The Pioneer 1 - Wired CbM Evaluation Platform provides a robust industrial wired link solution for the ADcmXL3021 Triaxial Vibration Sensor en/products/adcmxl3021.html. The ADcmXL3021 features ultra low noise density (26 μg/√Hz) and wide bandwidth to 10 kHz, which supports excellent measurement resolution and tracking of key vibration signatures. The EV-CbM-Pioneer1-1Z and EV-CbM-Pioneer1-2Z kits provide a complete plug and play solution for operating the ADcmXL3021 on an RS-485 network over meters of cable in harsh industrial environments. This wired evaluation platform is enables monitoring of industrial assets to improve up-time, accelerating the path to Industry 4.0.
Table 1. Solution Performance Trade Offs
Solution | Flexibility | PCB Area | Complexity | EMC |
---|---|---|---|---|
DEMO-CbM-Slave3-Z | Low | Low | Low | Medium |
DEMO-CbM-Slave2-Z | Low | Low | Low | High |
DEMO-CbM-Slave1-Z | High | Medium | Medium | High |
The following guidelines apply for the direct SPI to RS-485 link designs (DEMO-CbM-Slave2-Z and DEMO-CbM-Slave3-Z). The SPI to RS-485/RS-422 link designs include a SPI clock transfer over RS-422 (SCLK) and a power over data lines implementation (phantom power), where data and power share the same twisted pair (SPI MISO).
Cable Effects
Maximum SPI SCLK
Minimum SPI SCLK
Typical Performance
Figure 2. SPI SCLK vs. Cable Length Typical Performance
The following is supplied as part of the EV-CbM-Pioneer1-1Z demo kit:
The following is supplied as part of the EV-CbM-Pioneer1-2Z demo kit:
The following equipment is suggested as a vibration source, but not strictly required, as the system can also be tested manually:
The following steps describe a typical setup, as shown in Figure 3, to communicate over a SPI to RS-485 link using either DEMO-CbM-Slave2-Z or DEMO-CbM-Slave3-Z:
Figure 3. Typical Evaluation Setup
Figure 5. Hi-Rose Connector Orientation and S2 Pushbutton Switch
Two downloads are required, with all software available on the ADcmXL3021 wiki.
Figure 6. Typical Manual Time Capture Plot - corresponding to Figure 3 Typical Setup
Figure 7. Typical Manual FFT Plot - corresponding to Figure 3 Typical Setup
To provide a more easily visible Manual Time Capture (MTC) of a 'finger tap' vibration on the top surface of the ADcmXL3021, access the AVG_CNT register as shown in Figure 8. Write HEX 0x3 to AVG_CNT and perform a read back to confirm the write. This will decimate the signal so that the MTC waveform can be viewed easily. Figure 9 shows an example MTC measurement where a 'finger tap' vibration is applied to the top surface of the ADcmXL3021 module.
Figure 8. Write Hex 0x3 to the AVG_CNT Register to Decimate
Figure 9. Typical Manual Time Capture plot - Finger Tap Vibration Setup
Schematics, layout, BOM, and Gerber file for the DEMO-CbM-Slave2-Z: click here to download
Schematics, layout, BOM, and Gerber file for the DEMO-CbM-Slave3-Z: click here to download
Schematics, layout, BOM, and Gerber file for the DEMO-CbM-Master-Z: click here to download
Schematics, layout, BOM, and Gerber file for the DEMO-CbM-Expandr-Z: click here to download
A simplified power over data wires simulation circuit is provided in Figure 10. This circuit uses LTC2862 RS-485 transceiver LTspice macromodels and 1 mH inductors (Wurth 74477830). LTspice includes real inductor models, which include device parasitics, enabling closer correlation between simulation and real design performance. The DC blocking capacitor values are 10 µF. In general, using larger inductor and capacitor values enable a lower data rate performance on the communication network. The simulated test case is a 250 kHz data rate, which roughly corresponds to 100 meters of cabled communication when porting clock synchronised SPI over an RS-485 interface. The input voltage waveform used in the simulation corresponds to a worst-case dc content, with a 16-bit word and all logic high bits. Simulation results are presented in Figures 11 and 12. The input voltage waveform (VIN) matches the output at the remote powered device (no communication errors). Figure 12 presents a zoomed-in view of the bus voltage differential waveform (voltage A – voltage B) for droop analysis. The voltage at the remote sensor node, extracted from the L2 inductor (V(pout)) provides a power supply rail of 5V±1mV.
Figure 10. Engineered Power LTspice simulation circuit using LTC2862 (RS-485) and 1mH Wurth Inductor 74477830
Figure 11. Simulation Result with RS-485 bus differential voltage V(A,B) , and droop points X and Y
Figure 12. Droop analysis for point X and Y
The VDROOP, VPEAK, and TDROOP are measured using the Figure 11 and 12 LTspice waveform. The L and C values are then calculated using equations 2 and 4. It depends where you measure on the waveform, however, the calculated L value is 1 to 3mH as shown in Table 1. Measuring at point X (Figure 12) is most accurate and yields the correct inductance value of approximately 1 mH. The high pass filter frequency (equation 6) is simply a function of the droop time and voltage, and for point X is approximately equates to 250 kHz/32 for 1 bit (half clock cycle), which matches the input waveform (V3) shown in the Figure 10 schematic.
When simulating Figure 10 it is also worth noting that the C8 capacitor is recommended to reduce voltage overshoot at the sensor (Vpout at power extraction node). With C8 added the overshoot is maximum 47mV and settles to within 1mV of the desired 5VDC within 1.6ms. Simulating without a C8 capacitor results in an underdamped system, with 600mV overshoot, and a permanent 100mV of voltage oscillation from the 5V dc target.
LTSpice models are available here: click here to download
*July 2019. Initial Release.
*August 2019. Added additional information on demo power supply configuration.
*February 2021. Added LTspice models for engineered power.