I recently did a measurement of my MTM speakers in my living room. I posted the measurement which included a distortion plot to another thread.
Greebster commented on the peak in distortion at 600Hz. I'd not thought much about it, put it down to some sort of room interaction, but on further investigation (after a pm) He was right, and it is not the room.
Note that I have not actually noticed this distortion when listening. But it is there just the same. I'm starting this thread as an investigation into what might be causing it. It could well turn out to be the drivers themselves, but could also be something to do with my design, so it is worth investigating and I might learn something (or perhaps even improve the sound).
Below are some plots. The first is the in room measurement. The thing that seemed most odd with it was that there seemed to be multiple harmonics at 600Hz.
The second picture is a measurement I took outside (it's 1/8th octave smoothed). Note the same 600Hz 2nd harmonic (holm won't show all harmonics at once unfortunately). Other harmonics do not show up at 600Hz in holm. Note also that I am dubious about holms reporting of distortion at higher frequencies, tests of amps have shown a rising distortion that is not present with other measurement software.
The next two plots are of the impedance curve of the woofers in the box (no crossover). They clearly show a problem at 600Hz. I have to try and dig up the free air measurements to see whether it is present in those too.
The third impedance plot is a zoom of the impedance plot of the finished speaker with the crossover in place. Oddly the 600Hz glitch is gone..... I'm wondering if it is only one of the two speakers exhibiting the behaviour...
Finally a couple of pictures of the actual speaker.
So with that as a starting point, I think that I can say that the problem is coming from the midbass units and that it is not related to the crossover.
Greebster commented in PM that it could be dimention related, though that was when I thought the impedance plot showed nothing... last pic is my original plan for the baffle dimensions.
Tony.
Greebster commented on the peak in distortion at 600Hz. I'd not thought much about it, put it down to some sort of room interaction, but on further investigation (after a pm) He was right, and it is not the room.
Note that I have not actually noticed this distortion when listening. But it is there just the same. I'm starting this thread as an investigation into what might be causing it. It could well turn out to be the drivers themselves, but could also be something to do with my design, so it is worth investigating and I might learn something (or perhaps even improve the sound).
Below are some plots. The first is the in room measurement. The thing that seemed most odd with it was that there seemed to be multiple harmonics at 600Hz.
The second picture is a measurement I took outside (it's 1/8th octave smoothed). Note the same 600Hz 2nd harmonic (holm won't show all harmonics at once unfortunately). Other harmonics do not show up at 600Hz in holm. Note also that I am dubious about holms reporting of distortion at higher frequencies, tests of amps have shown a rising distortion that is not present with other measurement software.
The next two plots are of the impedance curve of the woofers in the box (no crossover). They clearly show a problem at 600Hz. I have to try and dig up the free air measurements to see whether it is present in those too.
The third impedance plot is a zoom of the impedance plot of the finished speaker with the crossover in place. Oddly the 600Hz glitch is gone..... I'm wondering if it is only one of the two speakers exhibiting the behaviour...
Finally a couple of pictures of the actual speaker.
So with that as a starting point, I think that I can say that the problem is coming from the midbass units and that it is not related to the crossover.
Greebster commented in PM that it could be dimention related, though that was when I thought the impedance plot showed nothing... last pic is my original plan for the baffle dimensions.
Tony.
Attachments
I tracked down the free air impedance measurements. No sign of a problem at 600Hz. Bit of a wiggle around 1Khz though that is also present in the recent measurements.
So it looks like it is probably something to do with the enclosure (or something has been damaged since those initial tests back in 2005).
Tony.
So it looks like it is probably something to do with the enclosure (or something has been damaged since those initial tests back in 2005).
Tony.
Attachments
If you look at the free air measurement, the glitch is near 1200 Hz (look at the where the phase dips). When the two woofers are combined (no crossover), it becomes 600 Hz. I think it is driver related, maybe something to do with the motor. When you put the crossover in, it swaps it out or damps it, perhaps? There is still some action that can be seen in the phase curve of the third impedance plot in post 1.
ok I did some impedance measurements tonight. First pic is each MW144 in the left speaker measured separately no crossover.
Second pic is both running in parallel in the left speaker. No crossover.
Third pic shows mw144's without crossover, vs the complete speaker with crossover.
I need to pull a driver to verify, but it looks like the 600hz glitch has to do with the cabinet itself (assuming they still don't show it with a free air measurement).
Tony.
Second pic is both running in parallel in the left speaker. No crossover.
Third pic shows mw144's without crossover, vs the complete speaker with crossover.
I need to pull a driver to verify, but it looks like the 600hz glitch has to do with the cabinet itself (assuming they still don't show it with a free air measurement).
Tony.
Attachments
Last edited:
You might try loosening all of the woofer mounting screws a half turn or so to see if this changes the magnitude or frequency of the 600Hz glitch. Often I have found that the the woofer basket/magnet resonance feeds into the cabinet. This driver resonance won't show up in the unmounted driver impedance curve, but if you mount the driver basket firmly to a panel and do a frequency sweep with your hand lightly touching the back of the magnet you can feel it. An accelerometer stuck to the back of the magnet also will show the peak of this resonance.
If this is the source, you can damp the resonance by applying some viscoelastic adhesive around the basket to top plate joint.
Last edited:
Thanks bolserst. I'll give that a try. I actually have a different cabinet (original prototype) and a test baffle as well so could do some experimenting. Getting at the back would be difficult in final cabinets as they are completly sealed, but mounting on the test baffle would hopefully show the same issues..
Tony.
Tony.
My friend Pieskapov tought to me this. With REW, activate Signal Generator and RTA, click Frequency tracks cursor in Generator settings.
Play various frequencies around the distortion peak (move the cursor in RTA window), listen and feel with your hand the speaker to find a resonance. If you can't detect anything, it is perhaps the motor or something inside the cabinet. That amount of "distortion" seen in your graph, should be easy to find.
Play various frequencies around the distortion peak (move the cursor in RTA window), listen and feel with your hand the speaker to find a resonance. If you can't detect anything, it is perhaps the motor or something inside the cabinet. That amount of "distortion" seen in your graph, should be easy to find.
Thanks guys. I think I've worked it out. The loosening of the driver did actually reduce the level of the distortion, but interestinly I couldn't feel any resonances in the magnet when I mounted a speaker facing out.
What I did find though was that the distortion on the left speaker was much higher than on the right. I then tried the drivers individually and one is shocking and the other is not too bad. It turns out the one that is shocking is the one that I put a screwdriver through the suround and patched up with flexible super glue
The distortion is coming from the surround. If I lightly touch the surround not quite opposite the damaged side, the distortion vanishes.
Below are two plots of the speaker running in open air out of the box (mic a couple of meters away. First just the speaker, second lightly touching the surround.
When the driver is mounted in the box the distortion is amplified dramatically. Third pic shows the driver mounted in the box, and fourth is the one that hasn't had the screwdriver treatement...
So it looks like I could reduce my 600Hz problem by replacing the speaker I damaged There will still be a certain level of distortion around this frequency as it seems that the 1200Hz is surround related, but it will likely overall be much better. I suspect now that I know what I am listening for I may now be able to spot this. it is VERY obvious when I touch the surround and the harmonics reduce dramatically.
I'm wondering if I can do something to the surround on the opposite side to even things up, but I realise anything I do that isn't reversible may make it worse, so hesitant to try.
One last thing I found as well. One of the four nuts for this problematic driver is missing, the fourth bolt was just sitting loosly in the hole. That in itself could be a contributing factor as well, though taking out the bolt didn't seem to have any effect, however uneven clamping of the driver to the baffle may well have done, and may be why it improved (marginally) when I loosened the screws.
Tony.
What I did find though was that the distortion on the left speaker was much higher than on the right. I then tried the drivers individually and one is shocking and the other is not too bad. It turns out the one that is shocking is the one that I put a screwdriver through the suround and patched up with flexible super glue
The distortion is coming from the surround. If I lightly touch the surround not quite opposite the damaged side, the distortion vanishes.
Below are two plots of the speaker running in open air out of the box (mic a couple of meters away. First just the speaker, second lightly touching the surround.
When the driver is mounted in the box the distortion is amplified dramatically. Third pic shows the driver mounted in the box, and fourth is the one that hasn't had the screwdriver treatement...
So it looks like I could reduce my 600Hz problem by replacing the speaker I damaged
There will still be a certain level of distortion around this frequency as it seems that the 1200Hz is surround related, but it will likely overall be much better. I suspect now that I know what I am listening for I may now be able to spot this. it is VERY obvious when I touch the surround and the harmonics reduce dramatically. I'm wondering if I can do something to the surround on the opposite side to even things up, but I realise anything I do that isn't reversible may make it worse, so hesitant to try.
One last thing I found as well. One of the four nuts for this problematic driver is missing, the fourth bolt was just sitting loosly in the hole. That in itself could be a contributing factor as well, though taking out the bolt didn't seem to have any effect, however uneven clamping of the driver to the baffle may well have done, and may be why it improved (marginally) when I loosened the screws.
Tony.
Attachments
yes funny that eh? I'd actually completely forgotten about it since I did it nearly 8 years ago... when I found it was a single driver that was worse I remembered, and looked and sure enough it was the one. http://www.diyaudio.com/forums/multi-way/68153-prize-stupid-act-week-goes.html
Tony.
Tony.
I hope that there isn't a next time! I've just been having a close look at the surround, and I'm suspicious that this issue may have been present without the screwdriver incident. The place on the surround where gentle pressure stops the resonance appears to be not evenly attached to the cone. Everywhere else on the surround there is a nice even radius on the surround, but in this place it is almost like there is a slight concave crease and the surround seems slightly higher, as though when it was glued to the surround it was slightly off centre.
When I put it back in I want to try some strategically placed blu tac and see if I can suppress this surround resonance without any other nasty repercussions.
Tony.
I've just been having a close look at the surround, and I'm suspicious that this issue may have been present without the screwdriver incident. The place on the surround where gentle pressure stops the resonance appears to be not evenly attached to the cone. Everywhere else on the surround there is a nice even radius on the surround, but in this place it is almost like there is a slight concave crease and the surround seems slightly higher, as though when it was glued to the surround it was slightly off centre. When I put it back in I want to try some strategically placed blu tac and see if I can suppress this surround resonance without any other nasty repercussions.
Tony.
LOL Well yeah, we all know how that goes
Should have seen the look on a clients face one day when I was working on his new babies. One thing I did to the woofers was to vent the VC. This was easily done with a soldering pencil melting 6 holes in the PPB underneath the dust cap. Told him to make this puppy work it's best, one had to let out a bit of the manufactures magic smoke cuz it was overdamped.
Should have seen the look on a clients face one day when I was working on his new babies. One thing I did to the woofers was to vent the VC. This was easily done with a soldering pencil melting 6 holes in the PPB underneath the dust cap. Told him to make this puppy work it's best, one had to let out a bit of the manufactures magic smoke cuz it was overdamped.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.