Micrologix 1400 Encoder equation

Christoph · Jul 26, 2021

Hey All!

Using the HSC:0 for an unwind. Not high speed or anything.
50 PPR at fastest 800 FPM.using this math
pulse count per second / 50 the ppr * 60 for seconds = rpm
When I look at the data and do the math by hand it is right on
But letting it do the math in a cpt block it is really off.
Any Ideas?

drbitboy · Jul 26, 2021

Christoph said:
Hey All!

Using the HSC:0 for an unwind. Not high speed or anything.
50 PPR at fastest 800 FPM.using this math
pulse count per second / 50 the ppr * 60 for seconds = rpm
When I look at the data and do the math by hand it is right on
But letting it do the math in a cpt block it is really off.
Any Ideas?

Give examples of the numbers that are off.

Is the [pulse count per second i.e. Hz] a 16-bit unsigned INT or a 32-bit LONG?

Try putting the Hz value in a float and go from there.

Try reversing the /50 and the *60, or use parens to enforce the intended order of operations, or just multiply by 1.2 (= (1/50) * 60)

Christoph · Jul 27, 2021

The pulse count was a Long and I changed it to an Integer, no affect.
I already use parenthesis in the equation (Hz|50)*60=
I have found this processor does not seem to like just a digital value in an equation
but would rather have a data type with the value entered in the data table
I will try your suggestions in a little while.

drbitboy · Jul 27, 2021

Can you give an example e.g.

At an a measured or controlled 90RPM, the HSC measures a rate of 75 pulses per second i.e. ppr=75Hz, but the compute instruction that evaluates (ppr/50)*60 gives an answer of 62.5RPM.

Also, can you attach your code (.RSS in a .ZIP) to a post?

Christoph · Jul 27, 2021

I changed the destination from an Integer to a Floating point and all seems well now!
Go figure!

drbitboy · Jul 27, 2021

Btw, you could also use a SCL instruction for this. Not that the result matters ...

bernie_carlton · Jul 27, 2021

If staying in integer format then executing the multiplication before the division will help avoid loss of precision. (PPR * 60)/50

Just ensure that PPR*60 does not produce an out of range (over 32767) result.

drbitboy · Jul 27, 2021

bernie_carlton said:
If staying in integer format then executing the multiplication before the division will help avoid loss of precision. (PPR * 60)/50

Just ensure that PPR*60 does not produce an out of range (over 32767) result.

Of course.

Were the "really off" results always a multiple of 60?

Christoph · Jul 27, 2021

No it seemed multiple of 10 or 20

drbitboy · Jul 27, 2021

Christoph said:
No it seemed multiple of 10 or 20

Huh. That's not what this suggests, but maybe MicroLogix 1400 handles it differently.

Micrologix 1400 Encoder equation

Christoph

Member

drbitboy

Lifetime Supporting Member

Christoph

Member

drbitboy

Lifetime Supporting Member

Christoph

Member

drbitboy

Lifetime Supporting Member

bernie_carlton

Lifetime Supporting Member + Moderator

drbitboy

Lifetime Supporting Member

Christoph

Member

drbitboy

Lifetime Supporting Member

Similar Topics