Recent Build of a Kit - 2 wire limits

[LRX]

Hi, I'm new here - but old to the electronics industry and age too

I built up my kit earlier this year, and as usual, I did my own thing.

The reason for this is that there were a few things I wanted to try, such as 2 wire limits - I've never done 3 wire on anything I've made, and I have made a bunch of things - a few of which were not rubbish.

Anyway, If anyone's interested, I could detail my build - will try, time is not my friend. (The machine is very stable (Un-shielded) - which goes against most of the stuff I have read to date.)

No, shielded cables (anywhere), 2 wire limits, no shielded limit wire - I used 4 core "cabtyre" (anybody familiar with that term?) for the motors and 6 core flat cable for the limits, vac pump, led's and solenoid - flat cable rolls nicely. I wanted as little resistance as possible with the Cat tracks. Oh, and 2 wire limits need a small tweak to the tinyG config file.

I also added 2 e-stops - though I have not needed them for a while But I do bump them accidentally from time to time which makes me crazy.
Some simple filtering on the reset line is needed to prevent spikes from resetting the tinyG.

The environment my machine is in is pretty electrically noisy, no, very noisy as there are several full time welders welding aluminium and steel below my work area. So I figure if it's stable under these conditions, it should be stable anywhere!

I have not yet mastered the optics of black and clear tape - still have failures there - but I'm working on it.

So that's where I am.

BTW: This is an awesome kit to build - very well thought out and put together.

[JuKu]

Cool! I recommend shielded cables and hard (direct) pull-up of limit switches, as in my environment and my cabling, the machine is not reliable without. And if I can't get it to work without, I guessed majority of the customers couldn't either.

[LRX]

The Machine

IMG_6914.JPG (118.48 KiB) Viewed 4243 times

The Head

IMG_6920.JPG (134.81 KiB) Viewed 4243 times

My Enclosure

IMG_6918.JPG (139.58 KiB) Viewed 4243 times

Close up of "flat wire"

IMG_6919.JPG (143.34 KiB) Viewed 4243 times

Some Wiring

IMG_6917.JPG (145.95 KiB) Viewed 4243 times

Here's my thoughts on interference issues.

I'm pretty sure it was because all my parts are earth together & the base is metal that I do not have interference problems.

Over the years I have noticed that any conductive parts in close proximity that are not physically connected together (Well, as in low impedance) create havoc with electronic systems.
Sort of like a big antenna array Or giant capacitively coupled earth loop. Or even Vandergraaf generator
(The 2nd image shows the earth cable on the right hand side.)

My initial thought was to connect the common earth to the negative rail, but this did give some issues - so ultimately it ended up in the power earth - after that, there were no more issues. Though, I'm not sure that there ever were issues, because at that time I was still learning about the machine - so I thought many things it did were glitches.

I also mounted plug&socket phoenix type connectors on the Z axis to make it east to connect up. You can see the limits are just connected with ribbon cable to the phoenix connector where where they then change to the flat cable.

Another thing that comes to mind is that I put a back emf diode on the solenoid at the solenoid so that the back emf did not make an inductive loop via the diode at the fets. If this were allowed to occur, the emf generated at switch off of the solenoid (probably several 1000v) would be shorted by the diode at the fets. That would mean that the energy would be dissipated in the 2 wires running to the solenoid, which means it would be induced in adjacent wires. I also made a small PCB with the 2 FETS on it. You can see it in the 2nd image below the tinyG board.

3rd image shows the ESTOP wire going to the tinyG. My 'fancy' filter was a resistor in both legs of the estop, this is to prevent induced noise from making it back to the tinyG - on either line, and a cap between the 2 wires on the switch side of the wire, i.e. the cap is in parallel with the switches (2 off) at the tinyG, then connected to the reset via a resistor in each line.

I work extensively with Atmel, and like any micro, they hate any noise on their reset line, especially negative going spikes (below gnd). So at all costs, you need to keep noise away from the device. Actually, I was quite surprised that it worked, I expected constant resets. I think an opto coupled ESTOP would be a much better plan.

Other things to consider are short and thick wires from the power supply to the TinyG - the motors pull heavy instantaneous current - so they will create havoc on the supply if there is any inductance (long wires / thin) between the board and the PSU. You can mitigate this somewhat, by putting big reservoir caps at the supply input to the board if you have long supply lines.

Looking at the images you can see the green earth wire that commons everything.

And that's my story in a nut shell,

[JuKu]

> I'm pretty sure it was because all my parts are earth together & the base is metal that I do not have interference problems.

I have that too, but I don't have a metal base.

> My initial thought was to connect the common earth to the negative rail, but this did give some issues - so ultimately it ended up in the power earth - after that, there were no more issues.

Hmm. You have something here, I'm sure!