RPi threading too slow

Question

So I have rpi wheeled robot, with wheel encoders and I want to always keep the track of the encoders and to be able to access the value when ever I need it.

Basically, I created a thread, which always keeps incrementing global variable on a change of encoder. There are 12 stripes on a wheel, I tested just on main and it works good, it kept track of all 12 and stopped at exactly the same point as it started. However, I need to be able to track both of them at the same time. With threads, its so slow, that it misses about 30% of High/Low and it is bad. However, with multiprocessing it kept track of both of them accurately, but I cant access the memory whenever I need.

def countRightEncoder():
global countRight
flag=0
while True:
    if encoderRight() == True and flag!=True:
        countRight+=1
        flag=True
    elif encoderRight() == False and flag!=False:
        countRight+=1
        flag=False
    print countRight

This loop will just always go, it wont stop and I just want to get the countRight value really often.

I was just wondering does anyone have any suggestions on how I could manage to do this without loosing any data? Maybe there is an easy way to get the value from multiprocessing?

Thank you

Answer 1

Perhaps try different software?

I've used (my) Python rotary encoder software to read multiple encoders.

If that doesn't work you will probably have to code your application in C.

If all you want to do is count pulses then just use a simple callback. The following code will set up a default callback on GPIO 2 and GPIO 19. The count (tally) of pulses is checked every second.

#!/usr/bin/env python

import time
import pigpio

pi = pigpio.pi()
if not pi.connected:
   exit()

cb1 = pi.callback(2)
cb2 = pi.callback(19)

stop_time = time.time() + 60

while time.time() < stop_time:
   print("cb1={} cb2={}".format(cb1.tally(), cb2.tally()))
   time.sleep(1)

cb1.cancel()
cb2.cancel()
pi.stop()