Firgelli Feedback Actuator isn't giving any feedback

[osolgineers]

Hi everyone!

I recently bought a Firgelli feedback linear actuator mounted with an optical encoder to track the stroke.

This is the product : robotshop.com/eu/fr/actuateu … metre.html

I tried to read the impulsions using Arduino and BeagleBone Black interuptions but nothing came with that. So I further tried to use an osciloscope to visualize wether impulsions could be read but it seems that I have nothing but a straight line, no impulsions…

What do you guys think about it? Could it be the optical sensor that is just dead?

Thanks guys, Max!

[scharette]

Hi,

To help you with this, it would be very helpful if you could reply with some pictures of your setup.

Make sure to show clearly all components (linear actuator, Arduino, power connections, signal connections, etc.).

Sincerely,

[scharette]

While waiting for your reply, we noticed we had one in stock so we tried it out. We also got no signal (both on logic analyzer and oscilloscope).

We contacted the manufacturer about it and here is copy of their reply:

@manufacturer

I am assuming the encoder was properly powered up with the 5VDC required.

Next, the output needs to be pulled up logic HIGH with either an external 10-15Kohm resistor or if using Arduino, set the input pin mode to INPUT_PULLUP

pinMode(XX ,INPUT_PULLUP);

Also, on Arduino, if using interrupts to read the pulses, ensure they run a shielded wire from the encoder output to the Arduino, to help with noise immunity.

And ensure the ISR incorporates some kind of “debounce” to prevent false pulse registration, Arduino interrupts can be quite sensitive to noise pulses from the high current DC motor, depending on the wiring.

With those extra details, it will most likely help you obtain a signal out of the output (green wire).

Try it out and let us know if it works any better this way!

Sincerely,

[osolgineers]

hello,
first of all thanks for your quick reply.

I performed the step you gave me and I’m happy to say that using a PULL_UP resistor, it works both on the oscilloscope and microcontroller, and I have fancy curves without noise.

Last question, which sampling frequency would be ideal?

Thanks again, Max.

[scharette]

Hey Max,

We are glad it worked out well for you.

As for managing this output, we recommend that you use interrupts. Just make sure to have a proper wiring to prevent noise. If you get spurious signals, you may want to add a filter (resistors, capacitors, inductors) at the input pin.

If you decide to poll the input for the pulses, you will want to over-sample enough so that you do not miss any of the pulses. The first step would be to determine the rate of the pulses.
In this case, the device produces 100 pulses/inches and moves at about 30 inches/s. Therefore, it will produce at most 3000 pulses/s, or a signal that is ~3 kHz. Normally, when sampling a signal, you want to sample at a rate higher than the signal to be able to reproduce it properly / not miss any parts. Many digital systems will over-sample 8 or even 16 times to be certain of the data being read / reduce error rate due to noise.

Sincerely,