Sonotube Subwoofer Output Problems

[arande2]

About a year ago, I built a sonotube-based subwoofer. I used an 18" Fi Car Audio Q18 (old version) driver, 30" sonotube with a volume of 800 liters (28cu ft), and MDF endcaps with an 8" port (should've used 12 now that I look back). The power is provided by the well-known EP4000.

It was originally a 12.5hz tune, but I have ended up with 17.5hz (I can add port length on top for 12hz, but I do not since it just takes away output).

My problem is that the output is way below what I think I it should be.

I used WinISD to model the output and it tells me that this driver in this alignment should be giving me right around 91dB through the passband from 17hz to 100hz for 1 watt (2v) of input in the ground plane. When I measured it after building it, I got an output of 84dB through 45-100hz, sloping gently to 76dB at the tuning frequency. I am confident in my measuring equipment as I tested speakers with a known response and got accurate results.

I can also verify that others have done a very similar build without any issues.

I have gone through many placements within my room and can verify something is off with the subwoofer. The original idea (with wall/corner gain) was that I could hit the SPL required at my listening seat for the most bass-heavy scenes in movies or transients in music while also having low distortion and what could be termed effortless depth and bass - all of this with maybe 100 watts at the most. Instead, it takes 100w even for medium -level listening.

My actual listening space is around 2500 cubic feet, but the whole space (huge opening) is about 8000.

In the current positioning with the subwoofer in the corner, I get about 77dB at best for 2v driving the subwoofer (4-ohm). It will get down to 68dB (these ARE corrected numbers) in the lowest ranges where I was wanting a lot of output capability. This is probably the most efficient position I could sit in without being in a corner or against a wall.

The absolute best sensitivity I get in the room is about 85dB and that is in the opposite corner at one frequency.

For a comparison, my main speakers will hit 115dB at my listening position at 50hz off of 250w between the two. ...

WinISD tells me I can expect 123dB ground plane at 18hz with full-tilt operation. Add room and boundary gain to this and BOOM! I didn't expect 100% results, but I also didn't expect what I am currently experiencing.


Sooooooo my question is of what I can do to find the problem (assuming there is one and I'm not expecting too much) or what may be causing this in the first place.

Thank you for comments. I will try to answer any questions.

[Brian Steele]

Originally Posted by arande2

Sooooooo my question is of what I can do to find the problem (assuming there is one and I'm not expecting too much) or what may be causing this in the first place.

Thank you for comments. I will try to answer any questions.

What are the t/s parameters of that driver. I suggest measuring the t/s parameters yourself, to see if they match published specs.

[blamus]

the spl value provided by the manufacturer is usually wrong, especially for car audio drivers. If you measured the T/S yourself, and then Vas/SPL, you will be amazed how different the values are.

[sreten]

Hi,

What is the driver Vas ? 270L ? ~ 9.5 cuft ?

A box/tube of 800L, 28 cuft is utterly pointless, it won't work well.

You have used the classic nonsense vented alignment for an ~ 0.6 Qts driver,
which always leads to boxes far bigger than driver Vas, not a good idea.
There is no point exceeding ~ 150% of Vas, its simply a waste of space.

Around 300L tuned to 18Hz or 200 to 300L sealed looks reasonable.

For 800L I'd used two tuned to around 15Hz.

Note 200L sealed will linearise the suspension over the larger boxes.

Note if the SPL figures below are for 2.83V/1m the real efficiencies
for low impedance drivers will be much lower due to low impedance.

rgds, sreten.

Q18

Fs: 24.5 Hz | 23.8 Hz
Re: 0.7 Ohms/coil | 1.4 Ohms/coil
Qms: 5.66 | 5.73
Qes: .64 | .63
Qts: .58 | .57
Mms: 322g | 342g
Cms: 0.13mm/N | 0.13mm/N
Sd: 1210cm^2 | 1210cm^2
Vas: 268 l | 268 l
Spl: 89.7dB 1W/1m | 89.4dB 1W/1m
Bl: 10.4 N/A | 15.1 N/A
Xmax: 27mm
Rms: 1000W
Sub OD: 18.500”
Cut ID: 16.750”
Mounting depth: 10.000”
Displacement: 0.24cuft

[arande2]

Brian, blamus, and sreten, thank you for your assistance.

Mr. Steele and blamus, I have been meaning to measure the T/S parameters, and I guess it shows how much of an amateur I am considering that I haven't yet. I haven't even measured the electrical characteristics of the alignment, either. I really should have done all of that when I first built it, but I was too excited to hear the final result and it got shoved off. That is on my list within the week. It also seemed to me that Fi has respectable quality control on their Q line, which is intriguing considering your comments on the often high variance of car audio drivers.

Sreten, based on your comments I am guessing that you disagree with the notions of infinite baffle and EBS alignments. Could you explain for what reasons exceeding 1.5x Vas is a bad idea? There is a lot of information out there and it is important to me to understand the WHY so that I don't make the same mistakes again or can make better decisions in the future.

I understand that in practice there are many variables (some that cannot be measured) that are not accounted for in simulations. It seemed to me that a large enclosure would give me a good amount of output for a given input, but considering my experiences, I am having to rethink that.

The SPL that models for half-space 1w/2v/1m is 91dB for my alignment so I would think the manufacturer has not listed the 2.83v numbers. Others have also used this alignment with good results, but I'm thinking there are legitimate reasons why it is not a good idea.


Lastly, I am having one more issue. I get a port chuffing issue with an 8.5" port at underwhelming output levels. Driven with 40v RMS, I start to run into bad chuffing issues at 17.2hz (the ACTUAL tuning frequency as measured in the current humidity and pressure). This models at 30 m/s port velocity, so that's why. I could use a 12.5" port without excessively lowering the resonance and this would allow the full 90v RMS drive level (which apparently I need to be able to use) with 27 m/s port velocity.


Comments are appreciated, of course.

[Brian Steele]

Originally Posted by arande2

Brian, blamus, and sreten, thank you for your assistance.

Mr. Steele and blamus, I have been meaning to measure the T/S parameters, and I guess it shows how much of an amateur I am considering that I haven't yet. I haven't even measured the electrical characteristics of the alignment, either.

I ran the numbers, and my sims are a close match to yours. I suggest as a start measuring the Re of the driver (I'm assuming that you've got both coils wired in series), to see if it matches the 2.8 ohm specification.

[arande2]

Yes they are in series (as it doesn't make sense to run at 1 ohm!!) and I measured 2.7 as the Re.

[sreten]

Originally Posted by arande2

Could you explain for what reasons exceeding 1.5x Vas is a bad idea? There is a lot of information out there and it is important to me to understand the WHY so that I don't make the same mistakes again or can make better decisions in the future.

Hi,

Because driver Vas will dominate the response, above ~ 1.5xVas the sealed
box Q will hardly change at all and consequently neither will vented responses.

consider a 0.6 Qts 6.5" driver with a Vas of 27L (near 1 cuft ft). Your
effectively using it in a 80L (near 3cuft box). Its pointless. With that
volume a very good 12" is totally on the cards and will be far better.

Any driver with Qts near 0.6 should be used sealed well stuffed IMO.

Your 800L sonotube could take a monster of a driver.

As soon as your heading over Vas you have the wrong driver IMO.

Sealed drivers should have very large Vas, very low Fs and low Qts.
Put in a box were Vbox << Vas to raise Qbox to reasonable, the
air spring will vastly linearise the drivers suspension, wanted ....

Its understanding horses for courses basically.

Use the driver for a sealed car sub and get a monster hifi driver.

rgds, sreten.

[tb46]

Hi arande2,

From your description it appears that you may be trying to make a defective driver perform up to factory specifications. Maybe you can contact Fi Car Audio and see if their customer service can be of help.

Regards,

[xmitterengineer]

arande2 "Yes they are in series (as it doesn't make sense to run at 1 ohm!!) and I measured 2.7 as the Re. "

Are the coils out of phase?
May I suggest reversing the connections on one coil.
I have (rarely) seen electrical polarity mis-marked on voice coil connections. This can generate considerable head scratching.
Good luck Sir.

Rickey.