion, which leads to lower distortion.
The distortion plot (see my linked report) confirms this - down to about 40 Hz, max. SPL is high with little distortion.
However, frequency response / sensitivity drops off somewhere above that, and audibly there's not much there around 40 Hz.
So as already stated, there's a narrow window between about 40 and 45 Hz where the response can be equalized, without restraints on max. SPL.
One has to be careful, because a) boosting means putting more power into the chassis, which could run it into its electrical / thermal limit b) below 40 Hz, resonator loading drops off quite sharply.
So equalizing this sub is not a trivial undertaking, and should only be done if you know exactly what you're doing - preferrably using measurement equipment. If you're not sure, just leave it alone - a lot of possible harm for little gain.
I actually experimented with that equalization just yesterday. I used a narrow PEQ with about 2 dB peak gain on 40 Hz. The results were promising - 40 Hz basslines didn't sound as shallow anymore as on the unfiltered box. But it would require a LOT of testing for me to give a clear specific recommendation.
Also mind my measurements are for the the box 15LB075-UW4. Other chassis might and will respond differently.
The sensitivity value for the Xsub which was posted with the rest of the EVX 180B information is inconsistent, stating 103dB half space, 100dB full space. One of those specifications must be 3dB off, as we know there should be a 6dB difference.
Including response to 400 Hz on the Xsub spec would indicate that the +3dB response in that frequency range shown in the EVX180B specifications would make most of the output below 100 Hz 100dB half space, -3dB puts 40Hz at 97dB, consistent with the dual EVX180B box sensitivity of around 98dB in larger boxes.
Looking at other dual 18” specification sheets:
The EV TX2181 (Fb around 35Hz) 40Hz 96dB half space /2pi.
The EAW SB1001 (Fb around 35Hz) 40Hz 96dB 2pi.
The Fulcrum Acoustic Sub218L (Fb around 32Hz) 98dB 2/pi.
The JBL SRX728S (Fb around 35Hz) 96dB half space.
As Mark suggests, your measurements appear to be 3dB high, possibly due to measuring at only 1 meter, at 2 meters the test mic would be out of the “mouth bubble”.
By the way, used to use a lot of EVX woofers around 1990 when "nothing came close" ;^)
Art
Attachments
Last edited: