Now to my problem (or problems as I have two which might be connected):
sometimes when jogging the axis will stop (in the last time its mostly the z-axis) and will only be able to move after I switch cables back and forth or restarted my computer.
The last two (and also my first) jobs ended in the router going rogue, going in crazy directions, so as the directions got lost…
Questions
a) when I touch the connection / cables I get “feedback” from the stepper, meaning maybe that the cables are not correct? Making sounds like “tak tak tak”
b) Could this be also attributed to the usb computer connection?
Just a shot in the dark, but I’ve seen some similar issues on this forum traced to electrical noise. Do some dry runs (no milling) both with the router running and not running. If the problem occurs, see if rebooting the CNC shield without rebooting the computer solved the problem. Swap USB cables and see if that solved the problem.
Unrelated, but some (all) of the Nano versions of the CNC shield have a wiring issue with the microstepper jumpers. One side is wired to ground rather than 5V, so the board will only do full steps unless modified.
the reason for my question around firmware, is I have had problems with some versions of GRBL dealing with the arc gecode commands in a weird way that results in the machine going all over the place.
the solution was to tell the CAM software to not use arc commands in the g-code.
If its the arc’s thing in GRBL, that I ran into, it will do straight lines fine, but go nuts when you try to do a curve.
i had to change my cam settings to make curves a series of short straight lines in the g-code and the problem went away.
So I did some testing.
I changed the cnc-shield, so far that seems to be solution for the constant hassle with stopping and restarting of the stepper motors.
I still have the problem that the router will not go the paths he should, although in Estlcam he shows to be on the correct path. So it might be electrical noise ( I have a Makita 0700).
If you open Estlcam, click on Setup in the top menu bar then ’ CNC Programs’ there are a couple of check boxes in the bottom left, uncheck the top one “Use arc commands G02/G03”
I changed the settings in Estlcam to the ones you proposed and also Ryan has on his site as I saw.
Made a new gcode and… still the same.
BUT I realised that I might have a different underlying issue.
I looked closely to the job and realised that sometimes the router stopped for a short period and after some jerks start over. Might be because of that that he losses his position.
Made some tests with just moving the router, and it seems as if straight movement will go mostly fine, but at some points the Y-axis stopps travelling and will only resume after I travelled a bit to the left or right on the x-axis.
after more extensive testing, it currently seems as as if my cable extensions are crab
I got back to the stepper cables which barely fit to the cnc shield, but for now no problems so far…
Will do more testing, but either I am pretty bad a cables with dupont, or my cable itself is for the garbage
Having bad motor wiring would help explain the skipping steps and random stops, but in the OP you mentioned it also traveled off to a random location as well… is it still doing that? The random position error is usually related to problems with the host/usb comms (another usb cable?), but it can also happen if gcode has an arc command and grbl does not have arcs enabled.
Can you post your grbl parameter list?
The fact that the y stops and will restart after moving x… to me sounds like a cable issue. That’s the trouble with diy cnc… we have like 100 connectors that can fail, and a lot of times those are non-industrial crap manufactured connectors (duponts made in China). Lord knows I have had more than my fair share of those fail. Anymore I avoid duponts like the plague when building a new nc machine (printer, cnc, sandbot… etc). I try an get board where everything is jst and screw terminals… and I always make my own harnesses with handcrimp jst and ferrules (no jst connector extensions etc to fail). Maybe it is overkill, but since doing that I have not had a single problem related to wiring.
I throw away all my self-made cables and made three dry runs, which went smooth as possible, without any stops and the bit stopped at the same position he started.
I have quite some hope now-
Unfortunately, I cannot use the router in “real life” without cable extensions, but will get new ones next week and will do then testing with real cutting
It seems that the problem with the router going rogue had something to do with the wiring as well, because of whatever problem the wiring was causing, now its gone.
I look forward to next week.
Thanks for your hands up!
Arne
Wiring problems are never fun. And they are never fun to fix. But they are often the issue and can cause some very funny behavior. I’m glad you’ve gotten some good results.
Awesome you seem to have found the problem! Honestly I have a rep among my friends for putting out a lot of “expert level projects” that have rats nest wiring.
I finally gave in and listened to some of them who swore I would benefit by changing my habits with wiring things. I bought all the DuPont and jst connectors, crimpers, and wiring to do signals and motor wires properly. I even invested in cable raceway for extra point lol. Any system that had caused me problems in the past (printers, Cnc, etc) eventually got new harnesses built using those tools… staying anal about redoing crimps that don’t go perfectly. Also a little secret of mine… add a film of dielectric grease to the pins before connecting them. Since doing this my maker headaches have reduced drastically!
So I urge all like minded “electrons don’t care what the wire looks like” folks to reconsider how much easier life can get if you just get the crimper stuff, learn how to use it properly, and use it every time. Also dielectric or noalox are I think underutilized on our machines. No way some over seas manufacturer is going to make sure your connectors are done the same… trusting those Amazon wires will lead to headaches later on, guaranteed.
Agreed. A lot of my electronic projects started to go together more smoothly when I started paying more attention to the wiring.
You’d think I’d have learned my lesson after all the proto-boards I had to build in college. It never failed that the more time you took running the wires on the proto-board the better the project worked and the easier it was to diagnose issues.
