KOP Mangetic Encoders Don't Spin Full 360

[JewishDan18]

Hello, today I was using the KOP magnetic encoders in analog mode. To my understanding, It should read 0 volts at 0 degrees, and 5 volts at 360 degrees, and vary linearly in between. However, It seems that they reach 5 volts at about 270 degrees, and reset after that. This is true for both the encoders in the kit. Perhaps the magnet is not properly centered on the chip? If so, it is strange that they are both reseting at a similar angle. Has anyone else experienced this? Thanks for any help.

[sparrowkc]

Our mag encoders do not have dead spots. We are reading them over SPI.

[JewishDan18]

Its not so much a dead spot, but it read 0 volts at 0 degrees, 5 (or 0) volts at 270 degrees, and if i keep spinning, reads 5 volts again at 180 degrees

[Alan Anderson]

What you describe doesn't make sense. How are you testing this? It might make things clearer if you showed us a picture of your setup.

[JewishDan18]

Perhaps the attached image will help. We are measuring the value with a voltmeter and outputing the number to the DS, and they match. The values are the voltage multiplied by 72 to convert it to degrees. So 0 volts is 0 degrees, and 360 is 5 volts

[sparrowkc]

Does the location of the border between 0 and 5 volts change over several rotations? Just to verify, you have a proper setup (magnet centered over the chip with an air gap of less than 2 mm) and aren't spinning the magnet in your fingers or something, right?

An encoder sends out a stream of 0V to 5V pulses as it rotates...Don;t look for anything voltages in between. They are digital not analog.. Not sure about the KOP encoder but typically you get 1024 pulses per revolution. .Your code counts pulses and the time between the pulses to get the information it needs. And there are 2 channels putting out pulses but they are out of phase with each other. Using that difference you can determine the direction the axis is turning. Use an oscilloscope to test the outputs and observe the pulse train...

BC

[Jared Russell]

Originally Posted by bcieslak

An encoder sends out a stream of 0V to 5V pulses as it rotates...Don;t look for anything voltages in between. They are digital not analog.. Not sure about the KOP encoder but typically you get 1024 pulses per revolution. .Your code counts pulses and the time between the pulses to get the information it needs. And there are 2 channels putting out pulses but they are out of phase with each other. Using that difference you can determine the direction the axis is turning. Use an oscilloscope to test the outputs and observe the pulse train...

BC

This is for a quadrature-type (optical) encoder. The original poster is describing the magnetic encoder that came in the kit of parts this year.

[Alan Anderson]

Originally Posted by JewishDan18

Perhaps the attached image will help. We are measuring the value with a voltmeter and outputing the number to the DS, and they match. The values are the voltage multiplied by 72 to convert it to degrees. So 0 volts is 0 degrees, and 360 is 5 volts

You're not telling us the whole story. Your green numbers can't get as high as the picture shows them reaching.

What is your magnet affixed to, and in what orientation relative to the sensor?

[JewishDan18]

The magnet is fixed to an axle, above which is the chip, mounted on a lexan bracket. The chip is pretty freaking close, but I'll check it out tonight and make sure it is close enough.

I made a mistake in describing the picture. The white numbers show how far it has turned, and the green numbers show how far it would think it has turned (both over multiple revolutions). Take that number modulo 360, and it'll give you the degrees it reads. Sorry about that. I'll try to get some pictures tonight of the setup, if it'll help.