200L Barrel Subwoofer

[kiwifi]

For a sub, i'd normally run a nearfield of the woofer and splice it with a nearfield of the ports. Jeff Bagby wrote a white paper on the process that's a great read.

If you have access to a parking lot or similarly large space, a groundplane or other far field measurement would be very helpful. I'd run this higher than the forecast pass band to check for issues from the barrel.
I'd also run a distortion measurement at 85 & 95 dB for h2-h5 + thd.

And I'd also grab an impedance measurement to check the tuning frequency and for any resonances.

You want to check all of the most problematic areas extra thoroughly in any design. In this design, I'd suggest that will be from about tuning frequency and 1.5 octaves up and quite honestly most of the pass band and even above because of the barrel resonances.

I've never done this, but it may be interesting to get a Nearfield from the side of the barrel. Put the mic a couple mm from it and fire it up. It may turn out to be meaningless or could give you information about what the side of the barrel is doing to a few hundred hz.

Thanks. The ground-plane measurement is not going to be feasible, but I should be able to do some quasi-anechoic measurements in situ and redo the DATS measurements, probably on Wednesday.

If there is barrel resonance, it's not audible. With the sub crossed-over at 80Hz, I can stand next to the sub and still not locate the source of the bass. I have to touch the cone to confirm that it is working!

I think the reason is the low Vas (18L) of the driver compared with the volume of the enclosure (200L) means that any pressure generated by the driver is "diluted" inside the enclosure. The Qts > 0.7 means that he cone movement is able to remain damped independent of the enclosure volume.

The offset position in the top also helps reduce the formation of internal standing waves. Hollow fill is required in the bottom of the barrel to help stop reflected sound from the bottom of the barrel exiting through the ports at the top. I also have the barrel standing on a rubber pad.

Subjectively, the sub sounds great. Surprisingly so, considering that it can be built for about $100 in parts!

 

[Adam_M]

Thanks. The ground-plane measurement is not going to be feasible, but I should be able to do some quasi-anechoic measurements in situ and redo the DATS measurements, probably on Wednesday.

If there is barrel resonance, it's not audible. With the sub crossed-over at 80Hz, I can stand next to the sub and still not locate the source of the bass. I have to touch the cone to confirm that it is working!

I think the reason is the low Vas (18L) of the driver compared with the volume of the enclosure (200L) means that any pressure generated by the driver is "diluted" inside the enclosure. The Qts > 0.7 means that he cone movement is able to remain damped independent of the enclosure volume.

The offset position in the top also helps reduce the formation of internal standing waves. Hollow fill is required in the bottom of the barrel to help stop reflected sound from the bottom of the barrel exiting through the ports at the top. I also have the barrel standing on a rubber pad.

Subjectively, the sub sounds great. Surprisingly so, considering that it can be built for about $100 in parts!

 

[kiwifi]

Thanks. The ground-plane measurement is not going to be feasible, but I should be able to do some quasi-anechoic measurements in situ and redo the DATS measurements, probably on Wednesday

I just realized that quasi-anechoic isn't going to work down to 20Hz inside, so it may just be the corrected/uncorrected in room responses. I'll see what I can do.

 

[JRS]

Thanks. The ground-plane measurement is not going to be feasible, but I should be able to do some quasi-anechoic measurements in situ and redo the DATS measurements, probably on Wednesday.

If there is barrel resonance, it's not audible. With the sub crossed-over at 80Hz, I can stand next to the sub and still not locate the source of the bass. I have to touch the cone to confirm that it is working!

I think the reason is the low Vas (18L) of the driver compared with the volume of the enclosure (200L) means that any pressure generated by the driver is "diluted" inside the enclosure. The Qts > 0.7 means that he cone movement is able to remain damped independent of the enclosure volume.

The offset position in the top also helps reduce the formation of internal standing waves. Hollow fill is required in the bottom of the barrel to help stop reflected sound from the bottom of the barrel exiting through the ports at the top. I also have the barrel standing on a rubber pad.

Subjectively, the sub sounds great. Surprisingly so, considering that it can be built for about $100 in parts!

Basically an infinite baffle alignment, which when done well is IMHO very good bass indeed. And maybe the best I've heard when it comes to articulate, clean, and palpable. I wish I had the pics handy as one can see just how insane audio can make a man--and in this case, it's me cutting a 2.5'x 1.5' hole into the floor of a brand new home with a big old circ and a maniacal grin. The hole is what 2 ultra long stroke 18" woofers were pushing air through from the crawl space into my great room. I believe those had a Qt of 0.59.

If the opportunity presents itself, I'd do it again. Something very natural and lifelike about the sound that I rarely hear from a box--no back wave to eat or that eats itself like the open baffle variants you see with Orion type speakers.

Looking forward to hearing more impressions.

 

[Adam_M]

I just realized that quasi-anechoic isn't going to work down to 20Hz inside, so it may just be the corrected/uncorrected in room responses. I'll see what I can do.

That's why I suggested the nearfield measurement of the driver, walls, and ports. You'll only have resolution to a couple hundred Hz, but that's fine. Here's a link to Jeff's white paper that describes the reasoning and process in detail. white paper on quasi-anechoic measurements. In all honesty, if you are diving in, I'd read every word Jeff wrote in every white paper he published, but this is the one you want for the topic at hand.

So - I'd take a nearfield of these and post them.
Woofer
Port. You can do some fancy math and measure only 1, but I'd probably just measure all of them because it only takes a few seconds longer. I'd measure these full-range to see where the resonance is. You can find that with sims too, but since you'll have the tools out, just measure it.
Wall

Jeff has a tool that allows you to blend the nearfields of the port and the woofer to give you quasi-anechoic sums. You could run the measurements through those and post that too.

 

[kiwifi]

That's why I suggested the nearfield measurement of the driver, walls, and ports. You'll only have resolution to a couple hundred Hz, but that's fine. Here's a link to Jeff's white paper that describes the reasoning and process in detail. white paper on quasi-anechoic measurements. In all honesty, if you are diving in, I'd read every word Jeff wrote in every white paper he punished, but this is the one you want for the topic at hand.

So - I'd take a nearfield of these and post them.
Woofer
Port. You can do some fancy math and measure only 1, but I'd probably just measure all of them because it only takes a few seconds longer. I'd measure these full-range to see where the resonance is. You can find that with sims too, but since you'll have the tools out, just measure it.
Wall

Jeff has a tool that allows you to blend the nearfields of the port and the woofer to give you quasi-anechoic sums. You could run the measurements through those and post that too.

Thanks for the white paper, I'll give it a read.

I don't understand why I would be concerned about resonances in the hundreds or thousands of hertz when the sub will only be operating up to 120Hz? Surely those higher frequency modes won't get excited.

 

[Adam_M]

Thanks for the white paper, I'll give it a read.

I don't understand why I would be concerned about resonances in the hundreds or thousands of hertz when the sub will only be operating up to 120Hz? Surely those higher frequency modes won't get excited.

120hz isn't a brick wall, you are probably rolling off either 2nd or 4th order, so at 240hz, you'll only be down around 12 or 24db. That isn't THAT far down.

Also you have to think about harmonics. Your driver, and the barrel, and the port, can all be excited by a fundamental tone and output (as distortion) at multiples of that fundemental.

Regarding the port, while the fundemental pipe resonance may be above the pass band, I'd be shocked if h3+ and maybe h2 doesn't activate the pipe resonance. It can be a peak 10+dB high, so if it is 1 octave above the pass band, the resonance could be basically at the level of the fundemental.

Same thing with the barrel walls. Except I'm not sure they even really would stop much of the fundemental being so light and flexible.

 

[kiwifi]

Here is the frequency response...for some reason this took me all day!


The red trace is raw and close mic (as shown in the pic) to eliminate the room as far as possible. The 22Hz port tuning is clearly visible.

The green trace is measured from the same point after the following EQ is applied...
PEQ - 63Hz, -6dB, Q = 2
Butterworth LPF (24dB/Oct) -3dB @ 160Hz

 

[Adam_M]

Thanks for measuring. Looks like there's some peaking there, and being that far away from the woofer, I have a really tough time telling if it's room or driver. I don't really think this is a valid nearfield, and it's certainly not a valid far field - so it's not a terribly useful measurement, I'm sorry to say.

I'd suggest mounting the driver cone-up (if you are concerned about wiring, during testing, we almost all universally just run the wires through the ports). The upside-down mounting isn't doing anything for you (it's also not negatively impacting you - just inconvenient for this purpose). For a proper nearfield, the mic needs to be something like 0.1x the driver's radiating radius from the cone (this is important - not the nominal diameter of the driver - rather the diameter of the driver that produces sound. Measuring from 1/2 way into the surround to 1/2 way into the opposite surround is a good approximation), and as centered on the driver as possible. That means the mic should be no farther than about 1/2" from the speaker cone in this case, and closer is better. I normally target 1/4". I'd suggest you re-read Jeff's white paper, notably page 4 where this is explained and ask questions. We'll help you.

You'll need to do the same with the ports. And then you will have to combine the responses together to really understand what's going on with the output. Measuring the driver only like you did (when you get a proper nearfield) only tells you about the output of the driver itself. You are missing port information.

I'd still suggest taking the same nearfield on the side wall of the barrel, just to see if you are introducing distortion through the unbraced cabinet as well.

 

[kiwifi]

Thanks for measuring. Looks like there's some peaking there, and being that far away from the woofer, I have a really tough time telling if it's room or driver. I don't really think this is a valid nearfield, and it's certainly not a valid far field - so it's not a terribly useful measurement, I'm sorry to say.

I'd suggest mounting the driver cone-up (if you are concerned about wiring, during testing, we almost all universally just run the wires through the ports). The upside-down mounting isn't doing anything for you (it's also not negatively impacting you - just inconvenient for this purpose). For a proper nearfield, the mic needs to be something like 0.1x the driver's radiating radius from the cone (this is important - not the nominal diameter of the driver - rather the diameter of the driver that produces sound. Measuring from 1/2 way into the surround to 1/2 way into the opposite surround is a good approximation), and as centered on the driver as possible. That means the mic should be no farther than about 1/2" from the speaker cone in this case, and closer is better. I normally target 1/4". I'd suggest you re-read Jeff's white paper, notably page 4 where this is explained and ask questions. We'll help you.

You'll need to do the same with the ports. And then you will have to combine the responses together to really understand what's going on with the output. Measuring the driver only like you did (when you get a proper nearfield) only tells you about the output of the driver itself. You are missing port information.

I'd still suggest taking the same nearfield on the side wall of the barrel, just to see if you are introducing distortion through the unbraced cabinet as well.

The measurement is close enough to the speaker to eliminate room modes and boundary interference at the (low) frequencies of interest. The resonant frequencies can be estimated by measuring the major dimensions of the enclosure and then located on the frequency response as shown...



As you can see, all of the resonances outside of the 160Hz pass-band. Equalization pushes the first major resonance (H1@361Hz) down by 35dB and the 80Hz AVR crossover will push it down by another 50dB or so.

I will be looking at applying some panel damping material to the sides of the barrel because there is some resonance from the circumference of the barrel (C1@210Hz) but it is minor compared to the other modes.

I disagree that the measurement is not "terribly useful." It's a measurement of the sub's response in the room, which is exactly what I needed in order to apply the appropriate equalization.

 

[Adam_M]

I'm glad you are happy with the sub. Hope you get many years of enjoyment from it.

 

[kiwifi]

I'm glad you are happy with the sub. Hope you get many years of enjoyment from it.

This large, ported sub generates very low frequencies (down to 20Hz) in a way that seems effortless in comparison to my other (sealed) subs. It's a great addition, but I won't be getting rid of my other subs any time soon, because I appreciate their tightly controlled bass. I think that having both ported and sealed subs in the same room can give you the best of both worlds. Maybe by having a large ported sub handle the bottom octave and allowing the sealed sub to just roll off naturally, rather than forcing more power into the sealed sub with VLF EQ boost, below its resonant frequency.

 

[kiwifi]

The Barrel Sub has been upgraded! The barrel has been wrapped in 15mm thick automotive, panel damping foam from AliExpress.



I was going to apply the foam inside the barrel but the glued side of this foam is extremely unforgiving. It sticks to anything it touches and can't be removed without tearing the foam. It would have been impossible to maneuver the foam inside the barrel, so it is on the outside.

I remeasured the frequency response and it is pretty much the same over the pass-band (up to 160Hz), however the sound quality is improved. I don't think this is confirmation bias, but the fact that now sound from inside the barrel is not able to leak out of the sides of the barrel.

Light-brown is the wrapped response, blue the unwrapped... I have not shown the equalized response because that has not changed.



FYI: I also did a close measurement of just the port (with the end of the microphone just inside the mouth of the port)

 

[alex-z]

Rather than looking at the frequency response, look at the decay times. That will tell you a lot more about what affect the damping provides.

Particularly if you do it outdoors with no room influence to muddy the results.

 

[iMickey503]

Hmmm. I wonder if you ran it sealed would it work like a distributed mode loudspeaker? Like those Flat panels with an exciter?

Now I read something in Popular mechanics back in the day that if you add water, the speaker is able to go down lower. Not sure how that works. But seems like something fun to try out.

FYI, aesthetics don't matter to you at all? And you want loud and LOW bass that travels everywhere? You can always get some abandoned car and make some crazy 6th order "box" that's masked as a Car.
Its pretty efficient too and cost effective. Great if you live out in the sticks. Not HOA friendly however.

It also begs the question, would Bricks make a good Transmission line box? Its cheap. Its solid as it gets.
Can be put together anywhere. The Baffle would be Liquid nails between a layer of bricks sandwiched in between for a nice look. To make it critter proof? Make your own "Drone cone" (passive radiator) with some Bike tires and a plastic lid from one of those barrel bottoms. It should be plenty stiff and last that way for years.

Add that Fiberglass matt they use to stop growing grass and weeds on pathways or use some metal screen grate material to stop any critters that like to munch paper and rubber woofer soft parts or insects from getting in.

I'd use the Jenson drivers as they are cheap and available everywhere for $30 bucks shipped to your door. Pretty stout too!

Really cool this project worked so well for you Kiwifi. I may have to try this formula in the future for fun.