What's causing violent woofer oscillations?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello,

can anyone explain to me what causes a woofer to have uncontrolled oscillations with each bass punch, even on small sound levels?

Here is my situation. I got this small active subwoofer as a support to my PC speakers. The first thing I noticed was a great level of distortion each time the bass would punch. It sounded really bad so I decided to replace only the driver. I ordered Visaton's W130S. To my suprise this driver too sounded more or less the same. It's like the membrane can't handle the bass and starts to have violent oscillations with each kick. And I am not talking about high listening sound levels.

What I noticed is that when I block bass reflex port, the oscillations become considerably less audible.

The amplifier that drives this sub has only 22W (bridged TDA1519A) while Visaton W130S is rated up to 50W and still cannot perform as it should. So the only thing that occurs to me is that these oscillations happen:
"when a speaker is in a ported enclosure and is driven with frequencies below the port tuning frequency."

If this is the case, is there anything I can do to reduce the oscillations?
 
It's possible that the oscillation is electronic in nature and due to faulty amplifier operation. However, it could also be that the acoustic system is seriously underdamped and tends to reverberate at some low frequency. If this is true, it may be correctable by a shunt resistor on the voice coil, or a different cone surround material. Decouple the rear and front of the speaker. You did that by blocking the port and it helped; maybe a port isn't called for.
 
Presumably this happens with every type of music.

Can you feed the sub from the output of something else? Such as an mp3 player? This would help confirm that the problem is related directly to the sub and isn't some processing that the computer is doing.

I can easily see some smaller computer speakers having a high pass at 40hz or something, to protect the drivers form over excursion at frequencies they cannot reproduce - hence they wouldn't be affected.

Sealed box loudspeakers will help reduce the excursion of a driver at very low frequencies when compared to a ported box, so this isn't a surprise.

The power rating of the driver has nothing to do with its maximum excursion level however. It may only take 5 watts to exceed it's mechanical limits at say 10 hz.

If there are any variable controls on the sub I would also try playing with those a bit. It's possible that you've created a situation that isn't stable within the analogue filtering before the amplifier. It's a long shot, but if one of the stages isn't properly designed, I could see it perhaps happening.
 
Any music below the woofers port resonance has no damping and will cause the type of severe excursion you describe. Clearly your amplifier isn't high-pass filtering it (check if your amp has a highpass switch).

If you cant filter the low frequencies, I would seal the port (plug it with a sock) and use it as a sealed enclosure, otherwise you eventually destroy the driver.
 
Thank you for the answers! I have tested the amp, connected it to another speaker box with Visaton fullrange FR 10. It acts the same as the other two drivers. The distortion sound is specific, I can clearly hear additional oscillations of the driver as it's trying to stabilize).

And when I connect all of these drivers to an external amp, they don't have such problems but the bass is not nearly as loud as with sub's internal amplifier.
Thus I'm concluding that the internal amp gives too much low bass for these drivers to handle. I don't have a schematic but I see that there is a second order low pass filter with op amp TL072. A a last resort, I could maybe try to put a high capacity electrolytic in series with the driver?

Anyway, I've never seen a crossover that cuts lower bass to a woofer. If a box is tuned for eg. 50Hz, does it mean that any signal below this frequency should be filtered? That is certainly not the case with many bass reflex designs that I've seen. How do they manage excess low bass?
 
I used a pair of Visaton W130S per side on my main system, 35L box for both, tuned to around 42Hz. They perform exceptionally well, given the driver size, though the cabinet requirements for max.flat in a ported box are large.

I'm feeding them with a 50w amp (per side), and, even at rather silly levels, I can't get much excursion on them, unless I go with a signal generator.

There's something very wrong with your amplifier, and plugging the port won't stop the driver cooking.

Edit - what music are you using when such a problem occurs?
 
There is very likely a subsonic filter designed into that amplifer. Any ported system done right has filtering below below cutoff. Since a commercial product can't guarantee the user will not drive it with a source with lots of power below cutoff, it has to be electronic. The first thing I would do is look for bad solder, especially if it's brand new. If it's old you may as well look for bad capacitors if there's no bad solder. If you could come up with a schematic or a model number maybe someone could help you pick this off fairly easily. The output capacitor is not a decent fix for this. A real fix is probably cheaper and is definitely higher performance.
 
Last edited:
One other thing to consider -- I noticed this with the national semi chips -- you can get them to thermally oscillate if the heat sink isn't sized properly -- it's a state between the protection circuitry and normal operation.

national also recommends 100nF//10uF//1,000uF as close as possible to their chipamp pins to prevent motor-boating.
 
What music are you using when such a problem occurs?
I find hiphop music suitable for bass testing.

I noticed this with the national semi chips -- you can get them to thermally oscillate if the heat sink isn't sized properly
No overheating is present. The heat sink was mildly warm during listening.


I spent the entire day trying to figure out what's wrong. The sub is 10 years old so I replaced all electrolytic capacitors. Nothing's changed. Even the sock trick won't do it in the long run - there's just to much unwanted oscillations. 100uF in series with the woofer - no go.

But I think I know whats wrong. Even on low sound levels - levels that you wanna use after you wake up, cone movement is visible. And if I crank it up a bit more (normal listening levels), the cone just goes crazy accompanied with distortions.
When I realized that W130S has excursion limit of +/- 8mm it was all clear to me. The box is only 7 liters and basically cannot produce strong bass so the bass needs to be super amplified. But the drivers cannot handle that.
That's my theory basically. And here is what it looks like, a conventional PC subwoofer made by Macase (they don't make speakers anymore).


An externally hosted image should be here but it was not working when we last tested it.
 
With such a tiny box, it's small wonder they've tried to eq it to death.
I'd give the driver a reasonable-sized cabinet, sealed, with a touch of eq to get an f3 of around 50Hz.

With the driver moving that far, I'd expect a lot of output, so there's a fair chance the driver isn't sealed properly, or the port tuning is all wrong.
 
Is the cone moving at a low rate, i.e. maybe only a couple cycles per second? Is this present in the source audio? If not, then your amp is faulty. Try leaving the subwoofer in the box it is in, and connecting it to a different amplifier. If the symptom is not present with that, then your amp is faulty again.

If your amp is faulty, and you plan on buying another plate amp to replace it, you should probably redesign a box optimized for the driver's parameters to get the most out of it, or at least decent performance at all.
 
I gave up on this subwoofer. And I really wanted to make it work. Redesigning would take me much time and patience, too much for the value of the sub. Now I'm left with a new driver that I can't use :(

I decided to buy a new active sub for 320€ (that's around 410$). And this is the very maximum I can spend. So if you good people have any recommendations within the specified price, please write.

I need it for a small room and I don't require much power. What about these:

Yamaha YST-RSW300 (340$)
KEF C4 (417$)
There's also Mission M3as for 370-380$.
 
For a fraction of the price, you could have a reasonable bass unit.

You'll need:
Some components for a Linkwitz Transform (f0=85Hz, Q0=0.7, fp=50Hz, Qp=0.7)
Some components for a high pass filter (optional, but recommended),
Enough wood to make an 8L cabinet,
An amplifier.

These would be cheaper than buying an subwoofer, and the result would be pretty reasonable.

The Linkwitz Transform will take the woofer to an f3 (-3dB point) of 50Hz, f6 of 38Hz, and an f10 (many consider this more useful) of 28Hz. Now, the cone excursion to get to those really low frequencies is quite a lot, so adding a HP filter would reduce the output at the very low frequencies, to save the speaker trying to reproduce sounds it physically can't. 12 or 18dB per octave set around 35Hz would be about right.

The components needed to make these circuits cost pennies, but you will need an amplifier of maybe 50w (should be plenty, but going lower powered wouldn't be advised).

The cabinet (if you like) would be a 200mm cube, internally (far smaller than those you mentioned). If you stuffed it, you could get away with a little less, but exactly how much less is subject to much debate. Most people quote 10% less, so, considering driver displacement, 8L internal would be about right. Bracing would be up to you, but if you built it from thick enough material, it's unlikely to need it.

Chris
 
Build your own. It will (should) perform better than a commercial unit, because you will build it according to your needs and tastes, and it will cost less.

Do you want low frequency reproduction (<35Hz)? Or are you more interested in bass mainly contained in music (>30-35Hz)?

You can go over to parts express and check out there deals. I just finished a subwoofer using the 15" Quatro QT385-4. I payed 130 for it, but now they have it on sale for only 75 bucks. Damn I hate it when that happens. Its an amazing deal.

If you don't want something that big,
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=295-550
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=295-555
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=295-545

These are just examples. If you do decide to build your own, I gaurantee you will be more satisfied, because you will have built it with your hands, and will appreciate it much more. We can come up with some simulations and recommended box sizes for you, along with plenty of guide lines for building a box.

EDIT: Sorry, I forgot you had a driver already. I guess the above might just be redundant.
 
Last edited:
I've never built a speaker box before so I'm not that confident in my "designing capabilities". And I think it wouldn't be alright if I asked someone here to do my homework. That's why I'm going with the safest solution - a final product.

I will build that 7L closed box anyway. That's not the main issue. Designing the active filter, however, is. That's a multiple stage filter:
-Input channel mixer
-Linkwitz transform circuit (as chris661 proposed)
-High pass filter

Does anyone have a complete circuit diagram by any chance? :)
 
Right,

Active Filters

For reasonable listening levels, you won't need the high pass filter. If you want it to go louder, you will need it, to stop the woofer bottoming out as it tries to play loud, LF sounds.

For a volume control, I use a mixer... Anyway, you'll need a LP filter, too, so only LF sounds get to the woofer. Anyway, the link above will give you everything you need. (except a pre+power amp)

Here's a thought for you. Buy a "faulty" amplifier off ebay (a hifi amp, with an integrated pre), where only one channel works. Tap off the pre-amp's PSU to power the crossovers, LT, etc, then connect the power amp to the output from those.

Chris
 
There we go.

This configuration does give you useful bass output quite low, but it's still only a 5" driver in a 7 litre box.

You can raise the cut off point of the high pass if this is a problem.

Note that the high pass comes first in the circuit with the Linkwitz transform (LWT) second, followed finally by a low pass. This is on purpose.

The LWT applies a lot of gain at low frequencies and could potentially clip if fed with a very low frequency signal. Putting the highpass before it stops this from happening.

Also LWT circuits can sometimes produce a reasonable amount of hiss/noise with certain component values. Placing the low pass after the LWT will filter this out preventing you from hearing it.

Obviously the low pass needs to be setup with respect to the main loudspeakers. As you can see the one I've attached to it is 4th order LWR @ 80hz.
 

Attachments

  • visasub1.jpg
    visasub1.jpg
    169.1 KB · Views: 97
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.