Help debugging hardware gone bad!

cvoinescu
Posts: 4442
Joined: Thu Jul 19, 2012 6:50 pm
Location: Camberley, UK
Contact:

Re: Help debugging hardware gone bad!

Post by cvoinescu » Wed Jan 27, 2016 11:28 pm

Static electricity is definitely a possibility, but you need a discharge into something connected to the breakout board, not just dust clinging to a hose. If you have limit switches, their inputs may share the same buffer chips as the motor outputs. It's enough to touch a limit switch wire, or even the limit switch itself, and cause a static discharge that damages a chip. This is more likely if there's a large source of static electricity nearby (the vac), but all you really need is dry air (common in heated spaces in winter) and a bit of static build-up from walking around.

The machine is made of several parts, which are insulated from each other by the V-wheels. The wires run close to some of these parts, so if you happen to touch the machine and discharge into it, the spark may jump to a motor wire or limit switch wire and fry your breakout board. This is also how static electricity from the vacuum hose could accumulate and cause a large discharge.

I would connect all the large metallic parts of the machine together and to ground (earth); that makes it less likely for a static discharge against the machine to affect the electronics.

Also, a touch probe input is particularly vulnerable, and I doubt the breakout board designers included extra ESD protection for it. Touch something connected to ground, to discharge yourself, before handling the touchplate. (This is good advice whenever working on the electronics, too.)
Proud owner of ShapeOko #709, eShapeOko #0, and of store.amberspyglass.co.uk

northbear
Posts: 252
Joined: Mon Apr 09, 2012 9:21 pm

Re: Help debugging hardware gone bad!

Post by northbear » Thu Jan 28, 2016 4:45 am

Static, hmmmm.

I do have limit switches including on the z axis where I noticed static from the vacuum so that could be possible. I am thinking to properly ground the entire machine I would need 4 wires. One to the z axis. one to the X carriage, one to the Y carriage and one for the base. Sound right? I could run these all to my grounded computer case (or would it be better to somehow bypass that and go directly to ground?) My guess is the highest priority would be grounding the z axis as that is probably the most likely source.

The V-wheels are insulated, but also my machine is made of anodized aluminum. Aluminum is conductive, but the anodized coating is not. Next chance I get I will take a continuity tester and see what is grounded and what is not. Plan would be to have one lead touching ground and the other touching various parts to see if they are grounded.
My buildlog is here

cvoinescu
Posts: 4442
Joined: Thu Jul 19, 2012 6:50 pm
Location: Camberley, UK
Contact:

Re: Help debugging hardware gone bad!

Post by cvoinescu » Thu Jan 28, 2016 10:12 am

You can connect the parts to each other, and just one of them to ground, rather than run four wires all the way.

Static discharge goes right through the anodic layer. The oxide may be a great insulator, but it has pores; at the voltages of ESD, there's little difference between anodized and bare aluminium.

Maybe someone with more electrical engineering experience than myself can weigh in and tell us how to better protect the limit switch wiring.
Proud owner of ShapeOko #709, eShapeOko #0, and of store.amberspyglass.co.uk

AnonymousPerson
Posts: 758
Joined: Sun Apr 26, 2015 1:16 pm
Location: 3753 Cruithne

Re: Help debugging hardware gone bad!

Post by AnonymousPerson » Thu Jan 28, 2016 10:27 am

cvoinescu wrote:This is also how static electricity from the vacuum hose could accumulate and cause a large discharge.
Milling plaster really kicks up the static problem with a vac. If anyone else intends doing plaster... be wary of the static issues. ;)
Shapeoko 3 #516

northbear
Posts: 252
Joined: Mon Apr 09, 2012 9:21 pm

Re: Help debugging hardware gone bad!

Post by northbear » Thu Jan 28, 2016 6:14 pm

cvoinescu wrote:You can connect the parts to each other, and just one of them to ground, rather than run four wires all the way.
Good point!
cvoinescu wrote:Static discharge goes right through the anodic layer. The oxide may be a great insulator, but it has pores; at the voltages of ESD, there's little difference between anodized and bare aluminium.
There goes my idea about using a continuity tester to see what parts of the machine are statically grounded!


The more I think/read about static in these types of situations, it seems like grounding the shop vac hose is probably the first most important thing to do. But now the question is what is the best/ proper way! Seems to be conflicting advice on if to put a bare wire tied to ground on the inside* or the outside** of the hose or some places recommend both!*** My concern a wire on the inside would be clogging. Another option would be buying a static dissipating hose.

Any thoughts on where the ground wire should be?

They have "dust grounding kits" **** But I am sure I can find these parts myself at the local Menards / Lowes / Home Depot store.

* http://www.waterfront-woods.com/Article ... static.htm
** http://images.meredith.com/wood/pdf/1st-page-art.pdf
*** http://www.leevalley.com/US/html/03j6112ie.pdf
**** http://www.rockler.com/dust-grounding-kit and http://www.woodcraft.com/product/812502 ... g-kit.aspx
My buildlog is here

northbear
Posts: 252
Joined: Mon Apr 09, 2012 9:21 pm

Re: Help debugging hardware gone bad!

Post by northbear » Thu Jan 28, 2016 6:22 pm

reread this link http://www.waterfront-woods.com/Article ... static.htm and from what they say the wire on the inside prevents static build up on the inside and wire on the outside prevents static on the outside (There is my duhh conclusion for the day -- your welcome :mrgreen: )
My buildlog is here

northbear
Posts: 252
Joined: Mon Apr 09, 2012 9:21 pm

Re: Help debugging hardware gone bad!

Post by northbear » Wed Mar 09, 2016 4:46 pm

Just a note, if you are looking for bare copper wire, Home Depot sells 18 ga by the foot, for $0.11 a foot . It is just bare stranded wire, so it is probably not quite as nice as the braided wire from Rockler, but inexpensive (and more convenient for me)
My buildlog is here

Post Reply