I currently have a hobby king DJI clone framed quad, but plan on designing a custom frame in the near future. One key design feature of this quad frame will be that the motors' vibrations are to be isolated from the rest of the frame, rather than having only the APM isolated from the frame. I plan on using the same motors/props, so I was going to find a spectrum on my current quad using my smartphone and a vibration app, but I'm not sure the app acquires data fast enough to give accurate resolution for the high frequency of the props. I know I've seen people who have done vib analyses on their copters, so I was wondering what people have done in the past that has worked. Any ideas?
Hard to do without instrumentation!
One could get dampers of differing durometer and see which one yields the most solid image?
Prop/Rotor balancing is obviously important
I definitely want to instrument the quad to get the vib data. Just not sure what to use. I was really excited when I found the android app, but it seems to top out at 43 fps. Do you know if ArduIMU has the resolution to effectively acquire vibration data in the high frequencies we see in quads? I would try with my Ardupilot but it's isolated in the current config, and I want to gather data with a separate non-isolated system over the course of a flight to get a good spectrum.
Honestly, my quad flies great as is. I balance my props and motors and the APM is isolated. I've never had any issues with vibrations. I'm just trying something new and hoping to learn something about vibration analysis and noise filtration.
We use some very 'spensive equipment to measure frequencies and modes..
not sure there is anything available to the hobbiest....
see if there is a university nearby with suitable research equipment...
I found a different android app called "speedy spectrum analyzer" that does FFT based on the noise from your microphone instead of the accelerometers. It costs $5, and I'm sure there's a free one if you do some looking, but the real point is that it at least gave me some useful frequency data.
I'm going to try to design the filter based on what I have, then bench test the new design against the old design with an arduino (nano, probably) and an IMU to validate. I'm sure the code to get the accelerometer data from the IMU is pretty simple, but I'm also sure it's already out there (and I don't have any experience coding for Arduino). Anyone have a simple code solely for recording the accelerometer data and time?
Fully with you in damping the vibration nearest to its source.
I would google for "sound card spectrum analyzer" and mechanically attach a light microphone to the frame. That would be cheap and will give you a baseline to start tuning the mounts. I have not been using those for years since I do have an oscilloscope with basic spectrum analyzer in it.
One example: http://zelscope.com/
good idea, but the micro will have as input the vibration and sound from motors. when you increase the throttle, then the motor noise will be harder and vibration also.how can you make a distinctions between one of those inout source?
same thing you can do with the windows sound recorder.
Open your laptop leed completely, put the Quad on the laptop, start sound recorder, start throtle, look at the oscillations. Or as you said, put an external computer micro on your quad and start props (before take off) and look at oscillations. Maybe try to put the micro on each motor position and repeat the test 4 (6-8) times for each motor and compare the results.
There is a new App on the iPad app store for full vibration spectrum analysis. There is also a app for full balancing of rotors. Just google iPad vibration analysis or iPad precision balancing. It comes with external acceleromtor and tachometer.