Help with CNC 3018 from Banggood

[JeanB]

Good morning Doddy,

Thank you for all your help.

As I can see you are not much in favour of Optocoupler so I was reading on noiseless end stops and came across the following which appear to have little noise, do you think these would be better.

1. Optical end stops as per Alexpress : https://www.aliexpress.com/item/32656670473.html

Thanks

[Doddy]

I don't know. Buy one and report back.

Personally I'd be buying 3 resistors but that's just me and my crazy ways.

[Kitwn]

I don't know. Buy one and report back.

Personally I'd be buying 3 resistors but that's just me and my crazy ways.

Agree. It seems that the lack of pull-ups is the problem and, if it's outputs are open-collector, the opto-coupler will need the pull-ups anyway!

[JeanB]

Good Morning Kitwn,

Thanks for your help.

I was going to install the resistors (cheapest solution) and separate 5v power supply, since I have one lying around. However, it will be a while before I can do anything as we are restricted to our homes due to the Corona virus. So I will let you if and when I get those resistors.

Thanks again.

[AndyUK]

After hours of reading, I changed all my limit switch wiring to shielded wires……..Nothing still same issues.

Hi Jean,

Agree with Doddy and Kitwin that pull up resistors are indicated and should be fitted - but haven't seen any mention about shielding of the spindle wiring? You mention limit switch wiring shielding, but not spindle?

If its not already, I'd recommend making that shielded too in order to reduce the noise output. Also be sure to ground your shields at one (and only one) end.

[JeanB]

Thank you, Andy,

I will shield the spindle and ground all the shields. I, at least, can do this for the time being.

[MikeG]

Hi Jean,

I am a complete novice as far CNC machines go and I have got myself a new cnc 3018 pro machine. I just assembled it and have been playing around with the GBRL and calibration stuff and thought I should enable the homing feature which would need endstop switches.

I was looking at the Optical Endstop that you have mentioned in earlier posts and thought it would be a nice clean method as opposed to using the micro-switch option.

Did you get round to solving the issues you were having? Will be grateful if you could upload a pic or two of what your switches look like and how you have attached them to the XY and Z axis.

Thanks,
Mike

[Muzzer]

Rather than bugger about with resistors, caps and optos, these can be very effective and require no electrical changes
https://www.aliexpress.com/wholesale...=ferrite+clamp
You can fit them over the mains cables, the motor wiring (make certain all 3 wires go through together though!) and any signal bundles you have in your cabinet.

Note that VFDs are really supposed to be fitted with an EMC filter on the INPUT wiring (L, N & E). Many / most people omit to fit them and wonder why they are finding noise issues. There's no standard size but as long as you ensure it's rated for the max current, you can't go too far wrong. For reference, officially recommended EMC filter on my Yaskawa VFD is the same footprint as the VFD itself.
https://www.aliexpress.com/item/3273...%2315608%23170

I have no experience with any of these suppliers but the products shown are in the right region, technically speaking.

[MikeG]

Thanks Muzzer for your insight on EMF interference. Yes, I agree with what you mentioned. As I am an Electronics engineer from the 1980's, I do have a good understanding of these issues.

I thought about the need for resistors, caps and opto to reduce the stray currents - it would be over kill for a small hobby cnc machine like what I have got. In my case I am powering the controlled using a 24vDC power adaptor as opposed to maybe others who might be using the high wattage Mains PSU's to power their CNC machines and my be succeptible to such noise.

The spindle motor I am using is a basic 2 wire 24vDC 8000 rpm type so it is unlikely to interfere too much. I might solder a capacitor across the terminals to reduce and startup spikes.

Cheers

[Doddy]

I thought about the need for resistors, caps and opto to reduce the stray currents - it would be over kill for a small hobby cnc machine like what I have got.

Knock yourself out with whatever works for you.

But here's some real-life experience:

I got one of our software apprentices to knock up an interface module using a Uno - idea being to support a variety of input devices (typically switches), output these encoded via serial to a Pi driving a TFT display and sending the input switch states over ethernet to a control system. The generic nature of this meant that, depending on switch configurations, one or more inputs could be disconnected / open-circuit. Unused, unconnected inputs could be driven either high or low through the physically adjacent pins that were connected. Solution to this was to use the internal pull-ups but a floating input was highly susceptible to parasitic drives - whether from capacitive coupling, or possibly through contaminants(grease, sweat?) on the board.

Another one, recently, despite my views on opto-isolators, another user on here was experiencing sporadic limit switch triggers on a 6040 machine, which we resolved by using a commercially available opto-isolator board on the inputs to the BoB. Prior to that, without the spindle operating, but under axis drive (so just stepper motors) that user was experiencing sporadic limit switch trips within around 8 seconds of motion... had to use mach3's debounce setting values of more than 1000 (40mS for a 25kHz kernel) to get anything remotely steady operation before we got the opto-isolated inputs stable.

Of course, your results may vary.

80's?, damn, you pre-date me by a couple of years.