Help me identify the problem! [Solved :Defective GBRLshield]

Bluemetal
Posts: 13
Joined: Tue Apr 10, 2012 9:03 pm

Help me identify the problem! [Solved :Defective GBRLshield]

Post by Bluemetal » Fri Sep 06, 2013 3:25 am

I am starting to get Shapeoko #8 (initial Kickstarter group) with GRBL shield to work (after a long pause) and learning by trial and error. It has been able to write its name on previous tests.

Being English a second language, I could not find anything that resemble this behavior in the forum, so here is a video.



The defect starts at 1:41 on this video, but it will start in any other times on the same file so it is apparently not gcode based.

The default speeds (feed and seek) are set to 750 mm/min on Arduino, but in CamBam they are set to 400 (cut) and 200 (plunge). The stepper controls have heat sinks and do not seem to be overheating.

The start part of the gcode is below as a reference.

What is causing this strange behavior?

Code: Select all

( Made using CamBam - http://www.cambam.co.uk )
( Seal Trial 9/5/2013 10:38:39 PM )
( T1 : 1.0 )
G21 G90 G64 G40
G0 Z3.0
( T1 : 1.0 )
T1 M6
( 3DSurface1 )
G17
M3 S1000
G0 X-2.3115 Y-12.7
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X2.3115
G0 Z3.0
G0 X-3.9964 Y-12.3
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X3.9964
G0 Z3.0
G0 X-4.962 Y-11.9
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X4.962
G0 Z3.0
G0 X-5.9277 Y-11.5
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X5.9277
G0 Z3.0
G0 X-6.7134 Y-11.1
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X6.7134
G0 Z3.0
G0 X-7.2347 Y-10.7
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X7.1
G1 X7.2347
G0 Z3.0
G0 X-7.7559 Y-10.3
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X7.7559
G0 Z3.0
G0 X-8.2772 Y-9.9
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X8.1
G1 X8.2772
G0 Z3.0
G0 X-8.7985 Y-9.5
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X8.7985
G0 Z3.0
G0 X-9.2687 Y-9.1
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X9.1
G1 X9.2687
G0 Z3.0
G0 X-9.5756 Y-8.7
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X9.5756
G0 Z3.0
G0 X-9.8825 Y-8.3
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X9.7
G1 X9.8825
G0 Z3.0
G0 X-10.1895 Y-7.9
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X10.1895
G0 Z3.0
G0 X-10.4964 Y-7.5
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X10.3
G1 X10.4964
G0 Z3.0
G0 X-10.8033 Y-7.1
G0 Z1.0
G1 F200.0 Z-0.25
G1 F400.0 X10.7
G1 X10.8033
G0 Z3.0
Solution Update: It turns out the GBRLshield (V3) that came with the kit had some fault in it (not fatal so it would still run, but serious enough to prevent it from performing correctly). Replacing it with a new GBRLshield V4 and using the same exact settings as the previous one and same gcode, worked instantly. The motors got very hot but did not fail. Some small adjustment to the current and they were much cooler. After that I also changed the USB cable to one that had ferrite cores on both ends and separated communication from power for good measure. Run a 8 hour job on ( spindle running on wood and all) without any problems what so ever. Now I need to get proper bits! :)
Last edited by Bluemetal on Sun Sep 15, 2013 2:11 pm, edited 1 time in total.
Proud owner of original Kickstarter Shapeoko #8 - still has original steppers, some end plates and screws but now looks like a Shapeoko 3.:)

WillAdams
Posts: 8608
Joined: Mon Apr 09, 2012 6:11 pm
Location: Pennsylvania --- south of the Turnpike, East of US-15
Contact:

Re: Help me identify the problem!

Post by WillAdams » Fri Sep 06, 2013 10:45 am

Check V-wheels to ensure your gantry moves smoothly w/o racking (which I think is part of the problem).
Check belt tension --- (which I think is the balance) I don't agree w/ the initial suggestions for setting it, believe they should be tight enough to pluck like guitar strings --- I also prefer stronger, more adjustable belt anchors.
Consider dual-drive, either dual motor or a drive shaft
Shapeoko 3XL #0006 w/ Carbide Compact Router w/0.125″ and ¼″ Carbide 3D precision collets

bjbsquared
Posts: 102
Joined: Mon Nov 26, 2012 10:48 am

Re: Help me identify the problem!

Post by bjbsquared » Fri Sep 06, 2013 10:48 am

The default speeds (feed and seek) are set to 750 mm/min on Arduino, but in CamBam they are set to 400 (cut) and 200 (plunge).
I think you have wrong thinking here. Cut and plunge are feeds. I don't think the seek of 750mm/min is being over ridden by CamBam. You might try lowering the seek rate ($5?) and see if the problem goes away.

G0: Rapid move Uses the 750mm/min setting.

And/Or..

It sounds to me like the motor is skipping. Either it needs more current by adjusting the current limit on the motor drivers or it is binding and can't move for some reason.

Can it be moved by hand after it does this?
#1016 - 1M x 1M Y: Dual Motor Drive w/Looped Belts, X:Dual Rail with wlanfox blocks, Z: Stock, Spindle:DW660

WillAdams
Posts: 8608
Joined: Mon Apr 09, 2012 6:11 pm
Location: Pennsylvania --- south of the Turnpike, East of US-15
Contact:

Re: Help me identify the problem!

Post by WillAdams » Fri Sep 06, 2013 12:13 pm

bjbsquared wrote:It sounds to me like the motor is skipping. Either it needs more current by adjusting the current limit on the motor drivers
I played the video w/o sound (and the electronics mystify me), so that's probably the hot ticket (though checking the things I suggested can't hurt).
Shapeoko 3XL #0006 w/ Carbide Compact Router w/0.125″ and ¼″ Carbide 3D precision collets

Bluemetal
Posts: 13
Joined: Tue Apr 10, 2012 9:03 pm

Re: Help me identify the problem!

Post by Bluemetal » Fri Sep 06, 2013 3:08 pm

bjbsquared wrote:
It sounds to me like the motor is skipping. Either it needs more current by adjusting the current limit on the motor drivers or it is binding and can't move for some reason.

Can it be moved by hand after it does this?
The belts are tight and everything moves smoothly and firmly when power is cut off after the failure (it is held in place by the motor while power is on after the failure), so it is likely not a basic mechanical issue. Lowering the feed and cut speed did not make the problem go away and indeed they are not affected by CamBam settings.

Following the suggestion, I decided to play a bit with the current limit on the X-axis in the GBRLshield.

More current (clockwise) and it seizes immediately (locked whenever gcode is telling it to move - when power is cut then moves smoothly) upon start of the gcode. All the way counterclockwise and it does not move at all (free to move by hand though). I had to leave for work before I could fine tune it, but it seems to be caused by heat shutdown of the stepper control after a certain time/work, so starts loosing steps and skipping. I will try to fine tune it tonight and see if I can get it to behave.

This first kits ( this is #8) had slightly beefier steppers for X and Y (larger than the Z stepper) - apparently the new kits have all Nema 17 motors. The specs may be too tight for comfort on the GRBLshield.
Proud owner of original Kickstarter Shapeoko #8 - still has original steppers, some end plates and screws but now looks like a Shapeoko 3.:)

WillAdams
Posts: 8608
Joined: Mon Apr 09, 2012 6:11 pm
Location: Pennsylvania --- south of the Turnpike, East of US-15
Contact:

Re: Help me identify the problem!

Post by WillAdams » Fri Sep 06, 2013 3:26 pm

Bluemetal wrote:it seems to be caused by heat shutdown of the stepper control after a certain time/work, so starts loosing steps and skipping. I will try to fine tune it tonight and see if I can get it to behave.
In that case, I'd point a fan for some active cooling at your stepper drivers --- my machine wouldn't work past the first few steps until I placed a 24v fan over the stepper drivers.

William
Shapeoko 3XL #0006 w/ Carbide Compact Router w/0.125″ and ¼″ Carbide 3D precision collets

Bluemetal
Posts: 13
Joined: Tue Apr 10, 2012 9:03 pm

Re: Help me identify the problem!

Post by Bluemetal » Sat Sep 07, 2013 1:58 am

WillAdams wrote:
In that case, I'd point a fan for some active cooling at your stepper drivers --- my machine wouldn't work past the first few steps until I placed a 24v fan over the stepper drivers.

William
There is clearly a relationship between the failure and the current settings, but I believe the solution may not be as simple as active cooling. I have printed a support for a 40mm 24V fan to cool the shield as can be seen here:
Active Cooling 1.jpg
Active Cooling 1.jpg (196.69 KiB) Viewed 1852 times


Before that I played further with the currents and managed to get to about line 900 of the 7000 line gcode file I have been trying to use - the previous record was around 180 or so. I took infrared pictures of the shield before and after the cooling following a failure:

Before ( ~ 46C )
IR_0212.jpg
IR_0212.jpg (37.19 KiB) Viewed 1852 times
After (~ 30C ) - the bright spot on the lower right is the LED.
IR_0214.jpg
IR_0214.jpg (35.77 KiB) Viewed 1852 times
After the cooling I have not been able to get past line 900 either. Considering the low temperatures to begin with, the fact that the motors ( Lin Enginnering 5718X-05E-03 rated @ 1A ) and that the shield can can handle 2.5A per winding, I am running out of options.

PS: I could not get links to Flickr photos to work as the Youtube one did earlier, so I downloaded the maximum 3 allowed.
Proud owner of original Kickstarter Shapeoko #8 - still has original steppers, some end plates and screws but now looks like a Shapeoko 3.:)

bjbsquared
Posts: 102
Joined: Mon Nov 26, 2012 10:48 am

Re: Help me identify the problem!

Post by bjbsquared » Sat Sep 07, 2013 12:06 pm

Is it always the same axis?

Make sure the power cable to the spindle is far away from the USB and Motor cables.
#1016 - 1M x 1M Y: Dual Motor Drive w/Looped Belts, X:Dual Rail with wlanfox blocks, Z: Stock, Spindle:DW660

Improbable Construct
Posts: 997
Joined: Tue Apr 10, 2012 3:21 am
Location: Fairhope, AL
Contact:

Re: Help me identify the problem!

Post by Improbable Construct » Sat Sep 07, 2013 2:37 pm

I am going to suggest my standard answer when something like this happens...

Check all of your connections to that motor.
A loose stepper wire will cause that kind of behavior.
It may look fine but as soon as it starts to loose connection the motor will stutter and sometimes even go backward.
Shapeoko #Classified some of the bolts may be original parts.
Shapeoko 1 # ???? Stainless plates, still in the box.
Shapeoko 2 # 3926 not stock
Shapeoko 3 # 0003
Store:
http://ImprobableConstruct.com
Twitter:
https://twitter.com/ImprblConstruct

Bluemetal
Posts: 13
Joined: Tue Apr 10, 2012 9:03 pm

Re: Help me identify the problem!

Post by Bluemetal » Sat Sep 07, 2013 3:12 pm

bjbsquared wrote:Is it always the same axis?

Make sure the power cable to the spindle is far away from the USB and Motor cables.
It is usually the X-axis given that it is the one doing most of the movement, but on occasion it happens on the Z-axis ( the second one doing most of the work). The Y axis moves most on the first few lines and then just the increments as the X is traveling the whole width of the par. The spindle has not been connected to power yet - I am just getting the machine to move according to the code.
Improbable Construct wrote:I am going to suggest my standard answer when something like this happens...

Check all of your connections to that motor.
A loose stepper wire will cause that kind of behavior.
It may look fine but as soon as it starts to loose connection the motor will stutter and sometimes even go backward.
I will recheck everything again but the fact that sometimes it stops mid Z movement too may indicate something more sinister. I just read your signature and realized that my Shapeoko ( #8) is almost as original as the day Robert shipped it. The only changes are the 3D printed support for the Z-axis. Maybe that setup was not ideal.

I will try to measure the voltage on the potentiometers to get a more precise measure of what is going on when I adjust the currents. Pololus are so much simpler to set up as you can measure while adjusting by the metal screw driver as the positive probe.
Proud owner of original Kickstarter Shapeoko #8 - still has original steppers, some end plates and screws but now looks like a Shapeoko 3.:)

Post Reply