DIY subwoofer repair

[suttondesign]

Regular readers of this forum may recall that I built two dual-opposed subs in 1.5L enclosures wrapped in large books. Both had scan-speak drivers, the 9" and 11" ones. The 11" (28wxxxx) drivers were very expensive. I built that sub in august 2024. It always had a weird chuffing sound from one of the drivers that I thought was air somehow leaking, like through the screw holes or at the flange, but I never could get the gremlin to go away. It was showing up in my REW measurements too, as a lot of distortion from that one sub at below 30hz.

So a couple weeks ago, with some time off, I built completely new, and actually quite a bit better boxes, this time using the 3/4" baltic birch for the tops and sides but 1" for each end. And I made super-sure, using gaskets, wood glue, and pocket screws, that the cabinet had no air leaks and no way for air to get around the offending driver.

Still the chuff-chuff. Very pronounced at 20hz, whereas the other 11" and the two 9" had zero chuff-chuff. Sigh. A shout-out to Madisound here. The driver was out of warranty, but they dealt with me very fairly on a replacement, and also gave me pointers on how to look at the bad driver to see if there's something repairable. Adam at Madisound has always been super-helpful to me with long orders (crossover parts, Linkwitz kits, the works).

The new 11" arrives in a few days. Since I know you care, I'll update the forum on whether I solved the chuff. Our hobby isn't free, so I'm not particular upset, but wow, those scan-speaks have gotten so expensive. Were I to do it over again, I'd probably just buy Daytons and call it a day. I think I was feeling flush a couple years ago. But maybe if Trump gives me $5,000, I won't feel quite so bad.

 

[Keith Flanker]

Since the chuff followed the driver into a brand-new, well-sealed box, it is the driver, not a leak. Distortion showing up in REW below 30 Hz, together with chuffing at high excursion, points at the voice coil/spider - a rubbing coil or a damaged former makes exactly that noise. Quick confirms: push the cone gently at 3-4 points around the surround and feel for scraping, and run a free-air impedance sweep against a good 28W - a double peak or asymmetry means the coil is off-centre. Taking Madisound's replacement was the right call.

 

[MAB]

Regular readers of this forum may recall that I built two dual-opposed subs in 1.5L enclosures wrapped in large books. Both had scan-speak drivers, the 9" and 11" ones. The 11" (28wxxxx) drivers were very expensive. I built that sub in august 2024. It always had a weird chuffing sound from one of the drivers that I thought was air somehow leaking, like through the screw holes or at the flange, but I never could get the gremlin to go away. It was showing up in my REW measurements too, as a lot of distortion from that one sub at below 30hz.

So a couple weeks ago, with some time off, I built completely new, and actually quite a bit better boxes, this time using the 3/4" baltic birch for the tops and sides but 1" for each end. And I made super-sure, using gaskets, wood glue, and pocket screws, that the cabinet had no air leaks and no way for air to get around the offending driver.

Still the chuff-chuff. Very pronounced at 20hz, whereas the other 11" and the two 9" had zero chuff-chuff. Sigh. A shout-out to Madisound here. The driver was out of warranty, but they dealt with me very fairly on a replacement, and also gave me pointers on how to look at the bad driver to see if there's something repairable. Adam at Madisound has always been super-helpful to me with long orders (crossover parts, Linkwitz kits, the works).

The new 11" arrives in a few days. Since I know you care, I'll update the forum on whether I solved the chuff. Our hobby isn't free, so I'm not particular upset, but wow, those scan-speaks have gotten so expensive. Were I to do it over again, I'd probably just buy Daytons and call it a day. I think I was feeling flush a couple years ago. But maybe if Trump gives me $5,000, I won't feel quite so bad.

Madisound was very helpful in getting me a replacement Seas Excel driver with a partially glued surround.

It was surprisingly difficult to figure out where the noise was coming from. I could hear it. I could see it in the impedance trace, distortion, and frequency response. I finally figured it out driving it at specific frequencies that caused the surround to burp.

The Seas are also a bit expensive.

 

[respice finem]

Madisound was very helpful in getting me a replacement Seas Excel driver with a partially glued surround.
View attachment 557907
It was surprisingly difficult to figure out where the noise was coming from. I could hear it. I could see it in the impedance trace, distortion, and frequency response. I finally figured it out driving it at specific frequencies that caused the surround to burp.

The Seas are also a bit expensive.

Looks like quality control took a day out...

 

[fpitas]

Looks like quality control took a day out...

Yeah, disappointing. Those aren't cheap drivers by any means.

 

[suttondesign]

New driver has the same chuffing, so it's not the driver. Now why would two identical drivers in a sealed enclosure, in dual opposed configuration, result in one chuffing audibly with a 20hz test tone, and the other not? All I can think is that it's something in the way I wired them in series?

 

[Sokel]

New driver has the same chuffing, so it's not the driver. Now why would two identical drivers in a sealed enclosure, in dual opposed configuration, result in one chuffing audibly with a 20hz test tone, and the other not? All I can think is that it's something in the way I wired them in series?

That would show up when you measured them with the multimeter.

Rest of the chain? Some filter, a subsonic one for example peaking down there by a wrong config?

 

[NTK]

Looks like quality control took a day out...

The compression from the movements of the driver diaphragm can create some pretty significant pressure rise inside the cabinet, especially for sealed designs. For example, if the pressure increase is 1 psi (pounds per square inch), which can be the result from a roughly 7% reduction in the net cabinet volume, with the surface area of an 11 inch cone being around 95 sq in, the pressure force on the diaphragm is 95 lbs.

1 litre is 61 cubic inches. It doesn't take that much displacement for a 95 sq in diaphragm to generate the 1 psi pressure rise for a small cabinet.

You'll want to be sure your drivers can handle this kind on mechanical loads.

 

[MAB]

The compression from the movements of the driver diaphragm can create some pretty significant pressure rise inside the cabinet, especially for sealed designs. For example, if the pressure increase is 1 psi (pounds per square inch), which can be the result from a roughly 7% reduction in the net cabinet volume, with the surface area of an 11 inch cone being around 95 sq in, the pressure force on the diaphragm is 95 lbs.

1 litre is 61 cubic inches. It doesn't take that much displacement for a 95 sq in diaphragm to generate the 1 psi pressure rise for a small cabinet.

You'll want to be sure your drivers can handle this kind on mechanical loads.

In my case, that Seas driver was brand new! Seas obviously had an issue with the adhesive bonding step in the manufacture of the surround. And their QC team didn’t identify that the had an issue in their factory or spot it on the finished product.

It was hard for me to spot since the surround lies so perfectly flat. Doing a frequency sweep made it super clear, to the point I could see the surround lift at certain frequencies.

 

[suttondesign]

Recall, it's two identical drivers in a dual-opposed configuration in a sealed cabinet. Only one is chuffing. I'm going to switch the leads on the two drivers to see if it's upstream.

 

[respice finem]

The compression from the movements of the driver diaphragm can create some pretty significant pressure rise inside the cabinet, especially for sealed designs. For example, if the pressure increase is 1 psi (pounds per square inch), which can be the result from a roughly 7% reduction in the net cabinet volume, with the surface area of an 11 inch cone being around 95 sq in, the pressure force on the diaphragm is 95 lbs.

1 litre is 61 cubic inches. It doesn't take that much displacement for a 95 sq in diaphragm to generate the 1 psi pressure rise for a small cabinet.

You'll want to be sure your drivers can handle this kind on mechanical loads.

OTOH, these are not cheap "no name" drivers, so I would expect them to "survive".

 

[suttondesign]

Update. I gave up on the dual-opposed scan-speak 11" configuration. One of my two original drivers from 2024 had an obvious problem - I could feel air passing across the rubber surround. Rebuilding the enclosure was a bust for that reason. Madisound was great and took back the replacement driver I ordered. For now, I sawed off one end of the existing enclosure and replaced it with a plain panel, glued and screwed. So for the time being, I'm using the same 1.5L baltic-birch enclosure with just the one 11" scan-speak driver at one end. At least it's running clean - it tests fine. What a headache.

I plan to build yet another 1.5l enclosure to try dual-opposed SEAS L26ROY, since I have 4 of those on hand. I guess I'll keep the one scan-speak 11" and do something with it some day.

 

[Penelinfi]

Pity about the driver; should be able to be fixed by the sound of things. But depends if you can find someone who fixes things if you aren't confident in your own speaker fixing abilities - it can sometimes be fiddly or tedious work

 

[suttondesign]

Got a new dual-opposed box built for the seas l26roys and got it up and running. huge output, and finally no chuffing. but I am sick of building 1.5l birch-ply boxes.

i plan to mothball the working single-driver 11” scan-speak. when i get a windfall, maybe i will try yet again to sister it in a dual opposed. it’s a cinch to rout out another circle at the opposed end.

hopefully this is my sign-off.

 

[suttondesign]

The SEAS L26ROY, dual-opposed, in a 1.5 cu. l. box, is a mighty beast when EQ'd to boost 20hz significantly and cut off beginning at 50 or 60hz (a "Linkwitz transform"). I also filter out below 15hz to avoid overexcursion. The dual-opposed is powered by a Hypex 500x, drivers wired in parallel since the amp can handle it.

I suspect that the Scan-Speak 28w 11" that I had in the same configuration was suffering significantly from an air leak from the surround on one of the drivers. I guess I'll never know - I threw the defective driver away in frustration. The air leak was under the flange, where the rubber surround is tucked under the flange. No way to fix it that I could tell.