Best Vibrational Data Collection Sensors

Hardware
1

Hi, I am creating a vibrational data acquisition system powered by an Arduino uno to attach to my 3d printer. This system will collect vibrational data from the system as it runs using IMU's (IMU: accelerometer that also includes gyroscope). It should be able to identify print faults due to vibration from the printer motor, loose print belt, etc.

I am currently using mpu 6050 IMU accelerometers. They work but sensitivity and precision are not high enough to separate small vibrations from the movement of the printer.

Any recommendations on high performance, high sensitivity IMU sensors that are compatible with Arduino and would be good for this use case?

2

Do you really need an IMU, or will just an accelerometer be sufficient?

3

It could be that the method you are using to analyze the vibration spectrum is not working, in which case a different accelerometer won't help.

What are you doing to analyze the data?

1 Like
4

ML? :thinking:

5

IMU is preferred. It's just accelerometer + gyro. It provides a bit more data and insight.

6

The eventual plan for analysis is signal filtering to eliminate low freq. motion and very high freq. noise. Then Fast fourier transform of results to try to isolate vibrational frequencies.

7

Eventual post-processing use case for the data is ML related. But I'm just trying to collect useful data at this point.

8

On what basis did you decide that the accelerometer "sensitivity and precision" is inadequate?

If you want informed suggestions, you really do need to provide some basic information.

As an aside, you do not need high pass filtering before taking the Fourier transform, but you do need to ensure that no frequency in the input signal exceeds (sample frequency)/2. For that purpose, adjustable bandwidth low pass filters are included in most modern accelerometers.

1 Like
9

The sensitivity must be good enough to distinguish system vibration from the overall movement of the printer. You can imagine how much a 3D printer may vibrate when its motor is running and the print head is moving. We need to be able to capture this oscillatory vibration and separate it from the general motion of the printer (accelerometer will be mounted on print head).

I'm not really sure the exact sensitivity and precision specs needed. The current sensors have an error of about 0.2 g's when stationary which is a big issue.

10

No, they don't. All accelerometers measure the acceleration due to gravity, as well as those due to other forces. The value of +/- g is reported on any vertical axis, and some component of g will appear along any axis that is not perfectly horizontal.

There are small constant offsets along each of the axes that can easily be measured and removed by calibration. This excellent, comprehensive tutorial explains how: Tutorial: How to calibrate a compass (and accelerometer) with Arduino | Underwater Arduino Data Loggers

The sensitivity must be good enough to distinguish system vibration from the overall movement of the printer.

When you finally learn how to use a MEMS accelerometer properly, you will find that most, if not all the modern ones are capable of doing what you want.

11

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.