Improve the accuracy of RPM measurement.

[vanquangtk 0]

Hello all.

I have a problem with RPM measurement for a spindle. I hope someone can help me.

My company's machine uses a disk with two pinholes to measure the speed.In the current program, it uses the following instruction to measure the speed:  SPD X0 K3000 D1

The issue is that we only receive 2 pulses per revolution, which requires using a longer sampling time for accurate results. Unfortunately, this leads to a delay in the response time.

I am considering trying a different method for measurement. Would using a PID controller be a good idea?

 

[pturmel 160]

If your measurement method is slow or imprecise, fundamentally, the only solution is to alter your measurement method.  If you have a high speed clock available, a common technique is to count clock pulses from sensor leading edge to leading edge, giving you the signal period.  From which you can take the reciprocal to get frequency.  If the available clock rate isn't high enough for the desired precision in one period, count it for multiple periods.

Using a PID controller or not has no bearing on the measurement of your process.

[vanquangtk 0]

2 hours ago, pturmel said:

Using a PID controller or not has no bearing on the measurement of your process.

I intend to use it to optimize the sampling time in order to ensure both response time and accuracy.

[pturmel 160]

And how can a PID controller do that?

[Ron_S 61]

put more pin holes in the disk and recalibrate

[vanquangtk 0]

3 hours ago, pturmel said:

And how can a PID controller do that?

This is code I wrote to do that

code.txt