Gear Tooth Sensors

[et1337]

Before I even start... I've already searched all over the forums.

So, our interrupt software works, and we know our sensor is hooked up and oriented right, but the software didn't register any clicks. So we hooked up an oscilloscope to it, and it's not registering any sort of pulse. So... we need help. Is it broken, or what?

[steven114]

Check that you're getting power, that it's the right voltage, and that it's clean. Check the datasheets to see the allowable voltage range, and make sure you're not using 5V when you should use 12, or vice versa.
One of the trickiest problems I've ever debugged electronically (I'm no expert, so this may be a 'duh' thing to anyone who knows what they're doing) was when my +5V power rail was too noisy - the multimeter registered 5 just fine, but when you hooked it up to a scope it actually oscillated between 5 and 0 at something ridiculous like 5MHz. That was playing strange games with my PIC's oscillator...

If you're sure you're getting good power, try a different one? Maybe it is broken.

[Alan Anderson]

Describe exactly how you have it hooked up. You say you know you did it right, but you'd probably say the same thing if you did it wrong.

[Ryan M.]

Originally Posted by Alan Anderson

Describe exactly how you have it hooked up. You say you know you did it right, but you'd probably say the same thing if you did it wrong.

An excellent point. I'm on his team... and haven't the faintest idea how we have it hooked up.

To keep this post from being completely worthless, I know that at least one professional electrical engineer looked at it. Beyond that... *shrug*

[DonRotolo]

Originally Posted by ETodd

Before I even start... I've already searched all over the forums.

Thank you!

Originally Posted by ETodd

So we hooked up an oscilloscope to it, and it's not registering any sort of pulse.

On the First hand, the hook up: From left to right, you want Ground, (nothing), +12 volts, Ground, +5 volts, and the last one is the signal output. That's right on the board.

On the Second hand: The GTS will sense the passing of a ferrous (Iron-based) object. Make sure a magnet will stick to your "gear".

On the Third hand: The Ferrous Gear must be within about 3 mm to be sensed. 1 mm is better and more reliable.

There is also a frequency limit, but it is doubtful you have reached that. Make sure the 'scope settings are such that you will see the pulses: The voltage will be around whatever is at the 5 volt input pin. You will see normally 0 volts with a brief pulse for each 'tooth'.

On our test rig, we used a flat steel bar, about 1/4" wide and 1/32" thick, 1.5 inches long, spun with one of the small fast motors in the KOP. We took great care to ensure the GTS was not struck by the bar as it spun.

Let us know how it turns out please! Also, we still can't figure out the interrupt code, so if you get something I'd be grateful if you could explain it to our programmer...

Don

[John Gutmann]

try hooking it up away from the gear and just waving a ferous metal in front of it while it is hooked up to an oscilliscope to seeif you get a wave. If you do get a wave the sensor is not broken.

[Kevin Watson]

Originally Posted by Don Rotolo

...Also, we still can't figure out the interrupt code, so if you get something I'd be grateful if you could explain it to our programmer...

There are a few documents in the white paper section. I know that Eugene Brooks and Dan Katanski have written papers, so you might search by author.

-Kevin

Edit: Here's a link to Dan Katanski's paper.

[Alan Anderson]

A "wave"? The GTS board produces a nice, clean, digital high-going pulse of less than a hundred microseconds each time it sees a gear tooth go by.