Counting the RPM from analog signal

[CoastalMaineBird]

Come on, this is basic math.

If you know that it's 527 samples between one rev and the next (samples per rev), and you know that your sample rate is 10000 Hz (samples per sec), then.....

Beware, though - if you just count the  number of revs in a block, you won't be accurate, since there are extra samples at the beginning and end.

[hexagonal]

Hi,

I don't understad what the sampling rate has to do with samples between revs. I have 58 peaks between revs and 10000 is just the rate of sampling of these peaks(actually the whole curve). I don't know how many samples I have per rev, since this will be the function of revs, more revs, less samples per rev. I will have to increase the sampling rate anyway since the engine will spin beyond 13000 rpm.

I am very sorry for these rookie questions, but I am new to the world of programming.

Thanks.

[CoastalMaineBird]

If your sampling rate is 10000 Hz, then it's 100 uSec between samples, right?

Suppose tooth #0 peaks at sample #113 of your block of data.

You find all the peaks and find that tooth #0 peaks again at sample #5913

The distance in angles is 58 teeth = 1 rev

The distance in samples is 5913 - 113 = 5800 samples

The distance in time is 5800 x 100 uSec = 580 mSec

therefore, you're going at an AVERAGE speed of  1 rev/580 mSec, or about 103.45 RPM.

This is assuming that your gear has 58 equally spaced teeth, and is NOT a 60-tooth gear (VERY common for RPM measurements) with a couple of teeth missing.

[hexagonal]

Can you show me the direction in which to go please? Maybe I am looking at this too much from engineering standpoint and not programming standpoint. To do this in analog world, I would take the stopwatch and see how many counts per revolution I have per second. Easy thing. What would be the ''stopwatch'' in LV?

[CoastalMaineBird]

You're not understanding something:  your SAMPLE RATE is your stopwatch. 

You asked it to sample at 10000 samples per second.  

Trust that it does that.