DiyMobileAudio.com Car Stereo Forum banner

1 - 5 of 5 Posts

·
Registered
Joined
·
18,416 Posts
Discussion Starter #1
Long story short, im impatient and curious and want to 150% make sure my product stacks up above the rest. I know Chris is doing his own testing, but idk when that will be done. On top of that, I'm also curious to see how his rig compares to say, an actual car door. Combine all of that with Matei moving back to NY and working at the shop again, I'll have some time free'd up to do this. That being said, I plan on setting up the Nissan Versa's rear door as a testing subject. My plan is not going to be nearly as intricate and flexible as chris's setup, but it will give an idea on how its actually behaving in the exact environment we use it in.. usually car doors. Car doors are not going to have the same natural resonance/response as a relatively small and flat panel that is bolted to a box. So here is my plan..

2 good contact microphones ( Metal Marshmallow Piezo Disc Contact Mic And Preamp ) will be used to gather measurements. You can read the description on this page, but the long story short is that contact microphones ONLY measure the vibrations of the substrate that they are mounted to, and ignore all sound that is traveling through the air. This is PERFECT to gather the info we need to know. These mics in particular are flat from 10hz all the way up to 20k. That said, One will be placed right behind the speaker (which will probably be a spare GB60), and one will be placed dead center of the door. This way we can get a better idea of what is happening across the entire panel. They will be mounted to the inside of the outer door skin (where you would normally install CLD). These models have their own built in preamp, so I can measure right from the analog input of the laptop. The inner door skin and will be FULLY sealed and treated before all testing. I only want results from the outer skin. The CLD being tested will be stuck to the outside of the outer door skin. Thats right.. right on the paint. The reason for this is so I do not have to touch the speaker or the inner door skin and can get consistent results, and to change products relatively quickly and easily. I am thinking of covering the door in painters tape so I don't potentially ruin the paint, but I will have to test to see if that changes the results of whatever is being measured. REW will be used to take generate sweeps and take measurements. The only thing I'm not sure of is if I should go directly into the laptop with the mic, or go into another preamp so I can do impulse response. Im not sure that's necessary considering I can still do waterfall graphs, but impulse will still be a halfway useful measurement.

TLDR: speaker installed in oem door location, fully treated inner skin while taking measurements of outer skin, contact mics permanently installed to the inside of the outer skin, deadener applied to outside of the outer skin directly to paint. All of it will be filmed to prevent the possibility of fudging results. Measured with REW to get frequency response, waterfall graph, and possibly impulse response.

Any questions, ideas, or critiques?
 

·
Registered
Joined
·
1,753 Posts
Baseline test with the blue painters tape on but no CLD applied of course. Would it be worth measuring in vehicle db measurements to see if there would be any significant ambient in car reduction in sound even though CLD isn’t designed to block sound.

FYI, blue tape has a specific release time so if you leave it on too long between testing then you could end up in a sticky situation.

A junk yard door would make a good test alternative if the tape doesn’t work out. If I had one I would give it to you for free.
 

·
Registered
Joined
·
18,416 Posts
Discussion Starter #3
Baseline test with the blue painters tape on but no CLD applied of course. Would it be worth measuring in vehicle db measurements to see if there would be any significant ambient in car reduction in sound even though CLD isn’t designed to block sound.

FYI, blue tape has a specific release time so if you leave it on too long between testing then you could end up in a sticky situation.

A junk yard door would make a good test alternative if the tape doesn’t work out. If I had one I would give it to you for free.
Yup, each run would include bare door, tape, and tape with cld IF i even decide i care about the paint lol. I'd use the green 3m tape though. Thats what we use here.

Sent from my SM-G975U using Tapatalk
 

·
Registered
Joined
·
177 Posts
I saw your post over on carav randomly. I think diyma might be a better spot for this kind of thread. Anyways, here's a copy/paste of my thoughts:

I have a few suggestions but maybe I should describe how someone would typically go about this test professionally. And when I say professional, this is a very basic explanation as this isn't my area of expertise.

To evaluate the performance of a CLD application an NVH engineer would conduct a model (vibration) test of the panel w/o treatments and repeat the test with your product installed. Model testing involves impacting the panel with an instrumented hammer (measuring the input) and then measuring the response of the panel with accelerometers. This is called a frequency response function (FRF) measurement.

The number of impact locations and accelerometer positions depends on the frequency response you're concerned with. Ever look at chladni patterns? What happens if your accelerometer is positioned at a null for a particular frequency?

I'll skip other aspects of the data analysis but you can find some pretty pictures online if you google.

If you actually want to go through with this, my suggestion would be the following:

1 . Pick a number of points on the outer door skin to measure with your contact mics. No idea if these mics are worth a shit but I'm not aware of cheap accelerometers that you could use.

2 . Play white noise from your speaker. If you're curious about why white noise is played, think about what the time domain and frequency domain signals looks like. I'm trying to think about why using a sweep could be useful. Your waterfall plots will also include other resonances from the input (speaker) which could muddy that data. However, this would also be consistent between test conditions.

You may also have to mic up the speaker to see what the frequency response of the driver is. If you do not know the input then what are you measuring on the door panel? Maybe a near-field measurement would be the way to go here? You might also have to stick the mic inside the door if you're sealing it up. This is pretty janky.

3 . Measure the response of the panel with the contact microphone at the chosen points.

4 . Apply CLD to panel and repeat measurements at the same points with the same input signal.

You suggested using painter's tape to prevent damage to your car. I would assume that the tape would negatively impact the performance the CLD. If this is just a baseline vs dampened test, I suggest installing the CLD to the inside of the outer panel and attach the contact microphones to the exterior of the panel.
 

·
New CAJ Member!
Joined
·
437 Posts
Looking forward to seeing this project go forward. I've been a little disappointed by the lack of progress on the other testing thread.
 
1 - 5 of 5 Posts
Top