To make a test like that among others is easy and quick, informative. A record? Having a long list of topics, I prefer to keep time for them. Sorry
Christian
Hello Eric,
I have had your question in mind but not the time to make a test. I made one today. On the graph, the impedance curve of the original post and the tap test from today (not the material to make an impedance test today). The exciter was removed from the membrane since the impedance curve.
Due to the height of the impedance peaks, I would have expected sharper peaks on the tap tests.
+ @xsuper9988
Here it is no? #6681
And here is the spectrum where the 3k bump on one side is visible #6682
Hello Steve,
Nice to read you!
Thank for sharing a new time this list.
I would split those "noises" in 2 categories :
- The noises for which the source is by somehow external to exciter : wires, extra strength on the exciter due to weight, loose nuts... All those things are quite easy to detect and correct (spine, wire routing, tighten screws...) even if ennoying at the beginning (I remember spending some times after a noise to find at the end a loose nut).
- The noises linked to the exciter design by itself (cavity, vent effect). By nature, they are more like resonators as they emit sound at a specific frequency (or small range) when excited.
- Green : The small piece of EPS as a reference. On it, the cavity noise at 2.3k is visible (the spectrogam shows it is not coming from the front side)
- Orange : A piece of the same dimension but with a hole in regard of the voice coil area.. No more cavity noise but no more HF reinforcement coming from the central area
- Blue : The third curve is a piece of the same EPS, about 25mm diameter just covering the voice coil. The cavity noise and the emission of the central area if visible. Even if it not exactly the same frequencies because the conditions are not exactly the same (boundary conditions), it could be a not too bad test to visualize those noises.
So avoiding the cavity noise by a hole modify also the HF.
If I remember, Steve, you have also tested that then searching solution to get back HF (tea bag dome?)
Against or to control the central area emission, some blue tack added at the center.
We could also list the workaround from @Eucyblues99 : the Eucy's dome and the felt on the exciter.
Just estimated but the roll off rate in HF of the orange curve seems 6dB/oct. Interesting because it opens to the possibility to find the cause, to compensate, or why not to add a tweeter (even if for this material it happens too early).
Please don't ask about the noise at 50Hz (stil not investigated, trouble with my mike?).
Christian,
Interesting. Typically, the tap test peaks are sharper.
What is the panel material, dimensions, and suspension details? Where was the exciter placed? Is it a material for which you have estimated the elastic moduli?
Have you tried any impedance measurements with the exciter in other positions?
Eric
And Christian.
Cone drivers have their own set of problems.
With dml it all depends on the exciter panel combinations.
For instance , the crate ply is very sensitive to exciter noise.
Which can be controlled by some small weights in certain positions.
But the Round flueted proplex panels do not seem suffer from panel exciter noise from the front surface.
The proplex has a naturally flat response from 40hz to 20k.
The panel exciter resonance in the 40hz region on my exciters also changes from panel to panel ,it can also move upwards in frequency when using weights.
I have never found the need to use dsp to alter the frequency response.
If the response is that bad , I would just throw the panel in the bin, and find a better sounding panel material .
Unless of course it sounded exceptionally good 😃
Steve.
Hello Eric
It is the panel described in the pdf of #10187
3mm corrugated polypropylene. Have a look to the pictures for the shape "JAlike".
For that tap test, the exciter was placed.
I should have somewhere results of tap tests done to evaluate the material characteristics.
Do you want to run a simulation?
Christian
Thanks Christian,
I was thinking about doing a simulation yes. But that shape and material just make it a little too complex for now. It would probably be better if I learned how to do my own impedance measurements. Then I could make my own comparisons between tap test resonances and impedance peaks.
Eric
Hello Eric,
Years ago when I built standard loudspeakers, impedance measurement was among my tools for the crossover design. Starting with DML, as it operates mainly in full range I was not seeing the interest. But digging the theoretical approach, I see it informative for the DML too. With REW it is accessible (see the extract below from : REW impedance measurement)
I just followed it. Rsense = 100Ohms. The signal comes from a headphone output of a cheap USB DAC (headphone output for its capability to drive low impedance). The signals are measured by the line input (R&L) of an other USB card. You just need an additional resistor of a known value (ie 10Ohm) for calibration.
I wired all that in a small platic box with speaker type terminals for robustness and easy connections.
Jaxboy.
Thanks 😊
I sent you a pm with a video of my garden and sounds.
Hope you enjoy.
I'm only putting a word or two in, as I have not been in my room for a while, I'm having a nice break from dmls.
Although I do have a few projects I intend to pursue at sometime.
My friend also finally lent me his tectonic bmr units to measure when I get time.
Steve.
Thanks 😊
I sent you a pm with a video of my garden and sounds.
Hope you enjoy.
I'm only putting a word or two in, as I have not been in my room for a while, I'm having a nice break from dmls.
Although I do have a few projects I intend to pursue at sometime.
My friend also finally lent me his tectonic bmr units to measure when I get time.
Steve.
Hello Audiofrenzy,Damn what happened to all the gurus they just up and left, no dedication.
For the past 8 years or so I've been experimenting with 5.1 channels for music. The rear channels are a problem as they interfere with the front channels as they are all playing the same signal and there seems to be cross talk cancellations. With movies using Dolby surround there is no problems with the rear channels because they are not using the same signal. So instead of using the rear channels for music I only use the front channels which include L/R main channels as well as the center channel. If you thought 2 channel center phantom imaging was good wait till you add a dedicated DML center channel it will bring the singer more to life as if she was actually there in the middle singing.
I use this soundtrack to test the center DML channel. If you thought 2 channel was good wait to you add a center DML panel and wow you will be amazed as the center channel just enhances the stereo left and right mains.
Do you use a track directly from YT? Which format?