Fixing false Positives with Geeetech Filament Sensor Switch

[macca448]

So silly me thinking I could fit the Geeetech Filament Run-Out Switch as a passive device and that would be it. After fitting the Geeetech Filament Switch the first few prints were fine but then it started to give false positives which is real annoying and time wasting.

THE SOLUTION: I like the Switch so hunting round I found this nice simple circuit

Opto_endpoint.jpg (25.19 KiB) Viewed 42086 times

The mod to add +5v to the X-MAX endpoint is here http://www.geeetech.com/forum/viewtopic ... 66&t=60915&
In a previous post I uploaded Marlin 1.1.9 with "Advanced Pause" enabled fir this feature to work. viewtopic.php?f=20&t=68977 (NOTE:In pins_GT2560_REV_A.h, PIN 24 is re-defined from X_MAX_PIN to FIL_RUNOUT_PIN for this)

The LED is optional. Figured a bit more bling doesn't hurt, aye!

I used shielded balanced audio hook-up wire (2 core plus ground/shield)
Made up the circuit and tested as these pics show

DSCF0539.JPG (332.26 KiB) Viewed 42086 times

DSCF0538.JPG (342.25 KiB) Viewed 42086 times

All working fine so then I installed it like so

2019_09220007.JPG (167.61 KiB) Viewed 42086 times

Hope this helps

Cheers

[_kaktus_]

HI.


Nice work.

I can also suggest a blocking capacitor for the input of the optocoupler diode. This will improve reliability.

[macca448]

Thank you _kaktus_!

Is this what you are suggesting? A 100nf ceramic cap between the diode input and ground?

Opto_endpoint_v2.jpg (24.93 KiB) Viewed 42057 times

Its a mission working on the electronics now its in a housing to tidy things up but next time I'm doing some work in there I will take your advice
Cheers

[Jarda3d]

The capacitor itself is enough at CPU side of course. You definitelly did not need optocoupler.

If you get still problems, then add one more parallel to the switch too. If still, then add resistor about 100-220 Ohms in series with switch NEAR connector of CPU board. Or if you are paranoid, use these ceramic caps in parallel: 1nF, 100nF, 1uF and schottky BAT54S if you know what to do with all.

If you buy 0805 SMD version, it could be add at bottom side of board to connector leads.

Really, this EMC problem did not need any optocoupler!

[macca448]

I appreciate your comments Jarda3d but for what its worth I'm not an engineer, just a hobbyist. I wanted a guaranteed solution and your ideas are not what I found to be the best solution. Because my 3D printing is to make a CNC engraver I chose to go with the GRBL groups "Best" solution as it has many benefits. Maybe you should read this thread about how they came to this way of noise filtering. https://github.com/gnea/grbl/wiki/Wiring-Limit-Switches

The components were cheap and readily available and easy to make a retrofit cable. My i3ProX was printing for 16 hours yesterday without fault so I'm now very happy chappy having solved the problem with a very simple easy to make circuit.

[_kaktus_]

Hi.


It's more like that:

Opto_endpoint_A.jpg (13.39 KiB) Viewed 41974 times

The capacitor at this point is designed to mitigate the effects of contact bouncing during switching.

I greet all hobbyists and electronics enthusiasts.

[macca448]

Thank you kaktus, I understand. It's a good day when you learn something new and today I learn something new!!

[_kaktus_]

Hi.


Oh, no problem.

The resistance limiting the short circuit current of the tact switch is also important.
It is already present in your circuit diagram because you have an LED.

[Jarda3d]

EMC.png (15.53 KiB) Viewed 41761 times

[_kaktus_]

Hi.


Excellent. That's what I was talking about.