RPM measurement

[gaks]

I want to measure RPM using 8-pin rotary encoder using PXI-6221. I am able to measure angular position but RPM is fluctuating.

[johnsold]

What is an 8-pin encoder? How many pulses per revolution? What is the range of speeds you are trying to measure?

The formula node looks like it might have been set up for 1024 pulse per revolution but the front panel control is set to 24?

If your loop succeeds in running 200 times per second, your speed would need to be rather high to avoid fluctuations. Look at an example: Let the encoder generate 24 pulses per revolution. Suppose you want the fluctuation to be less than 1% on successive readings (for a constant speed). Because the counter and the encoder are not synchronized, there will always be a 1 count variation. To meet the 1% criterion the average count must exceed 100.  At 100 counts every 5 ms the pulse frequency is 20000 Hz. Divide by 24 pulses per revolultion to get 833.33 revolutions per second. Multiply by 60 to get 50000 RPM. If your speed is less that 50000 RPM, then the fluctuation will be > 1%.

Lynn

[gaks]

8-pin encoder means it has 6 output and 2 pin for power input. Avilable points are, A,  A_bar,  B,  B_bar,  z,  Z_bar.  Currently I am using only A, B, and Z points. How to use all?

Sorry while uploading the vi I dint chk it. It produces 1024 pulses per revolution.

Range of the speed is 300-900 RPM

[johnsold]

A, B, and Z should be fine.  Are all the signals driven by logic devices or are they produced by slower circuitry? Sometimes counters and digital inputs on DAQ devices will experience spurious counts if the transition times on the input signals are too slow.  If your fluctuations usually tend to produce higher than actual speeds, this may be a primary suspect.

Repeating my calculation for the numbers you have now provided:

300 RPM/60 = 5 revolutions/second. 5 r/s * 1024 = 5120 Hz. 

900 RPM/60 = 15 r/s.  15 r/s * 1024 = 15360 Hz.

Both of these are below the 20 kHz I calculated would be required for < 1% fluctutation. At 300 Hz you would get ~4% fluctuation just due to the one count variation.

Lynn