It was inevitable. Welcome to the 12" subwoofers. Awesome designs can be made, with great prformance. but unless in big FLH, it wont play bass in todays standards. Been there. Up the tuning a notch, lower the expectations and nonsensical sims, you'll be good. ðŸ¤
I don't have a calibrated SPL meter, so couldn't calibrate my measurements. I have a cheap one that offers A or C weighting, but it was £30 around 10 years ago, so I have little faith in its accuracy. Perhaps I should buy one of those SPL calibrators (94dB in a box) - it'd be easier than using a seperate SPL meter.
The bad noises happened when the PRs were hitting their maximum excursion, so I suspect I was hitting Xmech there. This was in the final 10V of output from the amp, with sine tones - right at the ragged edge. Thankfully, nothing was damaged despite multiple high-power sweeps. A few of them pushed the amp a couple of dB into clipping.
Crash, I haven't written this project off just yet. I'll definitely be increasing the tuning frequency. At the moment, it's around 37Hz, which is definitely on the low side. From the limited testing I've done, I'm pretty pleased so far.
Next steps:
Figure out a way of calibrating the SPL of my measurement setup
Compare with the sealed 18" and ported 12". If it matches (vs the 18"s at 40Hz) or beats (the ported 12") both, I'll ditch the rest of my subs and build more of these.
Try to take some waveforms
Chris
The bad noises happened when the PRs were hitting their maximum excursion, so I suspect I was hitting Xmech there. This was in the final 10V of output from the amp, with sine tones - right at the ragged edge. Thankfully, nothing was damaged despite multiple high-power sweeps. A few of them pushed the amp a couple of dB into clipping.
Crash, I haven't written this project off just yet. I'll definitely be increasing the tuning frequency. At the moment, it's around 37Hz, which is definitely on the low side. From the limited testing I've done, I'm pretty pleased so far.
Next steps:
Figure out a way of calibrating the SPL of my measurement setup
Compare with the sealed 18" and ported 12". If it matches (vs the 18"s at 40Hz) or beats (the ported 12") both, I'll ditch the rest of my subs and build more of these.
Try to take some waveforms
Chris
You could try a strong neodymium magnet to remove the stuck disc, or maybe even a vacuum cleaner.
28L with that level of output is quite the feat.
Quick update: managed to remove the last disks. Poked around until they would rattle slightly, and then knocked the basket/back of the PR against the floor a couple of times.
They've essentially filled the dustcap with some rubbery stuff, which must have been slightly tacky when they installed the washers. Seems fine now, though.
According to the impedance curves, the new highest tuning frequency is 48Hz. I put one more disk back on, and got 42Hz. Douglas Adams would be pleased.
The REW file from the previous round of outdoor testing is attached.
Notes:
Upon reflection, I think part of the stronger-than-expected LF slope may be because I was measuring at approx 1m from the front of the cabinet, but the PRs were approximately 1.3m away from the mic.
Instead of borrowing it, could I use it to measure a speaker, and then switch to my Motu interface + Beyer mic and measure again? - That'd give me Dayton-reported SPL, Motu-reported SPL, and I can adjust for the difference. Shouldn't take long.
Next steps:
A longer-term conundrum: making this thing "tour-ready". Right now, there are 3x 12" cones, each with considerable excursion capability, open to the elements. It'd be good to find a way to reduce the likelihood of damage without increasing the cabinet volume too much. Oh, and force cancellation for the PRs is of course mandatory. I like how Bose approached it with their baby PC speakers: https://www.cnet.com/pictures/bose-computer-musicmonitor/4/
Which is 2x PRs in a slot-loaded configuration. The cabinet would need to get taller to allow airflow to the rear PR, but then depth could be reduced considerably
The question is whether the 2x PRs slot-loaded would take up more or less cabinet volume than simply adding spacers and wrapping a grille around the three sides.
Onwards!
Chris
They've essentially filled the dustcap with some rubbery stuff, which must have been slightly tacky when they installed the washers. Seems fine now, though.
According to the impedance curves, the new highest tuning frequency is 48Hz. I put one more disk back on, and got 42Hz. Douglas Adams would be pleased.
The REW file from the previous round of outdoor testing is attached.
Notes:
- SPL is not calibrated. I had my Motu M4 interface set for minimum gain on the mic input.
- First measurement is the impedance curve for the tuning used in the frequency response curves.
- There seems to be a small discrepancy between impedance minima and what's measured nearfield, so that might need investigation.
- Last couple of measurements are the 48Hz and 42Hz impedance curves mentioned above.
- The measurements were taken outdoors, around 5m away from the nearest building.
- The voltages are the RMS voltage measured by the Powersoft T602, reported by Armonia+.
- I applied 24dB/oct Butterworth filters at (IIRC) 33Hz and 150Hz. There's a broad-band measurement included.
- I did try pushing 2dB past the 101Vrms measurement, but the amplifier showed clipping and the frequency response showed broad-band compression.
- To keep within allowed file sizes, I've just shown the last 3 measurements at -8dB, -2dB and 0dB relative to the amplifier's maximum output level. The last couple made the cabinet sound pretty unhappy.
- The cabinet was on grass, which has been pretty wet recently.
Upon reflection, I think part of the stronger-than-expected LF slope may be because I was measuring at approx 1m from the front of the cabinet, but the PRs were approximately 1.3m away from the mic.
Appreciate the offer!
Instead of borrowing it, could I use it to measure a speaker, and then switch to my Motu interface + Beyer mic and measure again? - That'd give me Dayton-reported SPL, Motu-reported SPL, and I can adjust for the difference. Shouldn't take long.
Next steps:
- Calibrate SPLs
- Check if the frequency response is any better on solid ground
- Figure out whether the close-up frequency response, or impedance curve, is a better indicator of tuning frequency
- Figure out if the excursion limiter is required to keep the PRs safe
- Re-measure at high-power, compare to the sealed 18"s
- Test with music at high levels, subjective comparisons with other cabinets
- Waveforms
A longer-term conundrum: making this thing "tour-ready". Right now, there are 3x 12" cones, each with considerable excursion capability, open to the elements. It'd be good to find a way to reduce the likelihood of damage without increasing the cabinet volume too much. Oh, and force cancellation for the PRs is of course mandatory. I like how Bose approached it with their baby PC speakers: https://www.cnet.com/pictures/bose-computer-musicmonitor/4/
Which is 2x PRs in a slot-loaded configuration. The cabinet would need to get taller to allow airflow to the rear PR, but then depth could be reduced considerably
The question is whether the 2x PRs slot-loaded would take up more or less cabinet volume than simply adding spacers and wrapping a grille around the three sides.
Onwards!
Chris
Attachments
How about using really strong motors and a huge amount of eq to get the frequency response you like? There’s no shortage of power and dual drivers means more metal to heat.
Two opposing active speakers and two passive in a complete force cancellation setup. https://www.precision-devices.com/w...D-123C001-Technical-Specifications-070221.pdf
Two opposing active speakers and two passive in a complete force cancellation setup. https://www.precision-devices.com/w...D-123C001-Technical-Specifications-070221.pdf
The side firing PRs may be a problem in deployments where it would be advantageous to put the subs next to a wall, and won't allow tight pack arrays to be ratchet strapped together.
The Earthquake SLAP Pump-12 Module funnels the output of a pair of PRs with ~ 4" (100mm) peak to peak excursion, twice that of of yours, though a small slot.
Slot loading your PRs would only require a single front grill, and all the output would exit the front.
This would also facilitate cardiod arrays.
And your one meter tests would be more accurate 😉
Handles could be side mounted in the PR slots.
The diagram shows the driver front mounted.
A single back access panel could allow rear mounting both PRs and the driver.
Rear mounting the driver would reduce the volume given up to the grill frame.
Art
As you raise the drive level and the suspensions temporarily soften with excursion, you may find that the measured impedance will shift somewhat from it's cold state. Tuning may drop 1-2 hz.
Powersoft is above my hobby level pay grade, does their DSP software accomodate level based EQ? If so you can boost quite a bit at and above tuning at low levels to make the sub sound 'full' and taper off the boost at higher input levels. I've used the Behringer NX software to do this in some capacity.
Art's suggestion make sense. If needed, make the cabinet a bit taller, if making it slightly wider on the front, less deep and rear mounting doesn't give you the same amount of volume, it should still be a quite easy one person carry, and you can link multiple boxes together.
I'd recommend a pair of speakons in parallel on each box for easy daisy chaining if you go ahead with these, 3 boxes in parallel on one channel would be within the amp's capabilities.
Powersoft is above my hobby level pay grade, does their DSP software accomodate level based EQ? If so you can boost quite a bit at and above tuning at low levels to make the sub sound 'full' and taper off the boost at higher input levels. I've used the Behringer NX software to do this in some capacity.
Art's suggestion make sense. If needed, make the cabinet a bit taller, if making it slightly wider on the front, less deep and rear mounting doesn't give you the same amount of volume, it should still be a quite easy one person carry, and you can link multiple boxes together.
I'd recommend a pair of speakons in parallel on each box for easy daisy chaining if you go ahead with these, 3 boxes in parallel on one channel would be within the amp's capabilities.
The Faital 12RS1066 has similar BL, more excursion and more power handling than the PD driver you've linked.
FWIW, you generally need at least 2x the PR displacement compared to the active driver(s), so it'd end up being 2x main drivers and 4x PRs. Getting towards a big box!
Art, thanks!
I seriously looked at buying one of those Earthquake Pump modules, but decided against it: I couldn't find any documentation about cabinet size vs tuning frequency, nor did there seem to be a way of changing the tuning frequency. ie, I'd just have to buy one, put it in a box and see what I get.
I expect I'll mostly deploy these subs in 1/2/3-per-side situations, where I'll just stack them vertically. Small-format stuff, compared to your systems. I can't really think of a situation where the sideways PR output would be a huge problem, but intuitively I think having all the output from the front of the cabinet is a good thing.
Slot-loading the PRs is an option I'm considering. There are a few iterations that would work, including DIYing one of those Earthquake modules, so I'll do some sketches and figure out what's going to be the most compact overall.
An option that I quite like is similar to the ROAR layout, but with PRs instead of the folded path flanking the main driver. Benefits: extra output above 60Hz, main driver's magnet in fresh air (better than PRs acting as heatsinks), output is all at the front of the cabinet. The main drawback is the increased cabinet volume.
I'll do some sketches and work out the external cabinet volume of all the options I can think of. Then, of course, compare to a ported box. Who knows, that might win after all.
Chris
diyuser, interesting that the tuning might shift in-use. Do you have a source for that info?
It'd be easy enough to run an impedance sweep before/after a set of high-power sweeps and look for differences.
Yep, the Powersoft T-series (and X-series) amps have a single band of dynamic EQ available. I'm planning on using that to save the PRs from over-excursion just below the tuning frequency. A 24dB/oct highpass isn't quite enough to save them. I'm hoping to avoid using lots of EQ to get these subs to a usable state. Been there and done that with the previous sealed 18"s.
I'll be building 4-6 of these, and at least two of them will have 3x SpeakOns:
- 1x 4P input (sub connected to 1+/-)
- 1x 4P output (just wired 1:1)
- 1x 2P output, linked to input 2+/-.
That would allow me to easily split out to run a main speaker above, by using a single 4-core cable to power the stack.
Chris
Maybe some klippel/LSI measurements. But it is obvious, as cms/kms changes with displacement. Therefore having stiffer spring with the same mms MUST move with the Fs. The question is what is the real effect on SPL. I think it will not be big. It should half way self compensate for speaker getting more power in the bottom end as the impedance peak moves up. It should not move much, as the port here dictates a lot of the resonance modes. So it should be "different but same" all in all, regarding SPL.
http://www.quarter-wave.com/General/Two_Drivers.pdf
I remembered that wrong. With two drivers in series we do get double BL, double Sd but.. since we also double the Re so in the end the benefits simply isn’t as many as I first was thinking.
But force cancellation is nice. Heat managed by two coils is nice.
And the passive membranes rule of thumb, isn’t that related to the amount of air the active driver can move rather than membrane area. My suggestion have as you said lower excursion capacity.
Chris, the SLAPS-M12's use weights to tune, I assume they are the same units on the pump module. Specs are here:
Assuming Surround Losses of Qmd = 7 @ 20Hz
Have a look at the behavior of the TC Sounds VMP passives at higher drive levels in the Data-Bass testing, in particular look at the long term output compression magnitude graph.
Specifications
- Overall Diameter: 12.37″ 314.2 mm
- Cutout Diameter: 11.37″ 289 mm
- Mounting Depth: 2.25″ 53.3 mm
- Hex Nut Size: US 5/16″ (M8)
- Mass ID: 5/16″ (8mm)
- Max Mass OD: 2-7/8″ (74mm)
TS Parameters
All Measurements Are With Zero Added MassAssuming Surround Losses of Qmd = 7 @ 20Hz
- Mass to Deflection: 7.074 kgs
- Deflection in Meters, X: 0.02 m
- CMS, CMP: 0.000288201 Nâ‹…m
- Diameter Peak to Peak, D: 25 cm
- Active Piston Area Sd: 490.875 cm²
- Active Piston Area Sd: 0.0490875 m²
- Vas in Liters: 98.61116866 L
- Vas in Cubic Feet: 3.472224249 ft3
- Air Mass Mmr In Kgs: 0.006253515 kgs
- Air Mass Mmr In Grams: 6.253515291 g
- Moving Mass In Grams: 213 g
- Mms = Mmp In Grams: 219.2535153 g
- Free Air Resonance Fs = Fp: 20.02155789 Hz
- Peak to Peak Excursion in/cm: 4″ / 10.16 cm
- Qms: 8.5
- Qts: 0.35
Have a look at the behavior of the TC Sounds VMP passives at higher drive levels in the Data-Bass testing, in particular look at the long term output compression magnitude graph.
Diyuser, thanks for pointing that out! I will investigate the M12 passive radiator and see what Hornresp thinks.
In the mean-time, I've had a look online and found this: https://data-bass.com/#/systems/5b50df18c5c3340004aeaf17?_k=11zhn5
The last graph is Acoustic Power Per Litre.
It seems to use dB converted to Pascals, and then divided by the external volume of the cabinet. At 40Hz, the 2x B&C 21" IPAL subwoofer hits 0.265 Pa/L.
Taking the simulated 120dB@40Hz, and then doing the maths on my (unfinished, but 50L external so far), we get 0.4Pa/L. Looks like the search for output density is starting to pay off.
As noted earlier, there's still some work to make this cabinet useful in the wild world of sound reinforcement, but if I can avoid making the cabinet much bigger, this will be an impressive box indeed.
Chris
In the mean-time, I've had a look online and found this: https://data-bass.com/#/systems/5b50df18c5c3340004aeaf17?_k=11zhn5
The last graph is Acoustic Power Per Litre.
It seems to use dB converted to Pascals, and then divided by the external volume of the cabinet. At 40Hz, the 2x B&C 21" IPAL subwoofer hits 0.265 Pa/L.
Taking the simulated 120dB@40Hz, and then doing the maths on my (unfinished, but 50L external so far), we get 0.4Pa/L. Looks like the search for output density is starting to pay off.
As noted earlier, there's still some work to make this cabinet useful in the wild world of sound reinforcement, but if I can avoid making the cabinet much bigger, this will be an impressive box indeed.
Chris
Where are you getting your volumes from?
20in x 20in x 10.6in = 4,240in3.
4,240in3 / 1,728in3 = 2.45ft3.
2.45ft3 x 28.317L = 69.48L.
20in x 20in x 10.6in = 4,240in3.
4,240in3 / 1,728in3 = 2.45ft3.
2.45ft3 x 28.317L = 69.48L.
The Skhorn is 24 cubic feet, ~679.2 liters.
It hit 139.1dB (Max Burst) at 40Hz at 2 meters.
181/679.2=0.266 Pa/L (using 28.3L per cubic feet)
20/50=0.4Pa/L
At 2 meters, 120 drops -6dB to 114dB, 10Pa.
10/50=0.2Pa/L, but the Skhorn is higher.
If it hits 114dB at 40Hz at 2 meters, it will be impressive.
Once you have a calibrated measurement we'll see how impressive it actually is.
Has me curious!
Art