Group Actions
Engagement
  • Author Author: mvalencia32
  • Date Created: Date Created
  • Views 3520 views
  • Likes 5 likes
  • Comments 8 comments
Related
Recommended

Wilcoxon PC420V vibration sensor with a Raspberry Pi 5

mvalencia32
Wilcoxon PC420V vibration sensor with a Raspberry Pi 5

Vibration analysis is important for any industrial process, as it provides us with a lot of information about the mechanical state of the machines we use. It also allows us to use their signals as interlocks in PLC programming, etc. Therefore, one of my motivations for creating this small blog was to use a Wilcoxon sensor and adapt it to a Raspberry Pi with the idea of being able to view RMS measurements on the screen and log them in a .csv file.

Table of specifications:

Parameter Specification
Sensitivity 100 mV/g
Measurement Range 0 to 0.5 ips (inches per second)
Frequency Response 2.5 to 15,000 Hz (±3 dB)
Resonance Frequency >25 kHz
Transverse Sensitivity <5%
Power Requirement 18-30 VDC
Output Signal 4-20 mA
Operating Temperature -40 to 85 °C
Case Material Stainless Steel
Weight 145 grams (5.1 oz)
Mounting 1/4-28 UNF
Connector 2-pin MIL-C-5015

Pics:

Image 1; 

image

Project Comments:

  1. The Raspberry Pi 5 does not have an ADC, so I had to adapt an ADS1115 (I2C), which has a 16-bit ADC. (Image 2)
  2. The Wilcoxon sensor operates at 12 volts and has a current output of 4 to 20mA. Its datasheet provides a diagram of how to integrate it with a PLC, and using that idea, it is possible to integrate it with the Raspberry Pi 5. (Image 3)

Image 2

image

Image 3

image

Before processing the vibration signal, calibration tests were conducted between the ADS1115 and a reliable multimeter.

imageimage

Example Code to Anaconda Rsbaerry pi5:

import time
import math
import csv
import Adafruit_ADS1x15
import tkinter as tk
from matplotlib.figure import Figure
from matplotlib.backends.backend_tkagg import FigureCanvasTkAgg

adc = Adafruit_ADS1x15.ADS1115(address=0x48, busnum=1)
GAIN = 1
DATA_RATE = 860
V_REF = 4.096
MAX_ADC_VALUE = 32767
RESISTANCE = 120

root = tk.Tk()
root.title("Lectura del ADC ADS1115")
root.geometry("800x600")
root.attributes('-fullscreen', True)

speed_label = tk.Label(root, text="Velocidad (mm/s):", font=('Helvetica', 16))
speed_label.pack(pady=10)

fig = Figure(figsize=(8, 6), dpi=100)
ax = fig.add_subplot(111)
ax.set_title("Tendencia de la Velocidad")
ax.set_xlabel("Tiempo (s)")
ax.set_ylabel("Velocidad (mm/s)")
ax.set_ylim(0, 12)
line, = ax.plot([], [], 'b-')

canvas = FigureCanvasTkAgg(fig, master=root)
canvas.get_tk_widget().pack(fill=tk.BOTH, expand=True)

speed_values = []
time_values = []

start_time = time.time()

csv_file = open('vibration_data.csv', 'w', newline='')
csv_writer = csv.writer(csv_file, delimiter=';')
csv_writer.writerow(['Timestamp', 'Speed (mm/s)'])

def calculate_rms(voltage_values):
square_sum = sum(v ** 2 for v in voltage_values)
rms = math.sqrt(square_sum / len(voltage_values))
return rms

def current_to_speed(current_ma):
if current_ma < 4:
return 0
elif current_ma > 20:
return 12
else:
return ((current_ma - 4) / 16) * 12

def update_values():
voltage_samples = []

for _ in range(86):
adc_value = adc.read_adc(0, gain=GAIN, data_rate=DATA_RATE)
voltage = (adc_value / MAX_ADC_VALUE) * V_REF
voltage_samples.append(voltage)
time.sleep(1 / DATA_RATE)

rms_voltage = calculate_rms(voltage_samples)

current_ma = (rms_voltage / RESISTANCE) * 1000

speed_mm_sec = current_to_speed(current_ma)

speed_label.config(text=f"Velocidad (mm/s): {speed_mm_sec:.6f}")

current_time = time.time() - start_time
speed_values.append(speed_mm_sec)
time_values.append(current_time)

csv_writer.writerow([current_time, speed_mm_sec])

if len(speed_values) > 100:
speed_values.pop(0)
time_values.pop(0)

line.set_data(time_values, speed_values)
ax.set_xlim(min(time_values), max(time_values))
canvas.draw()

root.after(100, update_values)

update_values()
root.mainloop()
csv_file.close()

Results:

Image 2

image

Rasberry Pi Screen:

image

I hope you found this blog entertaining.

Regards

Martin

Parents

  • My first thought was:

    If you use the 'Insert' then 'code' functions you may be able to preserve the indentation on your code, without which Python is rather hard to follow.

        freq = list(range(101))
    ampl = list(range(101))
    
    for step in range(0,101):
        message = ":SOUR2:APPL:SIN {:.2f},5,0,0\n".format(fs)
        siggen.sendall(bytes(message, 'UTF-8'))
        freq[step] = fs
        #print("Frequency: {:.2f} \t".format(fs))
        print(":SOUR2:APPL:SIN {:.2f},5,0,0\n".format(fs))
        fs = fs * sm
        time.sleep(1)
        message = "MEAS:VOLT:DC? AUTO\n"
        hpdmm.sendall(bytes(message, 'UTF-8'))
        rpy =  (hpdmm.recv(1024))
        res = float(rpy)
        ampl[step] = res
        print ("DMM reading {:.4f}\t".format(res))
          
    hpdmm.close()
    siggen.close()

    It still looks wrong in the window while your are editing but it looks OK when posted..

    (The actual code is just the first example with formatting I found - not relevant to your blog !)

    The ADS 1115 isn't the ideal ADC for this work. It has a very nasty frequency response curve (page 12 of the data sheet) so at 860 s/s the response is about -3dB @ 400Hz but only - 18dB at 2kHz. (figures obtained by scaling from the graph on page 12).

    So you need an anti aliasing filter because out of band signals will be aliased back into the low frequency range.

    You could get some filtering in your differential amplifier..

    I didn't understand your current to speed function - the sensor is an accelerometer so I feel that there should be a  double integrator somewhere but I didn't spot it.

    MK

Comment

  • My first thought was:

    If you use the 'Insert' then 'code' functions you may be able to preserve the indentation on your code, without which Python is rather hard to follow.

        freq = list(range(101))
    ampl = list(range(101))
    
    for step in range(0,101):
        message = ":SOUR2:APPL:SIN {:.2f},5,0,0\n".format(fs)
        siggen.sendall(bytes(message, 'UTF-8'))
        freq[step] = fs
        #print("Frequency: {:.2f} \t".format(fs))
        print(":SOUR2:APPL:SIN {:.2f},5,0,0\n".format(fs))
        fs = fs * sm
        time.sleep(1)
        message = "MEAS:VOLT:DC? AUTO\n"
        hpdmm.sendall(bytes(message, 'UTF-8'))
        rpy =  (hpdmm.recv(1024))
        res = float(rpy)
        ampl[step] = res
        print ("DMM reading {:.4f}\t".format(res))
          
    hpdmm.close()
    siggen.close()

    It still looks wrong in the window while your are editing but it looks OK when posted..

    (The actual code is just the first example with formatting I found - not relevant to your blog !)

    The ADS 1115 isn't the ideal ADC for this work. It has a very nasty frequency response curve (page 12 of the data sheet) so at 860 s/s the response is about -3dB @ 400Hz but only - 18dB at 2kHz. (figures obtained by scaling from the graph on page 12).

    So you need an anti aliasing filter because out of band signals will be aliased back into the low frequency range.

    You could get some filtering in your differential amplifier..

    I didn't understand your current to speed function - the sensor is an accelerometer so I feel that there should be a  double integrator somewhere but I didn't spot it.

    MK

Children