A Little more advice needed!

[Picengraver]

Here is an image that I pulled from the web that shows how to install capacitors on limit pin inputs. You will want to do the same for all the input pins you have in use.

uno caps.jpg (135.71 KiB) Viewed 1918 times

Brushed DC motors can also be a source of EMI, so you might also benefit by adding caps to your spindle motor.

motor caps.jpg (36.16 KiB) Viewed 1918 times

EMI problems with Arduino Uno grbl installations are very common and can be difficult to resolve, and sometimes more than one source of EMI is present, so several "fixes" are often needed.

John

[willow]

Many thanks John, for the moment I am going with the 'ferrited USB cable' together with a powered USB hub - easy ones first (when they arrive that is) and I'll test each separately before combining the two to see what difference they make. If no success I will try the ceramic capacitors on the arduino board and motor, again separately - more to isolate the cause more than anything.

If all else fails, I will do a fresh install of Windows 7 (or maybe even XP) and see how that turns out - will keep you updated with outcomes as I test them. The pictures are great by the way - much better than trying to explain things... Should I also put caps on the Step/Dir, Probe and Analog input pins as well?

Regards, willower

[Picengravertoo]

This is why I did not post any pictures of caps wired into the UNO, or went into any further detail with it.

One of the reasons I purchased the Protoneer v3.51 cnc shield was the fact that they have amended their circuitry to stop noise effects on switches and probe lines - which I don't use (probe lines that is). They say you can use unshielded cables on home/limit switches. witness I have never had problems with homing or limits using this new shield.

Noise on those lines could now be an issue???

[Picengraver]

"Should I also put caps on the Step/Dir, Probe and Analog input pins as well?"
Not to my knowledge. Step/Dir and Analog are outputs. On the probe input can't hurt, but that is only active when probing, not when running a gcode file.

"Noise on those lines could now be an issue???"
Jeff, I agree with your thinking, but not knowing how effective the on-board filtering is, I figured it can't hurt to add caps also.

[willow]

Further update. I have today installed a powered hub to see if it made any difference (mainly because sorting through my Compter bits I found I already had a Belkin one here). Did it make a difference, it certainly did - it made the problem of Picsender 'hanging' far worse than it was before! Can't really work out why that should be - only perhaps that the stable voltage on the USB hub has had some effect on the possible EMI being generated to be more regular?

Have removed the powered hub now as I see it as a retrograde step at this point. That aside, my Tripp Lite Cable has not yet been delivered so won't be able to to try that until later this week. However, I have ordered a bunch of capacitors, and intend to apply these to the brushed spindle motor, which I think may be the source of the interference. More test test test and assess from there. If no improvement, from that point on, I intend adding caps to the Probe and Limits inputs and see what that does....

Will keep you folks updated with prgress as it goes along.

Willower

[willow]

Jeff/John, hopefully a final update to this saga.

Tripp Lite cable installed - but with only a moderate improvement, still hangs intermittently, but not as much as I have found to be 'Normal' of late.

Have since installed caps on inputs as described above and also installed them on the spindle motor at the same time. GREAT improvement!!! I have now completed my first project without incident - although motors still sounded a litte rough. Have now increased the microstepping to 1/8th stepping (from 1/2 stepping) and all still seems to be OK. I do have some doubts that there are alignment issues between drive screws and motors which are still to be resolved, but I have some flexible aluminium couplers on order which should clear that problem up.

I might even consider hooking up N/O switches for the Standard Abort/Resume etc. in the future, but I shall continue to assess this little machine a while longer before taking any further risks of possibly introducing other newer sources of EMI.

Many thanks for your support and guidance (not to mention patience!) throughout this exchange, I really was beginning to lose a bit of heart if I'm honest.

[Picengraver]

Good to hear Willower. CNC using an Arduino and grbl can be very frustrating at times, but the same can be said for other operating systems and controllers, too. All have similar problems, and similar fixes.

Please post pictures of your work here when you are ready. Will enjoy seeing them.
John