In the face of these accelerometer measurements I'm taking, I'm using that graph to reiterate just how effective the null measurement is. Like hifijim said, you can knock yourself out trying to kill resonances that ultimately just aren't audible. A null measurement will actually tell what resonances are becoming airborn, along with it's relative loudness, and when it comes right down to it, all that ultimately matters is what we're hearing from the panel.
We can turn this idea around. We can take an accelerometer measurement of the center of an actual driver cone, and use it as an example of what were hearing from it, instead of taking a real mic measurement taken at a distance. No one would ever recommend doing that, but that's kind what were doing when we use the accelerometer technique on the side panel.
This accelerometer thing is fun though, and I am by no means done with it yet. It's accurate, and it works well for about $50.00 with the buffer included.
We can turn this idea around. We can take an accelerometer measurement of the center of an actual driver cone, and use it as an example of what were hearing from it, instead of taking a real mic measurement taken at a distance. No one would ever recommend doing that, but that's kind what were doing when we use the accelerometer technique on the side panel.
This accelerometer thing is fun though, and I am by no means done with it yet. It's accurate, and it works well for about $50.00 with the buffer included.
Last edited:
Except me. 🙂 Because that exactly what I did to calibrate the piezo pickup in both amplitude and FR. That way I know the relative levels of cone vs box.
I believe that both the contact method and the null method have their uses and merits. And perhaps they serve as a cross check for each other.
I will try it. The poly cone with a concave poly dust cap I just bought should be perfect for attempting this. The main problem I foresee though, is that the accelerometer "spl calibration" will only be valid relative to one fixed mic distance, preferably as close to the cone center(Like the accelerometer position) as possible. Once the mic gets further away from that close up position, the "relative calibration" method you're using is no longer valid.
Last edited:
+1... or maybe +2
The cancellation null technique is new. It requires verification. accelerometers and contact microphones are crucial in this process.
I think you should correlate the accelerometer to the acoustic response in two steps. Step 1 is to correlate the accelerometer to a near field FR, with the mic at 1/4" from the cone. Step 2 is to correlate the near field response to the 1 meter response. I have done the step 2 process in the past and it is very repeatable.
And that is exactly what I did. I used an average of several measurement points. It's all documented right here in this very thread.
I'm uncertain how to use the contact mic readings vs 1m mic readings. Surface area has to be taken into account. The idea is worth figuring out.
I'm uncertain how to use the contact mic readings vs 1m mic readings. Surface area has to be taken into account. The idea is worth figuring out.
Hello Pano,
What kind of preamp circuit are you using withe the piezo disk? And what is the piezo diameter? Piezo disks come in a wide range of varieties, hence my asking.
What kind of preamp circuit are you using withe the piezo disk? And what is the piezo diameter? Piezo disks come in a wide range of varieties, hence my asking.
Surface area is critical, which makes accelerometer measurements in one tiny fixed part on the side panel so difficult to interpret relative to the output of the entire side panel. It's kind of like the old saying about looking for our lost keys only under the street light, because that's where the light is, except for the next light down the street, where we'll look for the keys again. The null method is kind of analogous to just waiting for daylight so you can see everywhere.
Last edited:
The disc is approximately 24mm. I have been using the instrument input of my Behringer UMC204 which has an impedance of 1 meg ohm. I have also used them with a Zoom recorder but it has lower input impedance, so less bass. The DI that Remlab found looks like a great idea to buffer mics like this.
I will do the cone accelerometer measurements vs nearfield and 1m today. We can compare data if you still have it.
If you say so. 🙂 You are obviously a big fan of the method and that's understandable. Sometimes a floodlight is needed, sometimes a spotlight. Just depends on what you are looking for. It's nice to have more than one tool in the toolbox.
I agree with needing more than one tool. Like I said, I am in the process of replicating your cone accelerometer method. I'll see what I can come up with.
I look forward to seeing what you get. I placed the measurement mic as close as possible to the piezo on the cone, and then at 3 or 4 or positions around the cone for an average. I didn't want that one spot to totally define the piezo vs cone response, tho maybe it should.
Anyhow, once you match levels then you'll know what the contact mic is giving you. From that the box wall surface area can be calculated to know how much is radiated. Knowing the very close mic/piezo level vs 1m levels will be handy, too. But surface area will still be important.
Anyhow, once you match levels then you'll know what the contact mic is giving you. From that the box wall surface area can be calculated to know how much is radiated. Knowing the very close mic/piezo level vs 1m levels will be handy, too. But surface area will still be important.
Okay, I have graph with pictures. First the graph. Remember that the linearity of my accelerometer, buffer system is extremely flat from 50hz and up which was shown in previous graphs.
Orange- Mic at 1"
Green- Accelerometer on cone with extremely thin double sided tape. I also made sure that only the accelerometer is touching the driver. The wires are being delicately held away.
I will obviously take any advice if I'm doing something wrong🙂
The last graph is signal vs noise on the cone accelerometer measurement, to show that noise is not effecting the measurement. I had to do the wires unshielded so diaphragm movement would be unimpeded.
Orange- Mic at 1"
Green- Accelerometer on cone with extremely thin double sided tape. I also made sure that only the accelerometer is touching the driver. The wires are being delicately held away.
I will obviously take any advice if I'm doing something wrong🙂
The last graph is signal vs noise on the cone accelerometer measurement, to show that noise is not effecting the measurement. I had to do the wires unshielded so diaphragm movement would be unimpeded.
Attachments
Last edited:
Thanks for the measurements. That looks very different from what I got. I did mine on the cone proper, but I don't think it would matter and long as it is well stuck. Looks like there is something hinky at 4.8K - may be best to ignore that, or smooth it away.
I'm not sure about theses piezos, but regular mic don't act like that, there aren't any super sharp peaks or nulls, so they can be ignored in calibration.
I'm not sure about theses piezos, but regular mic don't act like that, there aren't any super sharp peaks or nulls, so they can be ignored in calibration.
Linearity of my accelerometer taped to the cld speaker side panel. This is with the unshielded cable, so noise makes a small appearance at 60hz. I would have never dreamed that an accelerometer/buffer would be this linear. The second graph is a csd of the same measurement
Attachments
Last edited:
Jim, most of what I did starts in post 201. Seems I did not post my uncalibrated peizo on the cone measurement. I'll have to find it or just do another. I want to compare it with what Remlab is getting from his sensor.
Would it be an idea to use several piezo mics on one side panel, if some vibrate out of phase they'll cancel out ( electricly ) the in phase ones, just like the out of phase panel vibrations would cancel out the sound of the in phase ones. Also they could be wired to better match the impedence of the amp.