Home › Forums › TinyG › TinyG Support › TinyG edge drifts on series of small movements
Tagged: drift, pcb, reproduced
- This topic has 68 replies, 6 voices, and was last updated 10 years, 5 months ago by alden.
-
AuthorPosts
-
May 15, 2014 at 5:58 pm #5914cmcgrath5035Moderator
FYI, should you for some reason want to try 380.08 again, try tgFX build 3009, that combination worked OK for me before I moved to tinyG 412.01 and tgFX 3256. You can find build 3009 here
My GCode simulator took a while to import your files, but I was able to quickly see a lot of similarity to Flux’s test cases – significant areas of sequential very short segments (both XY and ARC).
My best guess is that you will benefit from tinyG build 426.02 when Alden makes it available.
Flux had better luck on some of his test cases with 380.08, but not all.Also, have you tweaked your stepper currents with tinyG? Since all this worked with GRBL, we know the rest of the machine is good.
When I first ran Flux’s test code, I found I was running at too low a drive current for the degree of motion he had. That led to extra “drift” in early runs.Also, files this long with the degree of detailed motion translates to heat – do you have a good fan blowing on tinyG (or equivalent thermal control)?
- This reply was modified 10 years, 6 months ago by cmcgrath5035.
May 16, 2014 at 12:54 am #5916datawayMemberThanks for your review.. I have heat sinks and two fans installed, one blowing directly down on the board and one blowing across the board as I have 3D printed a stand that suspends my tingG controller up 1 1/2″ in my enclosure so I don’t believe heat is an issue. As far as tweaking the currents, I did try that after reading through this thread and did some more testing but still encountered the drifting. I am hoping Alden makes 426.02 available soon as I have a few jobs backed up waiting to be milled. I will try 380.08 again with the build 3009 of tgFX see if that helps.
Thanks you again for your support…
John
May 19, 2014 at 6:04 pm #5929datawayMemberJust looking for an update as to when the release of 426.02 will happen to fix this issue …
Thanks for any update….
May 19, 2014 at 9:21 pm #5932aldenMemberI’ll get this posted to edge over the weekend. I’m traveling this week following up from Makerfaire.
In the mean time you can pull that release from dev and build it if that helps.
May 19, 2014 at 11:43 pm #5934datawayMemberGreat Thank you Alden hope Makerfaire was successful…Safe Travels…
May 24, 2014 at 3:49 pm #5953datawayMemberHello Alden,
Will you be building the hex file and making it available on your download site, I have attempted to build it from the DEV branch but I am running into issue loading the new build on my tinyG, I am sure it is something with my building of the file..
Thanks
John
May 24, 2014 at 10:17 pm #5954aldenMemberI have promoted dev to edge and posted the hex for build 429.01 on the synthetos.github.io download page. Please see if this is what you need.
May 25, 2014 at 7:36 pm #5957datawayMemberThank you Alden that is what I needed and it loaded just fine on my tinyG and am running a test now to see if the issue has been resolved.
I will update the thread with my findings..
Regards,
John
May 26, 2014 at 3:33 pm #5961datawayMemberwell everything seemed to be working good until the third pass on the roughing stage where the Z Axis seems to loose its mind and took a deep dive into the wood. I have tried using the generic gcode Post Processor and the tinyG mm Post Porcessor with the same results. When running the gcode for the roughing phase i get through two passes then on the third pass it starts out ok but then the Z Axis dives down the depth of the material, .75″ and I kill the job.
gcode :
https://dl.dropboxusercontent.com/u/3848551/Roughingtinygmm.taptinyG settings:
https://dl.dropboxusercontent.com/u/3848551/tinyg%20settings%2052614.txtPicture showing issue:
https://dl.dropboxusercontent.com/u/3848551/zdive.jpgMay 26, 2014 at 6:39 pm #5968aldenMemberLet me run this with coolterm to make sure it’s not on the board itself. An axis losing it’s mind points to a communications error, but I’d like to rule out everything else. Can you air-cut this with coolterm? Use Con/Xoff for flow control.
May 26, 2014 at 9:35 pm #5969aldenMemberI created a numbered version. Can tell me roughly where you see the plunge?
https://www.dropbox.com/s/815g3cu96bi5qw3/Roughingtinygmm.NUMBERED.txt
…and sorry if you have already said this, but what are you using to send the file, and what communications settings are you using – flow control, etc.
Also, how big is your work area? Mine hits the edge. I have an old Shapeoko.
Thanks — Alden
- This reply was modified 10 years, 5 months ago by alden.
May 26, 2014 at 10:07 pm #5971RileyKeymasterData,
Yah I would like to know where we are “losing it at”. I will run your gcode tomorrow. Can you describe the fail? Its always are to see what “should be happening”.
We typically can isolate if this is a issue with tgFX or TinyG in general by having you run this though coolterm >> via send file (make sure CTS is enabled). If it still messes up then its a TinyG issue. If not then its a tgFX issue.
Sorry you are having issues and lets get some solid data to help get you working!
ril3y
May 27, 2014 at 11:03 am #5975datawayMemberHello Alden I am using tgFX build 3256 and am using RTS/CTS I also tried with JCNC also set to RTS/CTS and both parameters for connection were the same set at Baudrate 115200, Data Bits 8, Parity NONE, Stop Bits 1.
I am using a shapeoko 2 with a work area of 11.25″ x 11.25″, I will run through coolterm today to see if i can determine where in the gcode it is failing and report back.
Thanks guys for all the help it is much appreciated, I am able to execute very simple jobs that only include profile tool paths but when i get into 2.5D roughing and finishing tool paths that is where I see issues.
I am loving tinyG and tgFX and if we can get this resolved I will be in CNC Bliss…
May 27, 2014 at 2:40 pm #5976datawayMemberThe issue happens at line 4193 when it starts to make the cut for the barrel edge design. It is running on coolterm now just waiting for it to get to that area of failure.
John
May 27, 2014 at 2:47 pm #5977datawayMemberHere is a rendered image with a red arrow pointing to the area where the failure is happening
also here is a link to the image:
-
AuthorPosts
- You must be logged in to reply to this topic.