Array size in Analog Tacho Processing vi

[tronne]

Hi

I would like to know more about what Analog Tacho Processing vi does particularly on the array size that it generates.

Attached is an example of vi that comes with LabVIEW. I used the array size function to show the size of the signal before passing thru the Analog Tacho Processing vi, and after the signal passes it. When we run the PC fan runup.dat file (which can be found in the LabVIEW order analysis Example Data), we can see that the array sizes for before and after passing thru the Analog Tacho Processing vi are 58368 and 1455, respectively. What I would really want to know is how the 1455 is generated.

Any help is greatly appreciated.

Thanks

[KyleTTU]

Hi, 

Am I correct in assuming you are measuring the array size of the Speed Profile output?  Which array in the cluster are you probing?  Basically this vi is going to take multiple points to determine rotational speed.  Thus you will never have a 1:1 ratio of input points to speed.  

Are you more curious as to what the formula behind the vi is?

Kyle 

[tronne]

Hi Kyle, 

Thanks for your respnse. Yes, I am measuring the array size of the Speed Profile output. I am curious as to what the formula behind the vi. Could you please explain more when you say the vi is going to take multiple points to determine rotational speed, and that there will never be a 1:1 ratio of input points to speed.

Thanks.

[KyleTTU]

Since the Express VI calculates rotational speed based on pulses per revolution, it will use those pulses to calculate the speed profile.  For example your input waveform may have 1000 data points to describe 3 pulses, the VI is going to just use those 3 physical pulses to determine rotational speed (based on the value you set for pulses per revolution).  Thus you will never see the same number of points on the input and output of this VI.  If you are trying to determine the exact formula it is using I will not really be able to help you with that.  

My best suggestion would be to use context help and the detailed help for that VI, you can also right click and open the front panel of the Express VI if you are so inclined.

Happy Programming!

Kyle 

[tronne]

Hi Kyle,

Thanks again for your response. I've emailed ni.com for some help (before I received your response), and they would be providing detailed information on the vi soon.

In your explanation, you mentioned that the Express vi. will use the pulses to calculate the speed profile. How do we determined or know how many pulses are needed to generate the speed profile? And how do you determine the value that you set in the 'pulses per revolution'?. I've searched the Help file, manual etc. For example, in the Help file, it just describes what Pulse/Revolution which is "specifies the number of pulses the tachometer generates during each revolution. The default is 1." Here it doesn't tell how and where to get this value. I simply assume it's 1 pulse per revolution.

Thanks so much for your help.

[KyleTTU]

Hi, 

How are you taking in these pulses?  Typically if you are using an encoder the encoder is spec'd for a certain number of pulses per revolution.  You would then use this information (specified by the encoder) to tell the Express VI how many pulses dictate a complete revolution.  Then the VI will be able to calculate RPM and the speed profile. 

Kyle

[tronne]

Hi,

Thanks again for your reply.

I use a tachometer that has an optical sensor which captures the coincidence with a reflective tape on a rotating tool. The tachometer is set with 1 pulse per revolution.

In your previous reply, you gave an example in which you said "an input waveform may have 1000 data points to describe 3 pulses, and the VI is going to just use those 3 physical pulses to determine rotational speed (based on the value you set for pulses per revolution).  Thus you will never see the same number of points on the input and output of this VI"

So, how do we know how many data points are needed to describe the pulses that determine the rotational speed?

Thanks so much for your help.

[KyleTTU]

Hi, 

I apologize for my delayed reply.  It is not easily determined how many exact data points are needed to represent a pulse.  Typically, you want to be sampling your incoming signal at at least twice the frequency it is running.  So, if your optical sensor is sensing something moving at 50 times a second, you would want to be sampling it at a rate of at least 1000 times per second (or 1kHz).  This will ensure that you can properly represent the pulses in software and that the vi does not reference any aliasing signals.

Kyle