Created Date: February 16,2015

Created by: Tinker

An interrupt should get you some improvement, if the interrupt routine is short you might manage as high a 1000Hz. I have successfully counted the pulses from a prox. triggered by a 6 tooth milling cutter at 4200rpm (420 PPS), I guessing at the 1KHz, but I suspect it is plausible.

Are you generating more than one pulse per revolution? even with conventional ladder and a 11-13ms scan time I'd expect pretty reliable counting up to somewhere around 30Hz or perhaps a bit more, problems at 11 seems odd.

EDIT: I just looked up the specs of the AK1-AN-1A, they say: "Switching Frequency 600 Hz ", so first, you will not get to 1KHz with that, I'm wondering what kind of object the sensor is detecting. I have seen (on a slow speed application) a prox. used to sense about a 1/2 " dia. bolt head on about a 2 " radius, making the on time less than 1/10 of a revolution. I'd be inclined to guess that "600Hz " is with a 50% duty cycle, at 10% I'd wouldn't expect reliable detection at more than 120Hz or so, maybe even less. However that is still far faster than the 11 you are having trouble with. Still, if you are hoping to measure high speed, you might consider making the detection surface 1/2 od a revolution.

Created Date: February 17,2015

Created by: milldrone

Does anyone have a idea how I might make improvements?

When I go to 11 rps the value falls right off to like 5 - 6 rps.

Thank you

I cannot tell from your post if the duty cycle of the prox target assembly is 50%. Meaning the on time is near the same as the off time. To get the most from your setup you need the off time to equal the on time.

Created Date: February 17,2015

Created by: kasky

Thank you very much for your responses.

I was dreaming about this last night (don't you hate that) and came to work this morning with pretty much the same idea, to increase the surface area of the pick up. So far it has made a significant difference in accuracy. From the former 10hz, I'm now seeing accurate numbers as high as 30hz. This is with about a 30% on time per revolution.

Making progress, thanks again guys.

Created Date: February 16,2015

Created by: kasky

I have a Click C0-02DR-D CPU.

I am trying to set it up to count revolutions per second of a electric motor.

I am using a AK1-AN-1A proximity sensor for the input to X001 on the Click.

In the I/O setup I have X001 setup for pulse catch, I have tried regular with no improved results.

My scan rate is 11 -13 ms.

I am able to count accurately up to 10 rps (for testing I am using a VFD to set the motor speed).

Does anyone have a idea how I might make improvements?

When I go to 11 rps the value falls right off to like 5 - 6 rps.

Thank you