4
\$\begingroup\$

My subwoofer (with an internal amp) is showing difficulty reproducing low frequency sounds (sine 30Hz ~ 70Hz). As I lower the frequency, around 70Hz, I start hearing very slight distortion (hint of buzzing instead of pure sine sounds). At 30Hz and 35Hz, the distortion becomes more like a fluttering sound.

The subwoofer's specs:

Model: SW12-I

Input Sensitivity for max Rated Output: Line input 60mVrms @ 30Hz. Speaker input 700mVrms @ 30Hz.

Maximum Rated Output: 150 Watts into 8 ohms.

Max. Hum: 3.5mV @ 60Hz.

Max. Distortion: 4% At Full Rated Output.

Freq Response: 27 to 120Hz +0/-3dB w/12dB per Octave Rolloff.

Bass Boost: +6dB Boost @ 30Hz.

Low Pass Crossover: Variable from 40 - 120 Hz. 24 dB/Octave slope above 120 Hz.

High Pass Crossover: Line outputs fixed at 80 Hz, 6dB/Octave slope. Speaker outputs fixed at 120 Hz, 6dB/Octave slope

Its service manual is here:

https://www.audiolabga.com/pdf/SW12-15%20I.pdf

Originally (last year), the subwoofer started making a crackling noise after being warmed up for 10~15 minutes. I changed some power resistors, a capacitor, and two transistors. In the process, I shorted the circuit at least once, so it's quite possible that I damaged some components. There is also a distinct possibility that the driver (12") may have been damaged (due to hard-driving) and cannot reproduce the signal the amp outputs faithfully. The original crackling noise seems to have disappeared.

I have collected some data (measured and plotted) in the last several days. I'd appreciate it if people here can look at them and post their opinions whether the amp is definitely NOT working right or the amp is probably OK and the driver is the most likely cause of bad sounds. (Again, sine waves above 70Hz sound fine, both in volume and purity.)

V, I, P :values in tables

The reason that I didn't try increasing the levels 9 and up with the 9.2-ohm dummyload was that the amp started making a faint clicking noise, which I assumed to be the sound of the amp's current limiting protection kicking in. I could increase the limit by adjusting one variable resistor.

Level resistance values in a table

V I P charts #1 and #2

V I P chart #3, Level resistance chart

The following are waveforms of the output of the subwoofer's amp that as displayed by Soundcard Scope:

50Hz waveform

40Hz waveform

30Hz waveform

30Hz waveform, 50ms

To have Soundcard Scope (soundcard based oscilloscope) display these waveforms, I had to reduce the input level from 30~40Vrms to to less than 100mVrms, using only three resistors without potentiometers or diodes. In the pictures, voltage numbers are just relative, not absolute, the max being 1.9 or so. The input signal strength fed to the subwoofer's amp was about 668mV. The THDs calculated by Soundcard Scope were mostly less than 1% for all frequencies tested except for the worst being at 30Hz, still less than 2%.

===================================================================

Update (10/2/18):

For these measurements, instead of dummy loads, I connected the original driver and fed sine waves ranging from 25Hz to 120Hz. I also fed the signal (off the driver terminals) to Soundcard Scope and took screenshots for some frequencies. With this driver, the sine waves at 25Hz and 30Hz definitely show serious distortion.

I include a table showing the THD for this setup at various frequencies as well. As before, P is calculated using the formula P=VI. Impedance is calculated by the formula R=V/I. (I don't know if this second formula is meaningful, however.)

V I P Imp table

V I P Impedance

THD table

50Hz

40Hz

35Hz

30Hz

25Hz

Even though the waveforms at 25hZ and 30Hz clearly show distortion, I don't think this level of distortion causes the noise I think I hear. The noise is like paper flapping. Also, with a real driver (instead of a dummy load), some alteration of the input signal may be expected because its voice coil generates some electromagnetic force. (Just guessing.) The limitation of Soundcard Scope and soundcard-based oscilloscopes is something that needs to be considered, too.

Do you think we can conclude that the original driver is broken and needs to be replaced?

Or is there something seriously wrong with the amp? Like 6db boost at 30Hz, low pass filter, etc.?

Or the bias (adjustable through R84 in sheet 3 of the schematic in the service manual) may need to be adjusted? (I've adjusted R34 and R30 per realignment procedure stated in the service manual, but I don't know how to adjust R84.)

Any opinions or comments are appreciated. Thank you.

\$\endgroup\$
  • \$\begingroup\$ Interesting question, but I can’t see any distortion in your oscillograms. \$\endgroup\$ – winny Oct 1 '18 at 18:15
  • 3
    \$\begingroup\$ Excellent question, plenty of background information, and a clear attempt to investigate the issue yourself with nice data presented. I wish more questions on this site were like yours. \$\endgroup\$ – evildemonic Oct 2 '18 at 15:31
  • \$\begingroup\$ Excellent S/W alternative to Audacity for a collection on audio port. When a scope is not handy. Audacity allows much higher audio sampling rates. \$\endgroup\$ – Sunnyskyguy EE75 Oct 3 '18 at 20:51
  • \$\begingroup\$ Does this low voltage of ~1V cause the noise or 10 or 20Vpp?? \$\endgroup\$ – Sunnyskyguy EE75 Oct 3 '18 at 22:20
  • \$\begingroup\$ "Does this low voltage of ~1V cause the noise or 10 or 20Vpp??" I restored the table that had vanished when I changed a chart in the Update section. Level on the sub's amp is set at 4 (10 o'clock) & input level is 668mV for all the measurement taken for the Update section. 668mV is the max vol. I need on the receiver. On sub, Level 5.5 is more than enough. The voltage readings are all RMS, so at 70Hz (where I start noticing distortion), Vpp is 24V. At 30Hz, 73V. Please note that V numbers displayed by Soundcard Scope are not absolute. I lower the voltage to <100mV before feeding the soundcard. \$\endgroup\$ – zeron Oct 4 '18 at 1:13
2
\$\begingroup\$

The output from the amp looks fine.

In terms of the current protection - most audio amps cannot drive a sine wave at full power for very long - typically music power will be like 20% of full power in terms of actual power, this is why the PSU transformers will usually be rated lower than the max power output. The same will probably be true for the heatsinking to some degree - I suspect this might be why the limiting is kicking in.

The fluttering you hear is the passage of air through the bass reflex box as the frequency becomes closer to the helmholz freq of the bass reflex tube

I definitely sounds like there is a speaker problem.

I would do the following:

  1. Check the speaker surround to see if it is torn. If so the speaker will leak around the edges causing cancellation. This will be less of a problem at higher frequencies because there is less excursion at these freqs
  2. Check the speaker is mounted properly and that the bass reflex port is not blocked
  3. Make sure to test the sound wherever you are used to listening to it - for example if you normally use it in a corner try it there. Subs will have real difficulty operating in the middle of a room due to standing waves / lack of proximity effect
\$\endgroup\$
  • \$\begingroup\$ The driver is off the enclosure. The noise seems to be coming from the cone of the driver. I know the driver itself can shake at these low frecuencies, so I tried to hold it against something to suppress the vibration of the driver, but I think the noise is caused by the cone vibrating in some unintended way. One thing I've noticed is that it sounds like something flapping, so I think the frequency of the fluttering noise is much slower than 30Hz, probably more like 10Hz or even lower. \$\endgroup\$ – zeron Oct 3 '18 at 4:42
  • 1
    \$\begingroup\$ Ahh the problem with testing with the driver out is that your driver is that low frequencies from the front and back of the driver are cancelling each other out. This does not happen at higher frequencies, as they have a beaming effect and will not wrap around the driver. This is why the minimum enclosure is a flat board with a hole in it - the purpose of the board is to stop the destructive interference between back and front of the driver at low frequencies \$\endgroup\$ – TheMachinist Oct 3 '18 at 14:40
  • \$\begingroup\$ "...the minimum enclosure is a flat board with a hole in it..." I'll mount the driver back in the hole in the original enclosure and test later.. However, the noise was present when everything was mounted on the enclosure, to a lesser degree. As you pointed out, these sine waves are too demanding to be reproduced satisfactorily for any prolonged period of time. Should most decent subwoofers be able to play low frequency sine waves for a second or so, at least? Can you look at the update (10/2/18), too? I tested with the driver driven this time. Thanks. \$\endgroup\$ – zeron Oct 3 '18 at 16:03
  • \$\begingroup\$ I tested with the driver mounted on the original enclosure, with the cutout hole for the amp on the back left open. The noise was still there. \$\endgroup\$ – zeron Oct 4 '18 at 0:29
  • \$\begingroup\$ I'm going to test the subwoofer in another location in my room and also in another room (with more open space with fewer pieces of furniture) this weekend. (In short, my room is not an ideal listening room.) I'm now suspecting that reflections of the sound bouncing off many tall bookcases in my room (similar to IKEA's Expedia), some standing in the middle of the room) may be affecting the driver's behavior. II have a question: if the driver is affected by reflections of the sound if creates, would the voltage readings at its terminals change, as captured by Soundcard Scope for 30Hz and 25Hz? \$\endgroup\$ – zeron Oct 5 '18 at 20:22
1
\$\begingroup\$

Speaker cones tend to buzz then the windings rub the magnetic core or when cone is ripped. **

The is most likely from speaker damage. But I don't recall if the speaker buzzes at this level or much higher.

If the voice-coil is misaligned with a gap of <<1mm , max excursion at the lowest frequency can cause winding buzz or whooshing from cone damage to the neoprene or cone suspension. If torn this can result in uneven suspension bias and cone tilt causing the some friction.

The Amp has a bass boost set to +3dB at 30Hz with total harmonic distortion (THD) visible and rolls off sharply at 25 Hz with even more THD visible flattening towards squarish.

Conclusion:

The Amp could be improved and could be the cause if this were tested at normal listening levels but if this voltage produces buzzing it's the voice coil or suspension ring that are visibly damaged perhaps the braided wire interfering, or torn.

example of your frequency response and THD

Due to the Fourier Series, comparing a sine to the extrapolated triangle is a fixed ratio of ~ 70% Vpp. When this ratio reduces, it means harmonic distortion has started to attenuate the peak due to squarish response, by adding harmonic content. The 25 Hz has reduced to ~50% rather than 70%, meaning that THD has begun at this low power level. This is chiefly caused by the PS cap ripple. It may get much worse at max power level. enter image description here

Which reminds me

** Some may know the phenomena or have your speakers buzz from mobile phones when syncs up to tower . dit-da-da-dit repeated every 5 minutes or so, when placed near the speaker coils in a car door or on a PC. They resonate the chirp burst RF and cause an EM response and in some cases a radial motion instead of axial one. This temporarily brings the coil closer to the rigid magnetic parts, meaning not only can you hear it with the power off , but it buzzes much easier as the coil touches the magnetic assembly.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.