Page 5 of 6

Re: twforeman's build

Posted: Wed Dec 25, 2013 10:43 pm
by twforeman
Over the past week or so I tore down my ShapeOko and rebuilt it with the new motor mount plates. I also tweaked it for more accuracy (you might sense a recurring theme here.)

Here are a couple of photos:

Joining the double Y Makerslide.
08-rail-join-2.jpg (183.76 KiB) Viewed 2438 times
Investigating options for the Z axis.
11-new-z-1.jpg (200.8 KiB) Viewed 2438 times
Making sure all the Makerslide is positioned correctly.
18-y-adj.jpg (169.56 KiB) Viewed 2438 times

As always, full details are on my blog.

Re: twforeman's build

Posted: Sat Jan 25, 2014 8:35 pm
by lordmundi
I have really enjoyed following your build progress. I'm in the process of installing some limit switches (mine are mechanical at the moment), but I'm curious how people are wiring them up. I understand electrically how everything needs to get hooked up, and I've got my header pins soldered onto my grblshield to attach the switch leads to, but I'm more curious about how folks _physically_ wired them up... for example, are people making a custom circuit board to tie all the grounds together and then tie the common signal line for each switch on an axis? or are folks making some sort of wire harness that connects the lines?

Anyhow... just curious how you wired yours up and how you connected ground lines and signal lines together.

Thanks for documenting your build!


Re: twforeman's build

Posted: Sat Jan 25, 2014 11:09 pm
by twforeman
So I'll try and describe the wiring for my switches. I used shielded cable with five conductors in it. This allowed me to run power and ground and all three signal lines in one cable.

My Z axis only has one switch on it with two flags to trip it.

The X and Y axes have two switches each, but they are wired in parallel so they only need one signal wire per axis.

The cable runs in the same energy chain as the motor cables (which also use the same 5 wire cabling) to the X axis switches on the motor plate. There I cut the cable sheath open and spliced the wires for the X axis switches in. Then the cable runs in the energy chain to the Y axis motor plate where it terminates to the Y and Z axis switches.

So, since there is only one cable for the limit switches, there is only one ground and one shield to connect.

I should take some new photos of the wiring, since the ones on my blog are not current. I'll try and do that soon.

Hope this helps.

Re: twforeman's build

Posted: Tue Jan 28, 2014 4:53 am
by Mgilbride
Also enjoy following your build. Two questions re: your limit switches.

1.) are you using resistors / capacitors to filter out noise? If so do you mind sharing a rough schematic?

2.) when you say you left the motor grounds floating while you grounded the shield of the signal line (from your blog link a few posts ago) you mean you left the motor cable shields floating, correct? Have you subsequent changed this or have you left those cable shields tied together and floating?

Re: twforeman's build

Posted: Tue Jan 28, 2014 1:53 pm
by twforeman
1. Yes, I put some capacitors across the power and ground wires at the switches. The capacitors I used were really old, so they may be in pretty bad shape. Not sure how useful a schematic will be, but I'll try and generate something.

2. Yes, I misspoke. I left the motor cable shields floating. They are still tied together and floating.

I do get the occasional random limit switch trip and I'm not sure what causes it. It doesn't happen very often.

I'm hoping to get an oscilloscope in the near future and I'll be hooking it up to the limit switch wires and doing some experiments to see what causes the noise issues. That will be interesting to see.

Re: twforeman's build

Posted: Tue Jan 28, 2014 5:00 pm
by cvoinescu
My guess is that your shields would be more effective if tied to ground/earth, or to the V- terminal of the motor supply (both of which could be tied together, too). It's only a guess, though.

Re: twforeman's build

Posted: Tue Jan 28, 2014 5:10 pm
by twforeman
I tried tying all the shields and grounds together (in a star) and the limit switches locked up as soon as the machine moved.

I do need to play around with it some more, but it's currently working pretty well.

Re: twforeman's build

Posted: Thu Jan 30, 2014 2:54 am
by Mgilbride
I'm in the same boat. Working pretty well but two nagging issues. One being the limit switches which still trip every 30 minutes or so. Really wished these were set up as normally closed switches. Starting to sound like that may be a best practice for these type of low voltage switches. I think that would make them a little easier to set up and more reliable. That sound right?

I tried leaving stepper cable shields tied together and floating but that had the reverse effect that you had. My limits started triggering immediately. So tied them back into rest of the grounds and crossed that off the list. Wonder why we are experiencing reversed effects?

"Schematic" of my system is below. I really don't know much about EE so am a bit blind when trying to ferret out ground loops and other potential problems / errors.
Schematic (Custom) (2).jpg
Schematic (Custom) (2).jpg (454.1 KiB) Viewed 2134 times

Re: twforeman's build

Posted: Fri Jan 31, 2014 2:21 am
by CrazyBillybob
I've noticed something in all of the circuit diagrams from folks with Limit switch issues show the filter caps close to the switches.
I'm not sure if this reflects the actual Placement of the capacitors in their build or not. But you really want those caps as close as possible to the Arduino not the switch. You also want a resistor in series with the cap. The Goal of the Resister Capacitor pair is to act as a low pass filter (The blocks the high frequency noise from motors and etc.) by moving the R/C pair close to the monitoring point you limit the amount of wire you have exposed add noise after the filter.

I have not connected any limit switches to my shapeoko yet so I don't have any first hand knowledge this is all electrical theory.


Re: twforeman's build

Posted: Fri Jan 31, 2014 2:42 am
by Mgilbride
The capacitor at the switches is my actual implementation. If you read this page ( you will see several suggested implementations of capacitors for noise control. A small cap across limit switch contacts is one of the mentioned implementations. However, there is no real explanation of why that's a good solution. in truth, I'm unclear on all the different goals people have for these components. What I think I understand is:

1.) Add stronger pullup resisters at end of signal lines closest to Arduino making it harder for noise to trip limit. Lower ohms, better noise immunity, worse signal bounce. Signal bouncing supposedly adequately handled by software debounce written into GRBL.

2.) Add a low pass filter at Arduino. Resister capacitor combinations that are much better able to filter out noise and only allow low voltage signals through to trip limit switches. These can be well designed by someone who actually knows what they are doing and is able look at a setup with an Oscope and knows how to use the info. (That's not me.)

3.) Add capacitor across switch contacts. (Again, I am likely wrong here.) Effectively cuts lengths of long signal lines in half making them bad antennas. My understanding is that when you ground the drain cable for your cable shielding you enhance EMI resistance but make your RF resistance worse. I think you are coupling the grounding and this placement of capacitors for a better result.

Different machines have different problems needs. I think machine size, build, environment, and a whole lot more make this one of the toughest DIY problems you will ever love.