Project Logs

Collapse

• First Measurements: Progress, Challenges, and Next Steps

  brokenbrokenpotatoes • 12/11/2024 at 15:36 • 0 comments

  Introduction

  This update covers the latest developments and challenges in the project, focusing on testing the strain gauge setup, investigating noise issues, and identifying areas for improvement. The testing process revealed critical insights into the power supply noise originating from the Pi Pico and prompted several modifications to improve the system’s performance. Additionally, the software has reached a basic functional stage, and plans for the next steps are outlined.

  Update Log

  Testing the Strain Gauge

  To test the strain gauge, I attached it to an aluminum bar using the following procedure:

  1. Surface Preparation: Sanded the aluminum surface with 800-grit sandpaper to ensure smoothness and proper adhesion.
  2. Cleaning: Cleaned and degreased the surface thoroughly with acetone to remove any contaminants.
  3. Wire Stress Relief: Soldered the wires to an attachment strip to relieve mechanical stress from the wires, preventing damage to the strain gauge.
  4. Final Cleaning: Cleaned the setup again with acetone to remove any soldering residue.
  5. Environmental Protection: Covered the strain gauge in a layer of silicone to protect it from environmental factors. For shaping the silicone, I used soapy water on my fingers to smooth the layer.

  Initial Test Results

  After setting up the strain gauge, I managed to obtain a functional measurement from the system. However, the data quality was not as high as desired, prompting further investigation into the sources of noise.

  Identifying the Noise Source

  1. Ripple on USB Power:
     • With the Pi Pico connected and all other power supplies on the board turned off, the ripple on the USB input measured 300mV peak-to-peak.
     • Disconnecting the Pi Pico reduced the ripple to 80mV, clearly identifying the Pi Pico as the primary source of the noise.
  2. Testing Modifications:
     • As a temporary test, I bridged the analog power supply and the Pi Pico's power supply. This modification reduced the ripple on the USB input and the 3.3V line to 100mV peak-to-peak, a notable improvement but not a permanent solution.

  Root Cause Investigation

  The noise likely originates from the Pi Pico’s onboard buck converter, possibly due to a cracked ceramic capacitor or inherent design inefficiencies. These findings emphasize the need for further testing and improved noise mitigation strategies.

  Planned Solutions

  • Proper Filtering: Implement CLC filters in the next PCB revision to comprehensively address power supply noise.
  • Pi Pico Evaluation: Test a different Pi Pico to determine if the issue is specific to the current board or inherent to the design.

  Software Development

  The basic software is now functional, outputting collected data in CSV format over USB. Work continues to refine the software for additional features and improved usability.

  Next Steps

  • Replace Pi Pico: Order a new Pi Pico to evaluate its noise characteristics and determine if the issue is specific to the current board.
  • Hardware Enhancements: Introduce proper filtering in the next PCB revision to address noise comprehensively.
  • Software Refinement: Continue improving the software, including advanced features like data compression and RS-485 communication.

  Summary

  This round of testing has provided valuable insights into the sources of noise and potential solutions. The temporary bridging of power supplies reduced ripple from 300mV to 100mV, but this is not a viable long-term fix. By implementing proper filtering and evaluating other Pi Pico boards, I aim to significantly improve the system’s performance and reliability. Despite these challenges, obtaining a functional measurement marks a critical step forward in this project.

• First PCB Revision

  brokenbrokenpotatoes • 12/01/2024 at 21:21 • 0 comments

  I'm excited to share a project I've been working on, a compact 8-channel strain gauge data acquisition system designed for use in experimental rockets. The goal of this project is to measure the loads and forces acting on the rocket during flight, and all of this is done using a Raspberry Pi Pico-based design.

  The Challenge

  The rocket that this system will be integrated into has a 100mm diameter, which means the design needs to be as compact and efficient as possible. We need a system that fits within these tight size constraints while providing high-precision measurements from multiple strain gauges, all while minimizing power consumption and weight.

  Key Features of the System

  • 8 Channel Measurement: The system is capable of monitoring up to 8 strain gauges (or other resistance-based sensors), making it ideal for multi-point force measurement.
  • Texas Instruments 24-Bit ADC: The heart of the system is an 8-channel, 24-bit ADC with adjustable gain (1 to 128) and a sampling rate of up to 32 kSPS. This gives us the precision needed to capture small changes in resistance from strain gauges.
  • Excitation Voltage: It provides up to 9V shared excitation voltage for all channels (250mA), which can be adjusted by soldering resistors onto the PCB. Future revisions are planned to support 10V excitation by using a higher voltage non-USB power supply.
  • USB Power and Communication: The system is powered via USB, and data is transmitted through the same connection for simplicity and ease of use. In the future, RS-485 support will allow multiple boards to communicate with each other for more complex setups.
  • Low-Noise Power Supply: For the power supply, I used the LM27762 Low-Noise Positive and Negative Output Integrated Charge Pump Plus LDO. This setup provides a bipolar power supply to get the nominal voltage output from the Wheatstone bridge around ground level, which is essential for ensuring that the output falls within the input range of the ADC.

  Compactness: A Critical Design Goal

  The current PCB design is 65mm x 45mm, but the aim is to shrink the system even further in future revisions to fit into the tight constraints of the rocket. By using careful component selection, efficient layout techniques, and a focus on simplicity, this system manages to pack a lot of functionality into a small package.

  Design Process and Tools Used

  The PCB was designed using Altium Designer. The project uses only components from trusted and established manufacturers to ensure high quality and reliability, which is critical for flight applications. The system also uses a mix of through-hole and SMD components, making it both easy to assemble and cost-effective.

  First PCB Assembly

  For the initial build, I decided to use the SMD stencil reflow method to solder the components onto the PCB. This process worked well overall, but I encountered a couple of challenges along the way:

  • I accidentally ordered the wrong LDO (Low Dropout Regulator) for the power supply, so I had to hack another LDO onto the board to get it working. While not ideal, it was a good learning experience.
  • I also found out that the Raspberry Pi Pico 2 and the original Raspberry Pi Pico are not exactly pin compatible. When I designed the PCB, I was working under the assumption that both models were pin-compatible. However, after soldering a Pi Pico 2 onto the board, I realized that GPOUT1 is not available on pin 13 of the Pi Pico 2, which caused an issue. This was an unexpected surprise, and I’ll need to adjust for that in future revisions.

  Images of the Soldered PCB:

  What's Next?

  • Smaller Form Factor: We plan to make future revisions even smaller to ensure the system fits within the rocket's constraints.
  • Excitation Voltage Control: Future versions will feature a DAC for more precise control over the excitation voltage.
  • Networking Multiple Boards: We're looking at adding RS-485 for networking multiple boards together, which would allow us to monitor more channels and share data...

  Read more »

View all 2 project logs