The Industrial Raspberry Pi Forum

Posted: **30 Jan 2025, 01:35**

I am investigating the use of the RevPi DIO encoder functionality to capture position information from a quadrature encoder. During my investigations I have found that when moving forward the encoder counts at twice the rate as when moving backwards (for the same frequency signal). Is this behaviour expected? Any ideas on how I can change this?

To test encoder functionality I am using a signal generator to simulate encoder wave forms, varying the phase of these signals to simulate moving forwards and backwards.

Code to capture output data
I have also confirmed the same behaviour using piTest.

Code: Select all

import time

# Testing an input output loop, with the output 1 wire connected to input 1
# PiCtory configuration needs to be set appropriately
# with OutputValue_1 and InputValue_1 set to 0-5V
import revpimodio2

# Initialize the RevPi communication
rpi = revpimodio2.RevPiModIO(autorefresh=True)
print(rpi.cycletime)

rpi.io["Counter_3"].reset()
rpi.io["Counter_4"].reset()

rpi.io["Counter_3"].signed = True
rpi.io["Counter_4"].signed = True

time.sleep(1)

start_time = time.time()

previous_count = 0
previous_time = start_time
while True:


    time_now = time.time()
    counter1 = rpi.io["Counter_3"].value
    counter2 = rpi.io["Counter_4"].value

    print("Count3: " + str(counter1) + " Count4: " + str(counter2) +" delta: " + str(counter1 - previous_count) + " time_delta: " + str(time_now-previous_time))

    previous_count = counter1
    previous_time = time_now
    time.sleep(1)  # Sleep for 1 second

Simulated encoder signals moving backwards

Code: Select all

Count3: 35534 Count4: 0 delta: -10 time_delta: 1.000213384628296
Count3: 35524 Count4: 0 delta: -10 time_delta: 1.0002143383026123
Count3: 35513 Count4: 0 delta: -11 time_delta: 1.0002446174621582
Count3: 35503 Count4: 0 delta: -10 time_delta: 1.0002529621124268
Count3: 35492 Count4: 0 delta: -11 time_delta: 1.0002257823944092
Count3: 35482 Count4: 0 delta: -10 time_delta: 1.000220537185669

Simulated encoder signals moving forwards

Code: Select all

Count3: 35045 Count4: 0 delta: 20 time_delta: 1.0002079010009766
Count3: 35065 Count4: 0 delta: 20 time_delta: 1.0002102851867676
Count3: 35085 Count4: 0 delta: 20 time_delta: 1.0002093315124512
Count3: 35105 Count4: 0 delta: 20 time_delta: 1.000211238861084
Count3: 35125 Count4: 0 delta: 20 time_delta: 1.0002107620239258
Count3: 35145 Count4: 0 delta: 20 time_delta: 1.0002062320709229

Note that output 1 on the signal generator is plugged into DIO I4 and output 2 is plugged into I3

Device: RevPi Connect 4

Firmware version:
pi@RevPi106155:~/projects $ piTest -f
firmware is up to date: 1.5 >= 1.5

Pictory Configuration

Posted: **31 Jan 2025, 11:38**

Hello daniel_123, Foremost, thanks for all the information that I am still collecting, as a first idea so far - can you try to solve the problem by activating debouncing? Here is the spot in the video tutorial:
https://youtu.be/cBxmCPLKnlk?list=PLwce ... hNZu&t=329

Posted: **07 Feb 2025, 06:21**

Hi Dirk,

Thanks for your reply.

I have tried with all different debouncing settings (no debouce, 25us, 750us and 3ms) the behaviour is still the same.

Let me know if you have any other ideas.

Cheers

Posted: **20 Feb 2025, 11:24**

Hello daniel_123, can you check the wiring? It may be a ground wire that is causing the whole timig to go out of sync. Do you have the possibility to look at the signals with an oscilloscope?

Edit: I can try to create a test setup. However, as a function generator is then required as an external signal source I think it will take a few days.

Posted: **25 Feb 2025, 09:23**

Hello daniel_123 my friend, so i have performed a test setup without a function generator.

I used this signal diagram as a template

I used a DIO and wired it like this

Output Channel 1 -> Input Channel 1
Output Channel 2 -> Input Channel 2

I have set PiCtory like this

InputDebounce: 3ms
InputMode_1: Encoder
InputMode_2: Encoder
Output 1: A
Output 2: B

Code: Select all

import revpimodio2
import time
import sys

rpi     = revpimodio2.RevPiModIO(autorefresh=True)
encoder = [(0,0),(1,0),(1,1),(0,1)]
state   = 0
i       = 0
iMax    = 50
iSleep  = 0.1

rpi.io.Counter_1.reset()
time.sleep(1)
print("Counter_1: " + str(rpi.io.Counter_1.value))

print("Counting up")
for i in range(0,iMax):
        state = i % len(encoder)
        rpi.io.A.value=encoder[state][0]
        rpi.io.B.value=encoder[state][1]
        print("Counter_1: " + str(rpi.io.Counter_1.value))
        time.sleep(iSleep)

print("Counter_1: " + str(rpi.io.Counter_1.value))

print("Counting down")
while i>0:
        state = i % len(encoder)
        rpi.io.A.value=encoder[state][0]
        rpi.io.B.value=encoder[state][1]
        print("Counter_1: " + str(rpi.io.Counter_1.value))
        time.sleep(iSleep)
        i=i-1

print("Counter_1: " + str(rpi.io.Counter_1.value))

If I start like this

Code: Select all

python3 EncoderTest.py > /tmp/EncoderTest.txt

This file comes out

EncoderTest.zip: (330 Bytes) Downloaded 640 times

I could not detect any error

My device configuration:

Code: Select all

pi@RevPi94369:~$ piTest -d
Found 2 devices:

Address: 0 module type: 138 (0x8a) RevPi Connect 5 V1.0
Module is present
     input offset: 113 length: 6
    output offset: 119 length: 7

Address: 31 module type: 96 (0x60) RevPi DIO V1.5
Module is present
     input offset: 0 length: 70
    output offset: 70 length: 18

Posted: **26 Feb 2025, 16:43**

Hello daniel_123, there may be an electrical problem. Can you visualize the signals with an oscilloscope?

The Industrial Raspberry Pi Forum

RevPi DIO Encoder - forward counts twice as fast as moving backwards.

RevPi DIO Encoder - forward counts twice as fast as moving backwards.

Re: RevPi DIO Encoder - forward counts twice as fast as moving backwards.

Re: RevPi DIO Encoder - forward counts twice as fast as moving backwards.

Re: RevPi DIO Encoder - forward counts twice as fast as moving backwards.

Re: RevPi DIO Encoder - forward counts twice as fast as moving backwards.

Re: RevPi DIO Encoder - forward counts twice as fast as moving backwards.